A Syntactic Study

A Syntactic Study
of Idioms
A Syntactic Study
of Idioms:
Psychological States in English

and Their Constraints
By
Anna Dąbrowska
A Syntactic Study of Idioms:
Psychological States in English and Their Constraints
Reviewed by:
Bożena Cetnarowska
(University of Silesia in Katowice, Poland)
Henryk Kardela
(Maria Curie-Skłodowska University in Lublin, Poland)
By Anna Dąbrowska
This book first published 2018
Cambridge Scholars Publishing
Lady Stephenson Library, Newcastle upon Tyne, NE6 2PA, UK
British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library
Copyright © 2018 by Anna Dąbrowska
All rights for this book reserved. No part of this book may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any means,
electronic, mechanical, photocopying, recording or otherwise, without
the prior permission of the copyright owner.
ISBN (10): 1-5275-0616-9

ISBN (13): 978-1-5275-0616-9
TABLE OF CONTENTS
Acknowledgments ...................................................................................... ix
List of Abbreviations .................................................................................. xi
List of Tables ............................................................................................ xiii
List of Charts ............................................................................................. xv
Introduction ................................................................................................. 1
Chapter One ................................................................................................. 7

Towards Defining an Idiom
1.1 Introduction .................................................................................... 7
1.2 The definition of an idiom ............................................................. 7
1.2.1 Dictionaries and linguistic encyclopaedias as the sources
of definitions of the notion of idiom .......................................... 8
1.2.2 Linguists’ and scientists’ definitions of an idiom ................... 9
1.3 The characteristics of idioms ....................................................... 16
1.3.1 Metaphoricity / figurativeness ............................................... 17
1.3.2 Analysability vs. non-compositionality ................................ 36
1.3.3 Fixedness of form and internal structure .............................. 40
1.3.4 Literalness, familiarity and predictability of idioms.............. 45
1.4 Hypotheses and models of idiom representation and processing ... 54
1.4.1 Non-compositional Models ................................................... 54
1.4.2 Compositional Models .......................................................... 57
1.4.3 The Hybrid Approaches ........................................................ 60
1.5 The working definition of an idiom ............................................... 67
1.6 Concluding remarks ....................................................................... 70
Chapter Two .............................................................................................. 73

Syntactic and Semantic Characteristics of Psychological Verbs
2.1 Introduction .................................................................................... 73
2.2 Syntactic typology of psych-verbs ................................................. 74
2.3 The lexical-semantic representation of a verb................................ 77
2.4 Aspectual classification of psych-verbs ......................................... 80
vi Table of Contents
2.4.1 Basic event categories ........................................................... 81

2.4.2 Aspectual classes of psych-verbs .......................................... 87
2.5 Syntactic properties and syntactic analyses of psych-verbs ........... 99
2.5.1 Belletti and Rizzi’s (1988) unaccusative approach to OE
psych-verbs ............................................................................. 101
2.5.2 Landau’s (2005, 2010) locative approach .......................... 108
2.5.3. Fábregas and Marín‘s (2015) layer theory ......................... 121
2.5.4 Grafmiller’s (2013) account of psych-verbs ........................ 123
2.6 Concluding remarks ..................................................................... 127
Chapter Three ......................................................................................... 129

In Search of Psychological Idioms: A Corpus-based Account
3.1 Introduction ................................................................................. 129
3.2 Towards the specification of the selection criteria ....................... 130
3.2.1 Belletti and Rizzi’s (1988) tripartite classification
of psych-verbs ........................................................................ 130
3.2.2 Top frequent psych-verbs in the COCA Corpus ................. 131
3.2.3 The top psych-verbs vs. basic emotion domains ................ 141
3.2.4 The final selection of the psych-verbs ................................ 146
3.3 Psychological idiomatic phrases: A Corpus-based study ............. 147
3.3.1 The aim of the research ..................................................... 147
3.3.2 The working definition of idiomatic units reconsidered ..... 148
3.3.3 Data collection and methodology applied ......................... 148
3.4 Data analysis and preliminary discussion of the results .............. 151
3.4.1 Idiomatic units with an Experiencer in the subject position .. 154
3.4.2 Idiomatic phrases for OE (class II) psych-verbs.................. 159
3.4.3 Idiomatic phrases for OE (class III) psych-verbs ................ 168
Chapter Four .......................................................................................... 171

Psychological Idioms: Syntactic Constraints and Aspectual Properties
4.1 Introduction .................................................................................. 171
4.2 Classifying and characterising idioms ......................................... 173
4.2.1 Semantic dimensions of idiomaticity .................................. 173
4.2.2 Idiomatically combining expressions vs. idiomatic phrases .. 189
4.3 Syntactic and semantic variability of idiomatically combining
expressions (ICEs) with psychological meaning .......................... 193
4.3.1 Alternations affecting the idiomatic object NP in
psychological idioms .............................................................. 195
4.3.2 Alternations of the syntactic configuration of elements
in psychological idiomatically combining expressions ........... 200
A Syntactic Study of Idioms vii
4.3.3 Co-occurrence dependencies in psychological ICEs ........... 214

4.4 Constraints on the syntactic structure of psychological idioms.
Previous accounts ......................................................................... 219
4.4.1 Nunberg et al.’s (1994) semantic alternative
to the Hierarchy Constraint ..................................................... 219
4.4.2 O’Grady’s (1998) Continuity Constraint ............................ 221
4.5 The structure of psychological idioms in the Phase Theory ........ 227
4.5.1 Defining phases ................................................................... 227
4.5.2 The vP-phase boundary and psychological idiomatic
phrases.................................................................................... 234
4.5.3 DP-phases and psychological ICE-idioms with a possessor
in NP objects........................................................................... 241
4.5.4 Psychological idiomatically combining expressions (ICEs)
with Double Object Constructions in the Phase Theory ......... 249
4.5.5 PP-phases in psychological idiomatically combined
expressions ............................................................................ 258
4.5.6 Psychological idioms with particles .................................... 263
4.5.7 Small Clauses as Predication Phases in psychological
idioms ..................................................................................... 266
4.6 An Experiencer in psychological idioms ..................................... 274
4.6.1 The position of an Experiencer in psychological idioms ..... 274
4.6.2 An Experiencer in psychological idioms as a mental
location ................................................................................... 277
4.6.3 An Experiencer in Baker’s (1989) Thematic Hierarchy...... 281
4.7 Aspectual properties of psychological idioms ............................. 282
Summary and Conclusions ...................................................................... 293
Appendix 1 .............................................................................................. 299
Appendix 2 .............................................................................................. 305
Appendix 3 .............................................................................................. 317
References ............................................................................................... 319

ACKNOWLEDGMENTS
First and foremost, I would like to express my heartfelt gratitude to

Professor Anna Bondaruk, for her immense work and time devoted to me
while I was writing this book. She has not only read and re-read the earlier
drafts of the book, commenting upon the content of the text, but also
helped me with its form, giving invaluable pieces of advice and clues at
every single step. Her comments on the text were illuminating and in
numerous cases made me look at it from a different perspective. Much
more notably, she has shared with me her enthusiasm for the research on
both the syntax of idioms and psych-verbs, and this has proved to be
contagious. Professor Anna Bondaruk has helped me in so many ways,
providing me with her encouragement, empathy, and kindness.
Besides, my sincere thanks go both to Professor BoĪena Cetnarowska and

Professor Henryk Kardela for their comments and suggestions while
reviewing the book, without which the present study would not have been
completed. Warm thanks are also due to my colleagues at John Paul II
Catholic University of Lublin.
Finally, on a more personal note, I owe a special debt of gratitude to those

who are dear to me. My deepest gratitude goes to my family and friends.
For the time, advice, and friendship over these long years, special thanks
go to you. I am forever grateful for your love, patience, and everything
else you have given me. All of you have been so gorgeous to me since
your confidence and faith in me have never strayed.
All shortcomings in this work are, of course, my own responsibility.

LIST OF ABBREVIATIONS
A - Adjective
Adv - Adverb
AP - Adjective Phrase
AdvP - Adverb Phrase
ApplH - High Applicative
ApplHP - High Applicative Phrase
ApplL - Low Applicative
ApplLP - Low Applicative Phrase
COCA - The Corpus of Contemporary American English
CP - Complementiser Phrase
Det - Determiner
DOC - Double Object Construction
DP - Determiner Phrase
FI - Feature Inheritance
Gen - Genitive
ICE - Idiomatically Combining Expressions
IdP - Idiomatic Phrases
iF[val] - Valued interpretable features
LF - Logical Form
O - Object
OE - Object Experiencer
P - Preposition
PF - Phonetic Form
PP - Prepositional Phrase
Pr - Predication
PrP - Predication Phrase
SC - Small Clause
SE - Subject Experiencer
SEM - Syntax-Semantics Interface
Spec - Specifier
TP - Tense Phrase
uF[ ] - unvalued uninterpretable features
v - Light / Little Verb
vP - Light / Little Verb Phrase
V - Verb
VP - Verb Phrase
LIST OF TABLES
Table 1-1. Types of lexemic idioms (Makkai 1972: 135-169)

Table 1-2. Types of sememic idioms (Makkai 1972: 172-179)
Table 3-1. The ranking position of the top 13 psych-verbs on the
COCA 5,000 frequency word list (above the number of
10,000, ordered from highest to lowest)
Table 3-2. Top class II OE psych-verbs from Grafmiller’s (2013) study
and their ranking position checked against the COCA 5,000
frequency word list (with the number of tokens after
Grafmiller’s filtering)
Table 3-3. Types of syntactic patterns for SE psychological idioms
Table 3-4. Types of syntactic patterns for OE (class II) psychological

idioms
Table 4-1. Properties IdPs and ICEs (Corver et al. 2017: 10; cf. also
Harwood et al. 2016)
LIST OF CHARTS
Figure 1-1. Sample sentences with superficial phrase-structure trees

(Bever et al. 1969: 225)
Figure 1-2. Logical structure underlying the sentences in Figure 1.1
(Bever et al. 1969: 226)
Figure 1-3. Literal First Hypothesis according to Bobrow and Bell
(1973) (own source)
Figure 1-4. Swinney and Cutler’s (1979) Lexical Representation
Hypothesis (own source)
Figure 1-5. The Direct Access Hypothesis according to Gibbs (1980,
2002) (own source)
Figure 1-6. Cacciari and Tabossi’s (1988) Configuration Hypothesis
(own source)
Figure 1-7. The Conceptual Metaphor Hypothesis (Gibbs, Bogdanovich,
Sykes, and Barr 1997) (own source)
Figure 1-8. Titone and Connine’s (1999) Model of Dual Idiom
Representation (own source)
Figure 1-9. The rough field of multi-word expressions (cf. Mäntylä 2004:
38)
Figure 2-1. The lexical-semantic representation of a verb
(Alexiadou et al. 2004: 11)
Figure 3-1. Plutchik’s wheel of emotions: two-dimentional circumplex
model (Plutchik [1980] 2001: 349)
Figure 3-2. Syntactic patterns for SE psychological idioms (in
percentage)
Figure 3-3. Syntactic patterns for OE (class II) psychological idioms (in
percentage)
Figure 3-4. Syntactic patterns for OE (class III) psychological idioms
(in percentage)
INTRODUCTION
In the field of English phraseology, linguists have shown a constant

interest in idioms (cf. Knappe 2004: 3). Undoubtedly, not only are idioms
an important part of the language and culture of the society (Ji-Xin 2009),
but they also carry more impact than non-idiomatic expressions because of
their close identification with a particular language and culture (Nida
2001: 28). The linguistic units which are the core of interest in this book,
will be referred to as “phraseological units,” “fixed phrases,” “idiomatic
expressions,” “idiomatic phrases,” or “idioms” for short. Definitions and
various aspects concerning the concept of idioms are reviewed and
integrated into a framework which is rooted in the linguistic discipline of
phraseology.
Since the general tendencies of present-day English are towards more
idiomatic usage (Seidl and McMordie 1978: 1), indeed, it seems to be
worth paying attention to the role phraseological units play in a language.
Undoubtedly, it is difficult to speak or write English without using idioms
(Seidl and McMordie 1978: 4), especially while describing one’s
emotional or mental condition. In the same vein, Wierzbicka (1972) says
that, in contradistinction to thoughts which have a structure that can be
rendered by means of words, feelings do not have it. All a person can do,
therefore, is “to describe in words the external situations or thoughts
which are associated in our memory or in our imagination with the feeling
in question and to trust that our reader or listener will grasp what particular
feelings are meant” (Wierzbicka 1972: 59). Therefore, it seems that it is
interesting and worthwhile to make an attempt to analyse both the
language of phraseological units and emotions. In other words, this book is
to focus not only on idioms, but also on one’s psychological condition.
However, the aim of the book is neither to discuss the issues of idioms and
emotions from the psychological point of view, nor provide a conceptual
analysis of emotional metaphors. Instead, the objective of the book is to
analyse idioms referring to psychological states in English from the
perspective of syntax, focusing particularly both on the syntactic structure
of this specific set of verbal psych-idioms, and on the constraints on the
way they are built.
2 Introduction
For the purpose of the book, the recent compositional model of

idiomaticity, represented by Cacciari and Tabossi (1988); Gibbs, Nayak,
and Cutting (1989); Gibbs (1990); Cacciari (1993); Cacciari and
Glucksberg (1991); and Keysar and Bly (1995, 1999), among many others,
is adopted. The compositional model objects to the standard view of
idioms as non-compositional strings, typical of generative grammar (Katz
and Postal 1963; Fraser 1970; Katz 1973; Swinney and Cutler 1979; Gibbs
1980; and Machonis 1985; among others). Most idioms are viewed here to
be flexible and able to undergo syntactic and lexical modifications, in
contradistinction to a few totally frozen phraseological units. Hence,
following Nunberg, Sag, and Wasow’s (1994) taxonomy of idioms, two
types of idioms are distinguished, i.e. (i) idiomatically combining
expressions (e.g. pull strings “to use connections”), and (ii) idiomatic
phrases (e.g. kick the bucket “to die”). Idiomatically combining
expressions (ICEs) are referred to as decomposable / compositional or
analysable idioms, and they comprise idioms with a derivable idiomatic
interpretation (normally or abnormally, literally or figuratively). Idiomatic
phrases (IdPs), in turn, are known as non-decomposable / non-
compositional, frozen, opaque, or unanalysable, and include idioms with
an idiomatic interpretation not derived from their constituent parts.
Furthermore, this book follows the theoretical approaches according to
which the syntax of a verbal predicate and the range of syntactic
realizations of its arguments are determined by a verb’s semantic
representations (e.g. Croft 2012; Dowty 1991; Goldberg 1995; Jackendoff
1990; Langacker 1987; Pinker 1989; Rappaport Hovav and Levin 1998;
van Valin and LaPolla 1997; and Grafmiller 2013). Consequently, the
study relates to the syntax-semantics interface within which phraseological
units are to be better comprehended. The semantic dimensions of
idiomaticity, the event structure of verbal predicates, and their aspectual
properties are to be discussed, as well.
The syntactic study is based on the database of 161 English verbal
idioms which describe one’s emotional / mental / psychological condition,
and hence correspond to standard psychological verbs (psych-verbs), e.g.
to love, to fear, to annoy, or to matter to. On the basis of the well-known
studies represented by Belletti and Rizzi (1988), Pesetsky (1995), and
Landau (2010), among others, it can be specified that psych-verbs express
(a change in) mental or/and emotional state and a relation between the two
arguments: an Experiencer and the Cause / Theme of such a psychological
condition. Cross-linguistically and within different languages, psych-verbs
are classified similarly to the three-way division offered by Belletti and
Rizzi (1988), as in (0.1) below.
A Syntactic Study of Idioms 3
(0.1) Belletti and Rizzi’s (1988) tripartite classification of psych-verbs:

Class I: Mark loves bats. (SE psych-verbs)
Class II: The bats frightened Mark. (OE psych-verbs)
Class III: This film appeals to Joanne. (OE psych-verbs)
As shown in (0.1), an Experiencer can be realized as either a subject

(class I) or as an object (class II and class III). In addition, in the overview
of some crucial syntactic approaches to psych-verbs, the focus is laid on
Object Experiencer (OE) psych-verbs, and their “special” syntactic
properties, called “psych-effects,” revealed in their non-agentive reading
(cf. Belletti and Rizzi’s 1988 unaccusative approach to OE psych-verbs,
Landau’s 2005, 2010 locative approach, Fábregas and Marín’s 2015 layer
theory, and Grafmiller’s 2013 recent account of psych-verbs). Indeed,
what makes OE psych-verbs special and worth analysing is their aspectual
ambiguity, (between stative, eventive non-agentive and eventive agentive
reading) rather than their Experiencer argument (cf. Arad 1998, 1999;
Landau 2010; Alexiadou and Iordachioaia 2014; among others).
With reference to psych-verbs, the psychological idioms under
scrutiny, such as those in (0.2) and (0.3), are to become the object of
syntactic analysis carried out in this book.
(0.2) The examples of idioms and the SE (class I) psych-verbs they

correspond to:
a. Y loves X:
carry a torch for X
fall head over heels in love with X
have a soft spot for X
set Y’s heart on X
b. Y enjoys X:
paint the town (red)
raise the roof
have a ball
kick (up) Y’s heels
get a buzz out of X
(0.3) The examples of idioms and the OE (class II-III) psych-verbs they
correspond to:
a. X annoys Y (class II):
get the hump
raise Y’s hackles
get a rise out of Y
4 Introduction
put Y through wringer

give Y the pip
drive Y batty / nuts/ bananas / bonkers / crazy
b. X depresses Y (class II):
upset the applecart
dampen/damp Y’s spirits
cast a gloom / a shadow over Y
take the wind out of Y’s sails
cut Y down to size
give Y a bad / hard time/ the blues / the run around
bring Y low
c. X appeals to Y (class III):
float Y’s boat
whet Y’s appetite
set / put Y on Y’s ear
Importantly, both the psych-predicates and the psychological VP-

idioms to which the predicates correspond, comprise a participant who
experiences some emotional or mental state, i.e. an Experiencer (Y), and a
Stimulus / Causer / Cause / Target (X), which has contributed to this
specific state or become a target of it. The Experiencer (Y) may be situated
either in the subject position, i.e. in Subject Experiencer (SE) psych-verbs,
as illustrated in (0.2), or in the object position, viz. in Object Experiencer
(OE) psych verbs, as shown in (0.3). In short, the idioms in question are to
correspond to the psychological states referred to by psych-predicates.
Moreover, the bipolar division of idioms into IdP and ICEs (cf.
Nunberg et al. 1994; Harwood et al. 2016) is of much significance in the
syntactic study of psychological idioms, and the constraints on the way
these idioms are built, which is to be undertaken in the book. Even though
some previous analyses of idioms (e.g. Nunberg et al. 1994; O’Grady
1998; and Bruening 2010) are expected to be useful, not all puzzles of
idiomaticity can be resolved by relying on them. Therefore, the most
current research, performed within the scope of the Phase Theory and the
Idioms as Phases Hypothesis (cf. Svenonius 2005; Stone 2009; Harwood
2013, 2016, 2017; Harley and Stone 2013; Kim 2014, 2015; and Corver et
al. 2017; among others) is chosen to address certain syntactic problems
that idioms pose. As a result, the analysis of psychological idioms in the
light of the Phase Theory, provides some evidence for DPs, ApplHP, and
PrPs phases that can be formed in verbal idioms, in addition to vPs.
The book is organised in four chapters. The aim of Chapter One is to
present the definitions of an idiom, taken both from dictionaries,
encyclopaedias, and put forward by linguists (e.g. Pulman 1993; Gläser

1998; Knappe 2004; Mäntylä 2004; Carine 2005; Liu 2008; and O’Dell
and McCarthy 2010; among many others), and to discuss different semantic
dimensions of idiomaticity. Among the most crucial characteristics of idioms
presented here, there is idiom metaphoricity, idiom literalness, familiarity,
predictability, and idiom (non-)compositionality, among others. This
chapter deals with idiom taxonomies, models and hypotheses of idiom
representation and processing, as well, offered by Makkai (1972),
Nunberg (1978) and his followers, i.e. Gibbs and Nayak (1989) and Titone
and Connine (1999), but also by Cacciari and Glucksberg (1991), Sag,
Baldwin, Bond, Copestake, and Flickinger (2002), and the classification of
idioms by Yoshikawa (2008), to list just a few. For the sake of this book,
the bipolar taxonomy of idioms, offered by Nunberg et al. (1994), and
adopted by Harwood et al. (2016), has been adopted, in which idioms are
divided into idiomatically combining expressions (ICEs), and idiomatic
phrases (IdPs).
Chapter Two sheds light on predicates which denote a mental or
emotional condition, such as fear, love, worry, frighten, or surprise, which
a human participant (Experiencer) experiences. Consequently, this part of
the book is devoted to providing a brief, yet not truly comprehensive,
analysis of psychological verbs. The chapter opens with a discussion
concerning the fundamental syntactic, semantic and aspectual
characteristics of psych-verbs. Then, both the working definition of psych-
verbs, offered by Landau (2010), and the tripartite syntactic classification
of these predicates (class I, II, and III), proposed by Belletti and Rizzi
(1988), are presented as the ones adopted for the sake of the book (cf.
Dowty 1991; Pesetsky 1995; and Landau 2010). Chapter II also deals with
the event structure of psych-predicates (Vendler 1967; Dowty 1979;
Grimshaw 1990; and Alexiadou and Iordăchioaia 2014; among others). To
be precise, the discussion concerns the aspectual typology of class I-III
psych-verbs within the Lexicon-Syntax Interface, and the syntactic tests to
distinguish between stative, eventive non-agentive, and eventive agentive
readings of class II OE psych-verbs (cf. Arad 1998, 1999). Finally, this
chapter offers a brief overview of the syntactic approaches to psych-verbs,
most discussed in the literature, i.e. Belletti and Rizzi’s (1988)
unaccusative approach to OE psych-verbs, Landau’s (2005, 2010) locative
approach, Fábregas and Marín’s (2015) layer theory, and Grafmiller’s
(2013) recent account of psych-verbs.
The aim of Chapter Three is first and foremost to elicit psych-verbs
which constitute a representative set of this type of predicates; and then to
determine idioms which correspond to the psych predicates listed, and
6 Introduction
which are to be analysed in Chapter Four. To meet these objectives, the

methodology adopted in data selection is explained. It involves two stages
of the corpus study. In the first stage the most frequent class I and III
psych-verbs (cf. Belletti and Rizzi 1988), with the top occurrence in The
Corpus of Contemporary American English (COCA), are elicited; the top
psych-verbs of class II are adopted after Grafmiller (2013). In the second
stage of the corpus study, idiomatic phrases corresponding to the
psychological verbs just selected are searched. Following the convenient
typology suggested in Belletti and Rizzi (1988), the data are divided into
three classes, in the same way as their psych-verbal equivalents. The
elicited idioms are arranged according to the twelve syntactic patterns they
exhibit, while the exemplary sentences for those idioms, taken from the
COCA and/or obtained via the Google Search, are listed in APPENDICES
1-3. The search results are extensively commented upon.
Chapter Four lays the theoretical foundations for the analysis of
English psychological idioms, elicited in Chapter Three. One of the central
concerns of this analysis is to decide, in the light of the recent approaches
within generative grammar, which of the theories provides the best insight
into the syntactic rules idioms are governed by. With this in mind, the
chapter first deals with the semantic properties of both idiomatically
combining expressions (compositional) and idiomatic phrases (non-
compositional) (cf. Nunberg et al. 1994), with special reference made to
psych-idioms. Then, syntactic and semantic flexibility of psychological
idiomatically combining expressions is thoroughly discussed, with the
exemplary sentences, taken either from the COCA Corpus, obtained via
the Google Search, or tested against native speakers’ judgments.
Furthermore, Chapter Four deals with syntactic constraints imposed on
idiomaticity by the grammar. The most important approaches to the
behaviour of idioms are reviewed, viz. Nunberg et al.’s (1994) semantic
alternative to the Hierarchy Constraint, and O’Grady’s (1998) Continuity
Constraint. Finally, the syntactic structure of psychological idioms is
examined within the scope of the Phase Theory (cf. Svenonius 2005;
Stone 2009; Harwood 2013, 2016, 2017; Harley and Stone 2013; Kim
2014, 2015; and Corver et al. 2017; among others) to check the validity of
the phase-bound approach for the data analysed. Last but not least, some
space is devoted to the position of an Experiencer and the aspectual
properties the idioms under scrutiny reveal, with some reference made to
the aspectual structure of psych-verbs to which psychological idioms
correspond.
Chapter Four is followed by the final part of the book, viz. Summary
and Conclusions, which gathers the main points from all the four chapters.
CHAPTER ONE
TOWARDS DEFINING AN IDIOM
1.1 Introduction
The aim of Chapter One is to present the definitions of the notion of an
idiom and its characteristics. The chapter comprises five sections. Section
1.2 provides a wide range of definitions of an idiom, taken from
dictionaries and encyclopaedias, followed by the definitions put forward
by linguists. Section 1.3 discusses different dimensions of idiomaticity,
meant here as the characteristics of idioms. This part of the chapter opens
with an analysis of idiom metaphoricity, and the notion of metaphor and
figurative language. Then, various taxonomies of idioms, idiom processing
mechanisms, their metaphorical interpretation, as well as analysability and
ambiguity of idioms are studied. Besides, this section deals with idiom
non-compositionality, i.e. the fixedness of form and internal structure of
idioms. Finally, among the characteristics of idioms analysed here, there is
idiom literalness, familiarity and predictability, with the focus laid on the
role of context, well-formedness of idioms, and the level of their
formality. Section 1.4 concentrates on the models and hypotheses of idiom
representation and processing commonly referred to in the literature. In
section 1.5, the working definition of an idiom is established, on the basis
of the aforementioned dimensions of idiomaticity. Finally, section 1.6
summarises all the aspects discussed in this chapter.
1.2 The definition of an idiom

The study of idioms is generally considered problematic for the majority
of linguists. In his book Idiomatic Creativity, Langlotz (2006) argues that
“idioms are peculiar linguistic constructions that have raised many
eyebrows in linguistics and often confuse newcomers to a language”
(Langlotz 2006: 1). They constitute a “subset of the fixed expressions in a
language community” (Glucksberg 2001: 68), and on account of their
complex nature, idioms give rise to a broad range of definitions. Thus, it is
extremely difficult to provide a brief definition of an idiom, encompassing
8 Chapter One
all entities subsumed under this label. What is more, linguists have not
reached any solution in form of a unified approach and view related to
idioms so far, nor is it possible to offer in this book an explicit description
of what the term idiom refers to. Nevertheless, in this section of the book,
an attempt will be made to provide some clues as to how to define an
idiom. The starting point in the discussion is an encyclopaedic and a
dictionary definition, presented in section 1.2.1, followed by the concepts
of linguists and scientists regarding the definitions of an idiom, outlined in
section 1.2.2.
1.2.1 Dictionaries and linguistic encyclopaedias as the sources

of definitions of the notion of idiom
To begin with, the word idiom, dating back to 1565-1575, derives from
Latin idioma “special property,” and from Greek ੁįȓȦȝĮ–idiǀma, “special
feature, special phrasing.” As defined by McArthur (1992: 495) in The
Oxford Companion to the English Language, idiom means a combination
of words which have a figurative meaning owing to their common usage.
Meetham and Hudson (1969) in The Encyclopaedia of Linguistics,
Information and Control describe an idiom as “a habitual collocation of
two or more words whose combined meaning is not deducible from a
knowledge of the meanings of its component words and of their
grammatical syntagmatic relations to each other” (Meetham and Hudson
1969: 667).
Besides, in their book English Idioms and How to Use Them, Seidl and
McMordie (1978) stress that, even though some idioms may be completely
regular and logical, “an idiom is a number of words which, taken together,
mean something different from the individual words of the idiom when
they stand alone. The way in which the words are put together is often
odd, illogical or even grammatically incorrect” (Seidl and McMordie
1978: 4).
Additionally, Simpson and Weiner (1989), in The Oxford English
Dictionary on CD-ROM, rank an idiom as a smaller unit within language,
defining it as “a form of expression, grammatical construction, phrase,
etc., peculiar to a language; a peculiarity of phraseology approved by the
usage of a language, and often having a signification other than its
grammatical or logical one” (Simpson and Weiner 1989: sub verbo idiom
n. 3a). However, this definition comprises not only idiomatic
phraseological units and idiomatic word-formation products, but also non-
lexical “idioms” or typical grammatical constructions (cf. Knappe 2004:
14).
Towards Defining an Idiom 9
What is more, in his Longman Dictionary of Idioms, Hill (1990)

underlines the metaphorical rather than literal nature of idioms, and
maintains that “[t]hey are also more or less invariable or fixed in form or
order in a way that makes them different from literal expressions. Because
they are metaphorical, one cannot usually discover their meanings by
looking up the individual words in an ordinary dictionary” (Hill Long
1990: viii).
Finally, in her Webster's New World Dictionary, Neufeldt (1991) offers
a more detailed definition of an idiom which is seen as “a phrase,
construction, or expression that is recognized as a unit in the usage of a
given language and either differs from the usual syntactic patterns or has a
meaning that differs from the literal meaning of its parts taken together”
(Neufeldt 1991: 670).
In a nutshell, encyclopaedic and dictionary definitions of an idiom treat
it as a habitual unit of language, the meaning of which cannot be deduced
by summing up the meanings of its individual components. Instead,
idioms are fixed phraseological units by their long usage and have to be
learned as a whole.
1.2.2 Linguists’ and scientists’ definitions of an idiom

Being aware of the fact that providing a definition for the term idiom is a
challenging and difficult task, a great tribute should be paid to linguists
and other scientists who have approached this problem from various
angles throughout the history of language. Some of the definitions given
by the specialists will be provided now before the specific characteristics
of idioms are discussed.
First of all, an idiom is conventionally defined as “a complex expression
whose meaning cannot be derived from the meanings of its elements”
(Weinreich 1969: 26). And some decades earlier Willey (1939) formed a
definition, saying:
Idiom or idiomatic phrase (...) is a phrase the meaning of which cannot be

deduced from its component parts. The following are examples of
idiomatic phrases: to bring about (accomplish); to bring to pass; to carry
out (make effective, accomplish); to come by (obtain); to go hard with (to
be painful or harmful to); to put up with (tolerate, endure); to set about
(begin). An examination of these phrases shows that the meaning of each
(when used in its idiomatic sense) belongs to the phrase as a single
element, and is not a composite effect made by joining the meanings of its
parts. The peculiarity of such phrases becomes apparent if we compare
them with phrases that are not in this sense idiomatic; as, “to get to the
10 Chapter One
city,” “to sleep late in the morning” where every word has a meaning that
is contributory to the meaning of the phrase.
(Willey 1939: 221)
In addition, for Weinreich (1969), “any expression in which at least

one constituent is polysemous, and in which a selection of a subsense is
determined by the verbal context, is a phraseological unit (…). [Thus,] a
phraseological unit that involves at least two polysemous constituents, and
in which there is a reciprocal contextual selection of subsenses, will be
called an idiom” (Weinreich 1969: 42). Consequently, Weinreich’s
definition recognizes idioms as phraseological units or multiword
expressions which comprise lexical items that function with two or more
related meanings, i.e. they are polysemous. These meanings (“subsenses”),
contextually dependent, may be combined to form either a literal or an
idiomatic meaning (cf. Everaert’s 2010: 83 example of the idiom kick the
bucket “to die”).
Similarly, Lipka (2002) identifies idioms with phraseological units;
and consequently, he forms a definition as follows: “A phraseological unit
is a semantic unit consisting of a group of word-forms not beyond the
sentence level” (Lipka 2002: 89). This definition makes “phraseological
units” comparable to word-forms as concrete realizations of lexemes. On
the abstract level, a phraseological unit recognized this way may be seen
as the realization of a “phraseme,” while the word-forms may be seen as
realizations of “lexemes” (cf. Lipka 2002: 84, 89-90, 94-96). A “lexeme”
will be treated then as an independent sign on the abstract linguistic level
of the lexicon. It embraces both “simple lexemes” as morphologically non-
composite lexemes, and “complex lexemes” as morphologically including
more than one segment (morpheme or formative). In Lipka’s (2002: 89-
90) terminology, the notion of “complex lexeme” also covers “phrasal
lexemes” or “discontinuous lexical items,” which relate to the notion of
idiomatic phraseological units as understood here (cf. Knappe 2004: 6).
Going further, Lipka (2002: 87) introduces the term “lexematic formative”
to distinguish phraseological combinations of formatives (e.g. put up with
“bear, tolerate”) or those containing one or more of such formatives (e.g.
tit and tat in tit for tat “revenge”) from word-formation products
containing so-called “cranberry morphemes” such as Fri in Friday and
cran in cranberry. Yet, these also fall under the definition of formatives as
“minimal formal units without identifiable meaning” (Lipka 2002: 87).
What is more, Adkins (1968: 149) names idioms as modes of
expression or phrases which are peculiar to a given language, and which
are the basis for understanding the language, since they constitute a large
part of it. Wadepuhl (1928) comments that “any construction that could
not be translated literally from one language into the other has been
considered an idiom” (Wadepuhl 1928: 68). Having noticed that idioms
are hardly ever translated literally, Adkins (1968) adds that often the
dictionary is of little aid to provide the meaning of a particular expression.
Idioms have meanings different from the meanings of the words which
compose them. Moreover, idioms cannot be understood from the way they
have been formed, which has been exemplified by Adkins ((1968: 149) by
means of idioms such as, make a beeline for, meaning “to take the shortest
route,” and be short-handed, meaning “to have insufficient help.” Other
idioms are composed of verbs and prepositions, such as to fill in, meaning
“to substitute for” or “to complete the blanks on a form,” or built of verbs
and adverbs such as to look forward, meaning “to anticipate.”
Besides, O’Dell and McCarthy (2010) define an idiom as a fixed
expression whose meaning is not immediately obvious from looking at the
individual words in the idiom (cf. McCarthy and O’Dell 2002). Likewise,
Fraser (1970) treats idioms as multi-word phraseological units, whose
meaning is not predictable from their constituent parts, “I shall regard an
idiom as a constituent or a series of constituents for which the semantic
interpretation is not a compositional function of the formatives of which it
is composed” (Fraser 1970: 22; cf. Makkai 1972). In short, Fraser (1970)
underlines the fact that the individual elements of an idiom cannot provide
the overall meaning of the idiom. Correspondingly, Palmer’s (1986: 36)
view of a genuine idiom implies a phrase covering more than one word,
whose meaning is unpredictable from the individual idiom constituents.
Additionally, he notices that even though idioms behave like single words
in semantic perception, grammatically they cannot be recognized as such
units since idioms normally do not undergo changes (e.g. cannot form the
past tense).
Furthermore, Kavka (2003) discusses the expressions with idiomatic
nature in general, defining them as “multiword chunks consisting of
elements, or constituents, which are bound together lexically and
syntactically” (Kavka 2003: 12). Yet, he further makes a division of these
idiomatic expressions into collocations (that are semi-compositional) and
idioms proper (genuine idioms that are characterized with non-
compositionality and invariability), and mentions that idioms are “a non-
literal alternative from possible options of a literal interpretation” (Kavka
2003: 14, 25).
Then, the picture of an idiom as “a unique and fixed combination of at
least two elements some of which do not function in the same way in any
other combination (of the kind) or occur in a highly restricted number” is
presented by ýermák (2007: 142). Moreover, he stresses that anomaly is
12 Chapter One
one of the core characteristics of idioms since “the more anomalies a

phraseme displays, the more idiomatic it is and vice versa” (ýermák 2001:
7). Thus, according to ýermák (2007: 84), syntagmatic and paradigmatic
deviations are mainly emphasized as the chief features of idioms.
On the other hand, Nunberg, Sag, and Wasow (1994) underline idiom
unpredictability and conventionalism when they say, “Idioms are
conventionalised: their meaning or use can’t be predicted, or at least
entirely predicted, on the basis of a knowledge of the independent
conventions that determine the use of their constituents when they appear
in isolation from one another” (Nunberg et al. 1994: 492). Having
presented the examples of the idiom kick the bucket which means “to die”
and spill the beans “to reveal a secret,” they conclude that the meanings of
both idioms and their forms do not result from any basic grammatical
principle or from our knowledge of the world, but simply have to be
learned.
In addition, Fernando (1996: 1) treats idioms as multiword expressions
which are conventionalised and usually with a non-literal nature, though
not in all cases. Besides, she believes that expressions which demonstrate
a tendency towards higher variability may show idiomaticity, but they
cannot be considered as genuine idioms. Fernando (1996) states her stance
as follows: idioms are “indivisible units whose components cannot be
varied, or only varied within definable limits” (Fernando 1996: 30). And
to develop her definition of the term, she adds that “only those expressions
which become conventionally fixed in a specific order and lexical form, or
have only a restricted set of variants, acquire the status of idioms and are
recorded in idiom dictionaries” (ibid.: 31). Thus, as specified by Fernando,
the invariance of idioms is one of the best characteristics of idioms.
Furthermore, idiomatic expressions are often treated by linguists the
same way as lexical units which function as one semantic entity and have
one meaning. As explained by Moon (1998), when a multi-word idiom is
recognized as a unit of one single meaning, it is lexicalized. Lexicalization
is a “process by which a string of words and morphemes becomes
institutionalised as part of the language and develops its own specialist
meaning and function” (Moon 1998: 36). In this process, lexicalization
and institutionalization accompany each other indispensably, and a string
of words is not properly lexicalised if its meaning or function is not known
widely enough. As soon as the meaning and function of the expression
have become accepted and generally recognized in a language, the process
of lexicalization comes to an end, and then the idiomatic meaning
becomes institutionalized. Additionally, institutionalization requires a
certain amount of frequency in use. However, as Moon (1998: 7) points
out, most idioms are rather infrequent, i.e. they may be restricted to certain
registers and uses of speech, or to certain accents or dialects of the English
language. Schraw et al. (1988: 424) conclude that both lexicalization and
familiarity contribute to the likelihood of idiomatic preferences, while
only lexicalization contributes significantly to the comprehension of
idiomatic meanings.
Likewise, both Everaert et al. (1995: 3-5; 2010: 81) and Fernando
(1996: 2-3) define idioms, or phraseologisms, so-called in Polish linguistic
literature, as always conventionalised multiword expressions, characterised
by semantic opacity, i.e. the fact that the meaning of the whole is not the
sum of the components (cf. SzymaĔska 2008: 116). “Idioms are
conventionalized linguistic expressions which can be decomposed into
potentially meaningful components and exhibit co-occurrence restrictions
that cannot be explained in terms of rule-governed morphosyntactic or
semantic restrictions” (Everaert 2010: 81). Moreover, for Everaert (2010),
idioms include “all formulaic expressions including sayings, proverbs,
collocations” (Everaert 2010: 77). Similarly, Kjellmer (1994) considers
idioms as a type or subset of collocation, while others consider “restricted
collocations” (e.g. cardinal error / sin / virtue / grace) to be a type of
idiom (Cowie and Mackin 1975; Cowie, Mackin, and McCaig 1983).
Similarly, for Saberian (2011a: 1231), the term “idiom” has been used
to cover a wide variety of different types of multi-word units (MWUs),
which are treated as vocabulary items consisting of a sequence of two or
more words. These words constitute a meaningful and inseparable unit.
Yet, Grant and Bauer (2004) state that the term MWU refers to both
idioms as well as open and restricted collocations, excluding phrasal
verbs. However, for Grant and Bauer (2004), open collocations are the
freest kind of MWU, while core idioms are the most restricted ones.
Similarly, Aisenstadt (1979) argues that collocations differ from idioms as
“R[estricted] C[ollocation]s are not idiomatic in meaning; they do not
form one semantic unit; their meaning is made up as the sum of the
meanings of their constituents. They have a much greater variability and
usually occur in patterns with a number of interchangeable constituents”
(Aisenstadt 1979: 1).
What is more, according to Fillmore, Kay, and O’Connor (1988), some
conditions should be met to name a phrase idiomatic or not, since
“constructions may be idiomatic in the sense that a large construction may
specify a semantics (and/or pragmatics) that is distinct from what might be
calculated from the associated semantics of the set of smaller
constructions that could be used to build the same morphosyntactic object”
(Fillmore et al. 1988: 501). Furthermore, Fillmore et al. (1988: 506-510)
14 Chapter One
distinguish between substantive (lexically filled) and formal (lexically

open) idioms. Substantive idioms are lexically fixed (e.g. once upon a
time), while formal idioms as abstract patterns show special semantics
and/or pragmatics, and permit some lexical variation (e.g. the sooner the
better, i.e. “the x-er the y-er,” where x and y can correspond to various
adverbs or adjectives). Fillmore et al. (1988: 505) also distinguish
grammatical idioms (when words can fill expected places in grammatical
structures) and extragrammatical ones (with anomalous structures, e.g. by
and large “generally speaking”).
SzymaĔska (2008: 116-117) adds that it is grammatical idioms and
formal idioms that, from the point of view of Construction Grammar,
contribute profoundly to the most revealing insights into the mechanism of
form-meaning pairings or constructions (cf. Lakoff and Johnson 1980;
Lakoff 1987; Goldberg 1995; Fillmore and Kay 1995; Fillmore 2001;
among others). To be more precise, the basic assumption of Construction
Grammar, as referred to by SzymaĔska (2008: 111), is the fact that the
linguistic knowledge of a language user is best represented in terms of
constructions, i.e. language patterns “dedicated to some particular
semantic or pragmatic purpose” (Fillmore 2001: 36). Besides, SzymaĔska
(2008) points out that some idiomatic expressions may show certain
systematicity, and may be internally structured, becoming recognizable to
language users as semantically more constrained options of more regular
patterns. She also states that, from the perspective of Construction
Grammar approach, the fact that grammatical structures (including formal
idioms) convey meaning independent of lexical items may actually prove
linguistic creativity of the expressions in question (SzymaĔska 2008: 146).
Additionally, due to the fact that some idiomatic strings have both a
literal and a non-literal meaning; contextual clues appear to be helpful to
distinguish whether a given MWU has a literal or an idiomatic
interpretation. Alexander (1987) defines idioms as “multi-word units
which have to be learned as a whole, along with associated sociolinguistic,
cultural and pragmatic rules of use” (Alexander 1987: 178).
Furthermore, Langlotz (2006: 2) admits that the heterogeneity of
linguistic terminology surrounding idioms encountered by linguists is
really troublesome. That heterogeneity of idiomatic expressions stands in a
dialectical relation to the abundance of linguistic terminology developed to
capture and classify these constructions. Langlotz (2006) defines an idiom
as “an institutionalized construction that is composed of two or more
lexical items and has the composite structure of a phrase or semi-clause,
which may feature constructional idiosyncrasy. An idiom primarily has an
ideational discourse-function and features figuration, i.e. its semantic
structure is derivationally non-compositional. Moreover, it is considerably

fixed and collocationally restricted” (Langlotz 2006: 5). Accordingly, by
treating an idiom as a multiword conventionalized expression which is
non-compositional, with some irregularity, Langlotz highlights its function
to communicate experiences or events. However, he concludes that any
definition of idioms is never finite as their discursive functions mutually
overlap, leaving for each of them a unique “degree of idiomaticity”
(Langlotz 2006: 5).
Besides, Glucksberg (2001) categorises idioms as “a subset of the
fixed expressions in a language community” (Glucksberg 2001: 68), aside
from other fixed expressions, such as compounds, names, film and book
titles. He also points out that idioms are different due to “their ‘non-
logical’ nature, that is, the absence of any discernable relation between
their linguistic meaning and their idiomatic meaning” (ibid.). Glucksberg
(2001) further notices that not all idioms are fixed or frozen, “Some
idioms are syntactically flexible, appearing, for example, in both active
and passive forms. (...) Some idioms can also be modified internally. (...)
Semantic variations that make sense (...) are also permissible” (Glucksberg
2001: 68, 73).
Nevertheless, since idioms have been mostly referred to as fixed
expressions whose figurative meaning is not clear from the literal meaning
of their individual constituents, most authors – especially generative
grammarians of the early stage of idiom research – have made an essential
distinction between literal and figurative language (cf. Chomsky 1965;
MacKay and Bever 1967). Thus, from the generative point of view, the
fact that an expression is not interpretable in a literal way inevitably
denotes that it is a fixed (non-compositional) expression. More
pragmatically oriented studies use the terms “true idioms” (Wood 1986:
II) or “pure idioms” (Howarth 1998: 28) to refer to the “idiomatic = fixed”
relationship and allow other non-literal phrases to be idiomatic but non-
fixed at the same time. Yet, Abel (2003) argues that these assumptions
only apply to a subgroup of idioms and that it is more adequate to think of
idioms as being represented in a dual way that combines not only the
lexical but also the conceptual level of idioms, and integrates their
representation in the first language (L1) as well as in the L2 lexicon.
Finally, as noted by Knappe (2004: 7), the status of idioms as
phraseological units has been much discussed between lexicology, syntax,
and word-formation. Earlier transformational grammar had to face up with
the problem of integrating idiomatic phraseological units within its
system, which sharpened, as a consequence, the awareness of the various
grades of both lexical and syntactic characteristics for different kinds of
16 Chapter One
phraseological units. Hence, some idiomatic phraseological units have

received from cognitive linguistics the key arguments supporting the view
that those units may be coded in the mental lexicon in the same way as
lexical units are (cf. Dobrovol’skij 1997: 10). As proposed above, idioms
and their figurative meanings are recognised as the units stored separately
in the mental lexicon, in the same way as the meanings of individual
words are listed in a dictionary, and that this meaning must be learnt as a
whole (Becker 1975; Gasser and Dyer 1986; Wilensky and Arens 1980;
Gibbs 1989).
To sum up, this section has focused on providing a definition of an
idiom, with its all potential arrays of diversity. In spite of the difficulties in
forming an accurate definition of an idiom, it is possible to find some
general characteristics that most of the definitions share, namely: (a) some
subset of idioms has a fixed word order, which implies they have a
restricted set of variants, and (b) it is impossible to guess the meaning
from the individual words that make up an idiom. Moreover, dictionary
and encyclopaedic sources, evoked as the starting point, recognize an
idiom as a habitual unit of language, the meaning of which cannot be
deduced from its components. Thus, not only are idioms varied as far as
their forms are concerned, ranging from two-word expressions to entire
sentences, but their inconsistent semantic and lexico-grammatical
properties can also bring about opposing views concerning their nature.
Idioms then show an excessive array of diversity (Suþková 2010: 3). At
this stage, an attempt to point out the specific characteristics of idioms
seems to be crucial before a working definition of an idiom is offered to be
adopted in this book.
1.3 The characteristics of idioms

This section concentrates on features typical of idioms and their various
taxonomies. Taken for granted is the fact that the range and intensity of
literalness and figurativeness (analysed in section 1.3.1), and their mutual
relationship appear to be crucial while characterising idioms. However,
there are other features of idioms that have also been considered in the
literature, such as their metaphoricity (referred to in section 1.3.1),
analysability (see section 1.3.2), as well as their fixedness of form and
internal structure (discussed in section 1.3.3). Lastly, section 1.3.4 deals
with idiom familiarity and idiom predictability. The priority of some
characteristics over others has varied, depending on the linguist’s
emphasis put on the role an idiom plays within a discourse (Nenonen
2002: 6). Yet, as mentioned by Mäntylä (2004: 28), although some
features of idioms are more noteworthy than others, several elements are
required for an expression to be categorised as an idiom (cf. various
taxonomies of idioms, outlined in section 1.3.1.2). Still, there are
expressions that are more prototypical idioms than others, and sometimes
it is a mission hardly possible to distinguish idioms from other types of
fixed and / or metaphorical expressions, which can be easily noticed in the
discussion below. The first property of idioms to be embarked on in the
subsequent section is idiom metaphoricity and figurativeness.
1.3.1 Metaphoricity / figurativeness

Idioms have been functioning under the aegis of frozen and dead
metaphors (Weinreich 1969; Fraser 1970; Swinney and Cutler 1979; and
Cowie 1981; among others) for a long time until that viewpoint has been
re-examined in the past few years (Lakoff and Johnson 1980; Lakoff 1987;
and Gibbs 1990, 1992: 485; 1993: 57-61; among others). Then, “dead”
phrases implied forgotten metaphorical and arbitrary meanings with
undetected origins, while “frozen” used to mean fixed in form and limited
as regards most transformations and variations. Thus, learning them
entailed memorizing them as entities since the link between their form and
meaning has not been recognized. More recent works, with a great input of
psycholinguistic studies (cf. Fernando 1996; and Moon 1998; among
others), have found a great number of idioms, far from being dead or
frozen, but instead, marked with possible alterations, metaphoricity and
noticeable origins of their meanings.
1.3.1.1 The notion of metaphor and figurative language
Adkins (1968: 149) explains that a language which is not literal, often
employing metaphors, is called figurative language. While no attempt has
been made to classify figurative language, it should be noted that the term
“figure of speech” or “figurative language” covers such examples as
simile, metaphor, personification, and hyperbole.
ýermák (2001: 5) adds that the very notion of metaphor, dating from
Aristotle, appears to have rather unclear boundaries, and doubts whether a
metaphor can be treated as a reliable and general condition to characterise
an idiom. Lakoff and Johnson (1980), on the other hand, underline that the
metaphor is omnipresent in everyday life, in whatever thought, action or
language. They argue that “our ordinary conceptual system, in terms of
which we both think and act, is fundamentally metaphorical in nature”
(Lakoff and Johnson 1980: 4), and add further that “the English
18 Chapter One
expressions are of two sorts: simple literal expressions and idioms that fit
the metaphor and are part of the normal everyday way of talking about the
subject” (Lakoff and Johnson 1980: 46).
In addition, Mäntylä (2004) stresses that figurativeness (metaphoricity)
is one of the most commonly acknowledged features of idioms, and that
the roots of metaphoricity often derive from real situations or acts (e.g.
hang up one's boots), or an image created by the connection between the
idiom and its meaning (e.g. the fat is in the fire). She also makes a remark
that the reason why idioms have been recognized as “dead,” or arbitrary, is
that the relation between the origins of an expression and its meaning has
faded (e.g. kick the bucket, which means “to die”), or the literal context
belongs to a special field unknown to the ordinary language user (e.g. kick
something into touch, which means “to send the ball out of play”), making
the association very difficult to discern (Mäntylä 2004: 28-29).
Moreover, Horn (2003), providing a thorough analysis of idioms,
metaphors and their syntactic mobility, credits Jackendoff (1997), and
Nunberg et al. (1994), for introducing the term “a sort of metaphorical
semantic composition” (Jackendoff 1997: 168; Horn 2003: 246) to
describe mobile expressions. Jackendoff (1997) states that idioms having
this property “can be partitioned into chunks that correspond to the “sub-
idiomatic” readings of the syntactic idiom chunks” (Jackendoff 1997:
168). Fixed idiomatic phrases, in contrast, lack this property, as
exemplified by Horn (2003) by means of the well-known and much-
discussed fixed VP idioms, such as those in (1.1), and the examples of
mobile VP idioms, as in (1.2) below:
(1.1) a. Bill kicked the bucket. [Bill died]

b. We shot the bull all evening.
[We were engaged in trivial conversation]
c. The bad guys flew the coop. [The bad guys escaped]
(Horn 2003: 246)
(1.2) a. Fred spilled the beans. [Fred revealed the secret]
b. Bill let the cat out of the bag. [Bill revealed the secret]
c. The Government drew the line with Milosevic.
[The Government established a limit / made / enforced a
distinction]
d. Bill pulled strings to get the promotion.
[Bill used his influential power to get the promotion]
e. They buried the hatchet after years of fighting.
[They reconciled/ended/settled a disagreement after years of
fighting]
f. They shall beat their swords into plowshares.

[They shall forge offensive weapons into peaceful tools]
(Horn 2003: 246)
Referring to the examples in (1.1) and (1.2), Horn (2003: 245-247)

explains that fixed idioms are the ones whose NP objects cannot undergo
syntactic operations, such as passivation. This is illustrated by the
unacceptability of the sentences in (1.3), used in their idiomatic sense.
(1.3) a. *The bucket was kicked by Bill.1

b. *The bull was shot all evening.
c. *The coop was flown by bad guys.
(Horn 2003: 245)
Mobile idioms, on the other hand, can occur in the passive, as in (1.4)
below:
(1.4) a. The beans were spilled by Fred.

b. The cat was let out of the bag by Bill.
(Horn 2003: 245)
The mobility of VP idioms is correlated by Jackendoff (1997) with a

property called “metaphorical semantic composition.” He states that
idioms with this property “can be partitioned into chunks that correspond
to the “sub-idiomatic” readings of the syntactic idiom chunks”2
(Jackendoff 1997: 168). Fixed expressions, in turn, lack this property,
which in fact is proved finally by Jackendoff (1997) himself to be an
insufficient condition for mobility (cf. Horn 2003: 246).
Furthermore, Horn (2003) makes an attempt to replace Jackendoff’s
(1997) property of metaphorical semantic composition with a property of
“thematic composition,” and explains that “an expression has thematic
composition if the thematic structure of the verb in its literal sense and that
of the verb in its idiomatic sense are identical” (Horn 2003: 246). Having
defined the thematic structure as a set of semantic roles that a verb assigns
1
The sign * [asterisk] is used in the thesis to mark the sentence / phrase as
unacceptable in terms of grammar.
2
This appears to be similar to the property of composition proposed by Nunberg et
al. (1994), who explain that if an idiom is compositional, then elements of its
interpretation can be assigned to its various components “in such a way that each
constituent will be seen to refer metaphorically to an element of its interpretation”
[after the meaning of the idiom is known] (Nunberg et al. 1994: 496, 499).
20 Chapter One
to its NP arguments, Horn (2003) further argues that the property of

“thematic composition” is “a sufficient condition for mobility” (Horn
2003: 245). Consequently, Horn (ibid.) introduces a property of
transparency that distinguishes two types of mobile expressions, viz.
metaphors and mobile idioms, a division discussed neither by Jackendoff
(1997) nor by Nunberg et al. (1994). In short, the properties of thematic
composition and transparency interact to define three classes of VP
idioms: fixed idioms, mobile idioms and metaphors. Yet, fixed idioms and
mobile idioms, need to be encoded as phrasal idioms in lexical entries,
while metaphors do not have to be encoded in this way. Finally, all
expressions which have the property of thematic composition are mobile
to some extent, and all expressions that lack the property of thematic
composition display highly restricted mobility (Horn 2003: 270-271).
To sum up, nearly all studies treat metaphoricity as an essential
property of an idiom (e.g. Cronk et al. 1993; Gibbs 1980, 1985; and
McGlone et al. 1994; among others). From the traditional viewpoint,
idioms appear to be frozen and conventional phrases, quite different from
metaphors, which are often “novel and fresh” (Glucksberg 2001: 67).
Nevertheless, scientists argue that idioms “may not be that different after
all. (…) some types of idioms behave exactly like metaphors, while others
behave exactly like literal language” (Glucksberg 2001: 67). This
idiomatic diversity is responsible for various taxonomies of idioms,
analysed in the subsequent section.
1.3.1.2 Various taxonomies of idioms
Idioms have been classified in multiple ways by different researchers

based on their semantics, syntax, and function (Jackendoff 1997; Horn
2003; and Grant and Bauer 2004; among others). This section describes
some taxonomies: the one of Makkai (1972); these of Nunberg (1978) and
his followers (Gibbs and Nayak 1989: 104; Titone and Connine 1999); the
one of Cacciari and Glucksberg (1991), the one proposed by Sag et al.
(2002), and the classification of Yoshikawa (2008). These taxonomies
seem to be more complex (cf. Saberian 2011a: 1232), compared to the
classifications proposed by other linguists, i.e. Alexander (1987),
Fernando and Flavell (1981), Cowie et al. (1983), Nunberg, Sag, and
Wasow (1994), Fernando (1996), Howarth (1998), Moon (1998), among
others.
A. Makkai’s (1972) taxonomy
According to Makkai’s study (1972: 117), there are two idiomaticity areas in
English to which an idiom can belong: lexemic and sememic. The lexemic
idiomaticity area, (the class of the so-called “polylexonic lexemes”)
comprises expressions of more than one word, which are “subject to a
possible lack of understanding, despite familiarity with the meanings of the
components, or the erroneous decoding: they can potentially mislead the
uninformed listener, or they can disinform [sic] him” (Makkai 1972: 122).
Disinformation or misunderstanding take place when an idiom is decoded,
or understood in a semantically wrong way. Instead, the semantic
idiomaticity area (the class of the so-called “polysememic sememes3”)
contains expressions of more than one word, which have both a logical
literal meaning and a moral or a deeper meaning, e.g. proverbs. On the basis
of this theory of idiomaticity, Makkai (1972) classifies all idioms either
under the category of lexemic idioms or under the label of sememic idioms.
Lexemic idioms are shorter and function as parts of speech, whereas
sememic idioms function as sentence idioms.
Consequently, Makkai (1972) divides all lexemic idioms longer than one
word and shorter than a sentence into six types, presented in Table 1-1
below.
Types of lexemic idioms Example Meaning

(1) Phrasal verb idioms give in to admit to be defeated or not
to be able to do sth; to agree
to do sth against one’s will
(2) Tournure idioms to fly off the suddenly and completely lose
handle one’s temper
(3) Irreversible spick-and-span very clean and tidy
binomial idioms
(4) Phrasal compound the White House the president of the U.S. and
idioms his officials; the official
home of the president of the
U.S.
(5) Incorporating verb to baby-sit to look after or mind
idioms somebody’s children
(6) Pseudo-idioms kith and kin somebody’s relatives
Table 1-1. Types of lexemic idioms (Makkai 1972: 135-169)
3
A sememe is a semantic language unit of meaning, analogous to a morpheme,
relevant in structural semiotics (cf. Stanojeviü 2009).
22 Chapter One
As shown in Table 1-1, class (1) of phrasal verb idioms includes both
phrasal and prepositional verbs, with the constituent structure of verb +
particle. A phrasal verb (e.g. put up) can carry a separate literal meaning,
apart from its one or more idiomatic meanings (put up “accommodate” /
“give the idea”) (Makkai 1972: 135-136).
Furthermore, class (2) of Tournure idioms (often verbal idioms) are
made of at least three words and have a phrase-like structure. “Tournure”
means a mode of expression, so tournure idioms are the type of
expressions that people generally identify idioms with, e.g. kick the bucket
“to die” (Makkai 1972: 153-154). Some tournure idioms have a
compulsory it, which differentiates these tournure idioms from phrasal
verb idioms, (e.g. to have it out “to discuss a problem to solve it” (Makkai
1972: 148)). In addition, tournure idioms often contain a compulsory
definite or indefinite article, and they can only show variation in inflection
(past tense, future tense, etc.) (ibid.: 148).
Moreover, class (3) of Irreversible binomial idioms have a fixed
structure since their word order cannot be reversed, e.g. spick-and-span
“very clean and tidy” but not *span-and-spick (Makkai 1972: 164; Travis
1984).
Class (4) of Phrasal compound idioms comprises nominal compounds,
that first have to be institutionalised and widely recognised in their
specific meanings, but which denote a specific, commonly known object,
using common nouns, e.g. “White House.” Makkai even claims that within
this approach all proper nouns could count as idioms, even personal names
(Makkai 1972: 168).
Class (5) of Incorporating verb idioms consists of the first element that
is either a noun or an adjective, which is attached to a verb, e.g. to baby-
sit, to sight-see. For Makkai (1972), due to their ambiguous literal
interpretation, incorporating verb idioms may be idiomatic. For instance,
when to baby-sit is interpreted literally, it may mean “to make baby or
babies sit” or “sitting with regard to, or on account of a baby or babies”
(Makkai 1972: 168).
Finally, class (6) of Pseudo-idioms represents the last type in Makkai’s
(1972) classification, encompassing all idioms in the lexemic idiomaticity
area which have a cranberry morph as a constituent, e.g. hither and yon “in
many different directions or places; here and there,” kit and caboodle “the
whole lot; everything,” tit for tat “an action of revenge” (Makkai 1972:
340).
Additionally, Makkai (1972: 172-179) divides his sememic idioms into
nine types according to their origin or function in a language, as presented
in Table 1-2 below.
Types of sememic idioms Example Meaning

(1) First-base idioms Never to get to To fail to achieve the first
first base. state of significance in an
activity, rendering future
success unlikely.
(2) Idioms of
May I ask who’s
institutionalized Identify yourself!
calling.
politeness
(3) Idioms of It seems that…, I’m unable to find my
institutionalized I can’t seem to glasses
detachment or find my glasses. (but I refuse to give up).
indirectness
(4) Idioms of proposals How about a I’m offering you a drink.
encoded as questions drink?
(5) Idioms of institutional

How do you do! Greeting. Good day!
greetings
(6) Proverbial idioms with Curiosity killed One may pay dearly for
a moral the cat. one’s curiosity.
A little more than
(7) Familiar quotations as
kin,and less than
idioms
kind. (Hamlet
I.ii.65)
(8) Idiomaticity in It wasn’t too bad. Approval.
institutionalized It wasn’t exactly Displeasure.
understatement my cup of tea.
As cold as a
(9) Idiomaticity in
witch’s tit Extremely cold.
institutionalized
He won’t lift a He is very idle.
hyperbole
finger..
Table 1-2 Types of sememic idioms (Makkai 1972: 172-179)
In Makkai’s (1972) classification of sememic idioms, depicted in Table

1-2, class (1) of First-base idioms derive from a cultural background,
sayings or proverbs which relate to culturally specialized fields, e.g.
American baseball, as in never to get to first base “to fail to achieve the
first state of significance in an activity, rendering future success unlikely.”
Secondly, class (2) of Idioms of institutionalized politeness covers
imperatives in traditional, polite forms, treated by Makkai (1972) as
idioms, due to the fact that they rarely expect literal replies, e.g. Do you
mind if I … Not at all / No, I don’t. The third class, Idioms of
24 Chapter One
institutionalized detachment or indirectness, covers traditional forms of

speech which hint at detachment or indirectness, e.g. It seems to be
snowing ĺ “It is snowing (but I hate to say so).” Fourthly, Idioms of
proposals encoded as questions, which, if answered literally, indicate
misunderstanding, or deliberate refusal to co-operate, e.g. Why don’t you
sit over here? “come and sit down here” ĺ “because I don’t like that
chair!” The fifth category, Idioms of institutional greetings, comprises
items lexemically unchangeable, usually used for greetings, and no literal
answer is expected as a reaction to them. Another class, Proverbial
idioms, with a moral and a standard form, is commonly recognized and
cannot be much altered as regards person, tense or anaphors. The seventh
group, Familiar quotations as idioms, has to be institutionalized and
known well enough to be easily recognized. It often happens that the
person using these items invokes authority. Class eight, Idiomaticity in
institutionalised understatement, reduces the impact of a dull statement or
denotes approval of something. Finally, Idiomaticity in institutionalised
hyperbole, often considered as vulgar, implies mainly exaggerated terms
that have been widely accepted and become idiomatic. They are used both
in speech and writing to exaggerate what is expressed with the intention of
making something sound more impressive than it really is.
B. Nunberg’s (1978) and his followers’ taxonomies
Before Jackendoff’s (1997) and Horn’s (2003) subdivisions of idioms,

relying on their metaphorical semantic/thematic composition and
transparency, were proposed, there existed an initial taxonomy of Nunberg
(1978), which takes into account the meanings of parts of an idiom that
contribute to the figurative meaning of the whole. On the basis of
Nunberg’s taxonomy (1978), semantic taxonomies have been postulated to
describe how idioms differ in their compositionality, and how these
differences may influence the process model of idiom comprehension
(Nunberg 1978; Gibbs and Nayak 1989: 104; and Titone and Connine
1999). Both Nunberg (1978) and his followers create semantic taxonomies
classifying idioms into: normally decomposable idioms, abnormally
decomposable idioms, and non-decomposable idioms.
The first category of normally decomposable idioms, whose individual
components contribute to the figurative meaning, includes the so-called
one-to-one semantic relationship between the words constituting an idiom
and the components of the idiom meaning. In this type of idioms, a part of
the idiom is used literally, or there are clearly noticeable metaphorical
correlations between the words constituting the idiom and the elements of
that idiom figurative meaning. Gibbs and Nayak (1989) and Gibbs et al.
(1989a) exemplify this category with the idiom break the ice, when the
word break links to the idiomatic sense of changing a mood or tense
atmosphere, while the word ice relates figuratively to social tension.
Similarly, in the idiom pop the question, the noun question quite clearly
refers to a “marriage proposal,” while the verb pop to the act of expressing
it (cf. Gibbs et al. 1989b: 59; and CieĞlicka 2004: 95; among others).
The second category of idioms includes the so-called abnormally
decomposable idioms, whose individual components have some
metaphorical relation to their idiomatic referents (e.g. buck in the idiom
pass the buck, with its meaning “to attribute to another person or group
one’s own responsibility”) (Gibbs and Nayak 1989: 109). Likewise, as
argued by Gibbs et al. (1989a), “we can understand the hitting of certain
buttons in hit the panic button as a conventional metaphor for how we
react in extreme circumstances” (Gibbs et al. 1989a: 578). Consequently,
an abnormally decomposable idiom may be viewed as somewhat lexically
flexible (cf. Gibbs et al. 1989b: 65). Additionally, the difference between
normally and abnormally decomposable idioms lies in the fact that,
compared to the former category, in which the words constituting an idiom
denote directly some component of the idiomatic reference, the latter
contains such idioms which only refer to some metaphorical relation
between the individual part and the referent (cf. CieĞlicka 2004: 95).
Thirdly, semantically non-decomposable idioms are the ones whose
individual constituents do not contribute to the overall figurative meaning,
e.g. chew the fat, which means “to gossip or make a small talk” (Titone
and Connine 1999). This category resembles the traditional approach to
idiomatic expressions, which are not compositionally derived from their
constituent words that build the string. Thus, Gibbs et al. (1989a) provide
a definition of semantically non-decomposable idioms in the following
words, “the individual components of phrases such as kick the bucket or
chew the fat are not in the same semantic field as their respective
figurative referents (i.e. “to die” and “to talk without purpose”) and should
not be viewed as semantically decomposable” (Gibbs et al. 1989a: 578;
and cf. CieĞlicka 2004: 95).
Having provided the division of idioms founded on Nunberg’s (1978)
initial taxonomy, Gibbs et al. (1989b: 59) mention the syntactic
variability of idioms. On the basis of the results of their experiments, they
prove that the more decomposable an idiom is, the more syntactically
productive the idiom is expected to be. Therefore, normally decomposable
idioms (e.g. pop the question) have been found much more syntactically
productive than either abnormally decomposable (e.g. carry a torch,
26 Chapter One
which means “to love or to be romantically infatuated with”) or

semantically non-decomposable idioms (e.g. chew the fat “to talk without
purpose”). These conclusions largely support the predictions of the idiom
decomposition hypothesis, making a suggestion that the syntactic
behaviour of idioms can be analysed through examination of their internal
semantics.
To wrap up, in their subsequent studies of idioms, Nunberg, Sag, and
Wasow (1994), followed by Espinal and Jaume (2010), and Harwood et
al. (2016), among others, propose a bipolar classification of idioms,
dividing them into idiomatically combining expressions (ICE) and
idiomatic phrases (IP). While the typical example of idiomatically
combining expressions is pull strings (“to use connections”), in which the
overall idiomatic interpretation is distributed among its parts, even though
these are associated with conventional meanings (e.g. pull ĺ use, and
strings ĺ connections); the example of idiomatic phrases, commonly
cited in the literature, is kick the bucket “to die,” the meaning of which is
completely not derivable from its components (both “kicking” and “the
bucket” do not contribute to the overall meaning of the idiom).
C. Cacciari and Glucksberg’s (1991) and Glucksberg’s (1993)

taxonomies
In their taxonomy of idiom compositionality, Cacciari and Glucksberg

(1991), and Glucksberg (1993) categorize idioms as opaque /
compositional-opaque, transparent / compositional-transparent, quasi-
metaphorical, and non-decomposable / non-compositional.
First of all, opaque idioms are phrases, such as kick the bucket, in
which there is some degree of semantic constraint on interpretation of the
idiom. Here the meanings of individual elements can still constrain the
way in which the idiom is interpreted and used. Hence, even if the idiom
kick the bucket is opaque, in that its constituent words do not map onto its
figurative reading, the semantics of the verb “to kick” can still constrain
the interpretation of the idiom. Since kicking is a distinct act and involves
a swift action, saying that “he was kicking the bucket all week” is
inappropriate, even if saying that “he lay dying all week” is perfectly
suitable (cf. CieĞlicka 2004: 97).
On the contrary, transparent idioms are phrases with a direct mapping
of literal constituent meanings to idiomatic meanings. For example, spill
in the idiom spill the beans, which is literally translated as “to divulge a
secret,” straightforwardly draws our attention to the verb “divulge,” and
the word the beans directly implies “a secret” (cf. Saberian 2011a: 1232).
Furthermore, idioms classified as quasi-metaphorical phrases are those

in which the overall literal meaning of the phrases metaphorically maps
onto the idiomatic meaning. To provide Glucksberg’s (1993) example,
“giving up the ship is simultaneously an ideal or prototypical exemplar of
the act of surrendering and a phrase that can refer to any instance of
complete surrender” (Glucksberg 1993: 18). Likewise, carry coals to
Newcastle denotes an idea of bringing something to a place which has a
wealth of that thing, while bury the hatchet indicates any example of
making peace, not essentially including the once ritual action of burying
physically the hatchet. Such quasi-metaphorical idioms, reveal their
meaning via allusion (Glucksberg 1993, 2001), which means that at the
same time they denote an ideal paradigm of a concept and the contextually
determined referent in a specific communicative situation (cf. CieĞlicka
2004: 98; and Saberian 2011a: 1232).
Finally, the last category of non-decomposable / non-compositional
idioms embraces expressions where there is no relation between the idiom
constituent parts and the overall figurative reading, as in the phrase by and
large “in all possible circumstances; generally speaking,” or trip the light
fantastic “to dance or move to musical accompaniment.” Such non-
compositional idioms are opaque due to the fact that, in contradistinction
to transparent idioms, the idiom literal meaning does not show even the
slightest reference to its figurative interpretation (cf. CieĞlicka 2004: 98;
and Papagno and Romero Lauro 2010: 22).
Similarly, depending on the degrees of figurativeness, Mäntylä (2004:
28-29) mentions another, although parallel, way of categorising idioms.
They are usually divided into three categories, somehow overlapping with
one another, without strict border lines, and they mark how easily the roots
of figurativeness are to be detected. These are: transparent idioms with
their literal meaning clearly linked to the figurative meaning, e.g. give the
green light; semi-transparent idioms, i.e. the expressions where the literal
meaning gives some hint as to the figurative meaning but the link is not as
noticeable as with fully transparent idioms, e.g. quake in your shoes,
which means “to feel nervous or afraid.” Finally, there are opaque idioms,
where the motivation behind the figurative meaning is impossible to
perceive without knowing the etymology, e.g. be home and dry, which is
translated as “succeeding at something and not expecting any further
problems” (cf. Colin 2005, and Peacock 2009: 2, who mention also the
fourth group, viz. semi-opaque idioms).
28 Chapter One
D. Sag et al.’s (2002) taxonomy
Sag, Baldwin, Bond, Copestake, and Flickinger (2002) in their article

“Multiword expressions: A pain in the neck for NLP” included a long
passage on multiword expressions (MWEs), defining them as phrases
which comprise at least two words, can be syntactically and/or
semantically idiosyncratic in nature, and which act as a single unit at some
level of linguistic analysis (Sag et al. 2002: 1). Besides, MWEs can be
treated as lying at the interface of grammar and lexicon, usually being
instances of well productive syntactic patterns, showing, on the other
hand, a peculiar lexical behaviour (Calzolari et al. 2002: 1934).
Furthermore, Sag et al. (2002: 3-8) classify MWEs into lexicalized
phrases, which have at least partially idiosyncratic syntax or pragmatics,
and institutionalized phrases, which are syntactically and semantically
compositional. The former group of phrases, i.e. lexicalized phrases, can
be further divided into (1) fixed expressions, (2) semi-fixed expressions,
and (3) syntactically flexible expressions. While fixed expressions are fully
lexicalized, viz. totally fixed, and can neither vary morpho-syntactically
nor be modified internally (e.g. in short, by and large, every which way);
semi-fixed expressions have a strictly invariable word order and
composition, but may be modified as regards inflection, variation in
reflexive form and determiner selection.
Interestingly, Sag et al. (2002: 4) make further exemplification of
semi-fixed expressions when they mention (a) “non-decomposable idioms”
(i.e. idioms in which the meaning cannot be assigned to the components of
the expressions), such as kick the bucket “to die,” in which the verb can be
inflected only in a specific context: he kicked the bucket, or varied in the
reflexive form: wet oneself. However, non-decomposable idioms normally
do not show syntactic variability, i.e. a passive form: *the bucket was
kicked, or internal modification: *kick the red bucket in the sky, are not
possible with the same idiomatic meaning (Sag et al. 2002: 5). Another
type of semi-fixed expressions are (b) “compound nominals,” syntactically
unalterable but inflected for number, such as: car park ĺ [car park]s.
However, for left-headed compounds such as attorney general,
congressman at large and part of speech, the inflection employed would
result in anomalies, e.g. *[congressman at large]s (Sag et al. 2002: 5).
Finally, semi-fixed expressions include also (c) “proper names,” which are
syntactically highly idiosyncratic. U.S. sports team names, for instance,
are canonically made up of a place or organization name (probably a
MWE in itself, such as San Francisco) and an appellation that assigns the
team uniquely within the sport (such as 49ers). The name of the U.S.
sports team can undergo optional elision, e.g. the San Francisco 49ers can
occur as the 49ers, or as a modifier in the compound noun a 49ers player,
etc.
The last subclass of lexicalized phrases, within the taxonomy of Sag et
al. (2002), apart from fixed expressions, and semi-fixed expressions,
comprises syntactically-flexible expressions, which have a wider range of
syntactic variability than semi-fixed expressions. Syntactically-flexible
expressions occur in the form of (i) decomposable idioms; (ii) verb-
particle constructions; and (iii) light verbs. “Decomposable idioms” can be
syntactically flexible to some extent, but it is difficult to predict what kind
of syntactic variation a given idiom can undergo. Moreover, “verb-particle
constructions,” such as write up and look up are made up of a verb and one
or more particles. They may be either semantically idiosyncratic, as brush
up on “to improve,” or compositional as break up in the meteorite broke
up in the earth's atmosphere. In some transitive verb-particle constructions,
as call someone up, an NP argument can occur either between or following
the verb and particle(s): call Tom up or call up Tom, respectively. Besides,
adverbs can often be inserted between the verb and particle as in fight
bravely on. Finally, in the case of “light verb constructions,” as make a
mistake, or give a demo, it is hardly predictable which light verb is
connected with a given noun. Although they are highly idiosyncratic, they
have to be distinguished from idioms: “the noun is used in a normal sense,
and the verb meaning appears to be bleached, rather than idiomatic” (Sag
et al. 2002: 7).
Finally, the taxonomy of Sag et al. (2002: 7), beside lexicalized
phrases, also includes institutionalized phrases which cover
conventionalized phrases / collocations, such as salt and pepper, traffic
light and to kindle excitement. They are semantically and syntactically
compositional. Regarding the phrase traffic light, traffic and light both
retain simple senses but produce a compositional reading by being
combined into constructions. As institutionalized phrases are fully
compositional, they can show full syntactic variability.
E. Yoshikawa’s (2008) taxonomy
Yoshikawa (2008) groups idioms into five different types: A, B, C1, C2

and D, with the last idiom type added by Saberian 2011b. In this
taxonomy, the main criterion of classifying English idioms is the degree of
L1-L2 structural and semantic similarity. If the major L2 constituent
words could be literally translated into L1, and if the L2 idiom is
semantically similar to an L1 idiom, since it shares the same central
30 Chapter One
concept used in the same contexts (pragmatically congruent), then an L2

idiom is structurally similar to an L1 idiom (Cedar 2004). Type A idioms
include English idioms with both structural and semantic similarity to L1
idioms. Type B contains idioms with some structural similarity and
semantic parallelism to L1 idioms; whereas Type C1 covers idioms with
structural “resemblance” but semantic “dissimilarity” from L1 idioms.
Type C2 includes idioms which both structurally and semantically differ
from L1 idioms; while Type D contains idioms with structural
“dissimilarity” but semantic similarity to L1 idioms. Nevertheless, Type D
idioms cannot be translated literally into L1, because their literal
translation is not logical in L1, yet their literal translation may give
language users some clue to predict the idiomatic denotation (cf. Saberian
2011a: 1232).
In brief, this section has focused on some taxonomies of idioms,
especially those widely used or referred to in the literature. The common
denominator of all taxonomic subclasses is the degree of an idiom literal
or / and figurative meanings, which in most cases overlap with one
another. A literal meaning of the phrase metaphorically maps onto the
idiomatic meaning with different intensity moving up and down the
idiomatic scale, making ground for different types of idioms.
1.3.1.3 Idiom processing and metaphorical interpretation
In the discussion concerning metaphoricity of idioms, Abel (2003: 347)

assumes that in the course of processing of some idioms conceptual
metaphors, in the sense of Lakoff and Johnson (1980),4 are activated, such
as “anger is heated fluid in a container” or “anger is fire,” which motivate
the meaning of idioms like smoke was coming out of his ears, she was
spitting fire, he was fuming, etc. Some idioms, e.g. miss the boat or pass
the hat, where one component has a metaphorical interpretation, probably
activate conceptual information with regard to this interpretation. Other
idioms are not conceptually motivated at all, e.g. kick the bucket; thus, this
fact proves that conceptual or metaphorical motivation cannot be
automatically equated with predictability of idiomatic meaning (cf. Gibbs
1992, 1995; Gibbs and O’Brien 1990; Nayak and Gibbs 1990; Glucksberg
et al. 1993; and CieĞlicka 2004; among others).
4
Lakoff and Johnson’s (1980) research has evoked many studies referring to
conceptual metaphors, such as, e.g. ARGUMENT, TIME and LOVE, which are
said to be used by people continually, and thus stimulating more extensive
linguistic studies.
Remarkably, CieĞlicka (2004) mentions another example within the

conceptual metaphor framework “LOVE IS JOURNEY,”5 which entails
comprehending one abstract domain of experience (love) in terms of
another, more concrete domain of experience (that of journey) (CieĞlicka
2004: 63). She adds (ibid.: 64) that there is thus a tight mapping between
entities in both domains, since the entities in the domain of love (e.g. the
lovers, their relationship, their common goals, and life difficulties)
correspond to their parallel entities in the domain of journeys (e.g. the
travellers, their vehicles, destinations). This ease and naturalness in
comprehending conceptual metaphors without an effort, and even without
conscious reflection is the result of “pre-existing conceptual metaphorical
mappings between conceptual domains that structure our experience and
perception” (CieĞlicka 2004: 63).
On the other hand, as referred by Mäntylä (2004: 29), since
figurativeness depends on the judgement of the individual language user, it
may sometimes be very difficult to define the degree of metaphoricity of a
single idiom, since knowing its meaning, or other similar expressions in
either the foreign language or the native one, including their context, may
influence the assessment. Even though it is easier to see the link to the
literal meaning once the figurative meaning is known, Laufer (1997)
warns about “deceptive transparency” that is, words that “look as if they
provided clues to their meaning” (Laufer 1997: 25). In fact they do not,
e.g. false friends, such as put words into somebody’s mouth, which means
“to suggest that someone has said something when in fact they have not.”
Polish learners, instead, would understand it as “to help someone say
5
As noted by Anudo and Kodak (2017: 168), Lakoff and Johnson’s (1980)
metaphors allow us to comprehend a more clearly delineated concept, e.g. the
concept of love can be comprehended in terms of a “journey.” Moreover,
metaphors should involve two different kinds of activities, as in the “love is
journey” metaphor, love involves emotions, but a journey refers to travelling.
Additionally, metaphors structure our everyday concepts, since “love” is partially
structured in terms of a “journey.” Furthermore, metaphors enable us to understand
one domain of experience in terms of another, e.g. the target domain of love is
referred to by the source domain of a journey. Besides, in the “Love is a journey”
metaphor, different means of travel can be used to explain a love relationship, i.e.
the journey could be undertaken using a train, a car, or a ship, etc. Since these
means of transport are different kinds of vehicles; thus, “vehicles” have become a
superordinate category evoking in human minds rich mental images and rich
knowledge structures. A mapping at the superordinate level enlarges the chances
for mapping rich conceptual structures in the source domain onto the target domain
(cf. Anudo and Kodak 2017: 168; Kövecses 2000, 2005, 2010; and Gavelin 2015;
among others).
32 Chapter One
something that is expected or needed”). Fortunately, despite one’s

individual or national creativity and historical uniqueness, there exist some
similarities among languages with the same or similar pictures evoked by
idioms. They derive from biblical, mythological and everyday life scenes,
and constitute a common language store that builds close connections
between language users, regardless of their origin, culture and nationality
(cf. Kellerman 1999; and Sornig 1988: 281).
Other factors that make the recognition and comprehension of idioms
more difficult relate to the fact that some expressions bring their literal
reading faster than their figurative connotations, depending on the context
(Cacciari 1993: 27; Marschark et al. 1983; Moore 1982; Popiel and
McRae 1988; and Needham 1992; among others). Moreover, both
distinguishing unfamiliar idioms and reading well-known idioms with
their figurative meaning create problems. In other words, if a language
user depends too much on metaphoricity, s/he may attach meanings or
features to an idiom that are not present there (cf. Mäntylä 2004: 29-30).
Pulman’s (1993: 250) example of cat among the pigeons suggests a
possibility when a language user might expect a connotation of cruelty,
while constructing the image the expression denotes, but instead, the
phrase has a different meaning (a disturbance caused by an undesirable
person from the perspective of a group).
Consequently, as CieĞlicka (2004) notices, it is essential for each
language user to acquire the ability to “deal with figurative language,”
called figurative competence (cf. Levorato 1993: 104; CieĞlicka 2004: 19).
The ability implies not only such language skills as understanding
figurative usages of a word, or the relationship existing between the literal
and non-literal meanings, but also the ability to produce figurative
language by creating new figures of speech.
However, the most recent psycholinguistic models seem to point out
that some amount of literal activation is required in the course of idiom
processing. While the activation of literal meanings of idiom components
appears well-documented in the idiom comprehension literature, the
lexical access of idiom components in idiom production has not been
made familiar enough, yet. On the other hand, linguists focused on the
processing of idioms and their storage in the mind (e.g. Fraser 1970; and
Gibbs 1980; among others), and have not been inclined to deal with the
problem of defining an idiom, but have taken its definition more or less for
granted.
All in all, the very nature of idioms, as well as the literal and figurative
character of idioms, evoked significantly in idiom processing, are closely
related to the next feature of idioms, viz. idiom ambiguity, resulting from
idiom metaphoricity, which is to be analysed below.
1.3.1.4 Idiom ambiguity
Chafe (1968) notices that many idiomatic expressions are ambiguous, with
one interpretation (the literal meaning), deriving from the meanings of the
words involved, and the other–the idiomatic meaning.
To begin the discussion concerning idiom ambiguity, it is worth
recalling transformational grammarians’ classification of ambiguity first.
They distinguish (i) lexical; (ii) surface structure; and (iii) underlying
structure ambiguities (cf. Chomsky 1965; MacKay and Bever 1967; Bever
et al. 1969: 225; and Bobrow and Bell 1973; among others). The lexical
ambiguity implicates alternative dictionary meanings of a word with no
differences at the other grammatical levels (e.g. in “The cold was
bothering John,” cold may refer either to John’s illness or the weather
condition which is worrying John). The surface (or derived) structure
ambiguity depends on how words are grouped together into phrases, i.e.
structured (e.g. “Visiting relatives can be nice” may be understood as
follows: the activity of visiting (relatives) is nice, or the relatives who are
visiting us are nice). The underlying structure ambiguity entails a change
in the essential relations between words (e.g. in “The mayor told the police
to stop drinking,” drinking was stopped either by the mayor who
announced it to the police, or the police was ordered to stop drinking) (cf.
Bobrow and Bell 1973: 343).
In addition, the underlying structural level of sentences represents the
essential logical relations which the words bear to each other. In other
words, at a deeper level the logical relational concepts, subject, predicate,
and object, express the internal relations among the words and phrases of a
sentence (cf. Bever et al. 1969: 225). Some explanation and illustration of
ambiguity between alternative syntactic structures underlying a sentence is
provided by Wundt (1900: 268) and Bever et al. (1969: 225-226). The
actual order of the words in a sentence does not always correspond to the
underlying relations. For example, in (1.5) active and (1.6) passive
sentences, presented below, the underlying relations are the same although
the word orders differ:
(1.5) Caesar crossed the Rubicon.
(1.6) The Rubicon was crossed by Caesar.

34 Chapterr One
In the two sentences, inn (1.5) and (1 1.6), the actinng person (su
ubject) is
Caesar in both cases. Buut this phrase is the topic off the statemen nt only in
the first, butt not in the seecond sentencee (cf. Wundt 1900: 268). These
T two
examples arre analysed as a having thee same underrlying phrase structure
represented in the tree diagrams in n Figure 1-11 and in Fig gure 1-2,
respectivelyy.
Figure 1-1. Sample sentennces with superfficial phrase-strructure trees (B

Bever et al.
1969: 225)
In Figure 1-1 each brranching “node” of the phrase structure tree

correspondss to a phrase structure “co onstituent,” w
while each co onstituent
correspondss to the relatioon between thee words and pphrases that it contains.
For examplee, in the first sentence the structure reprresents the facct that the
words “Caeesar” and “Ruubicon” are morem closely rrelated than thhe words
“crossed” aand “the.” Heence, the leveel of “underlyying phrase structure”
s
represents thhe “logical” relations
r whicch the words pperform to eaach other.
The two senntences, in (1.5) and (1.6), illustrated in FFigure 1-1, arre further
shown in thhe tree diagram m in Figure 1-2
1 below, whhich depicts th he logical
structure undderlying the sentences in Figure
F 1-1.
T
Towards Definiing an Idiom 35
Figure 1-2. L he sentences in Figure 1-1 (B

Logical structurre underlying th Bever et al.
1969: 226)
Furthermmore, Benjafiield (1992: 255) 2 in his interpretation n of the

classical prooblem of struuctural ambigu uity raises thee following questions,
q
“Why does tthe same surfa face form havee systematicallly different reeadings?”
“What causses such diffeerences?” Then he explainns, “[m]eanin ng is not
given on thhe surface of a sentence, but b is given bby the deep structures
s
interpretatioon of sentence. When we un nderstand a sen
entence, we traansform a
surface struccture into a deeep structure. When we prooduce a senten nce we go
the other wway: from a deep structure to a surface structure” (B Benjafield
1992: 255-2256).
In addiition, some insight hass been alsoo provided into the
‘psychologiccal reality’ ofo the structu ures postulateed in transforrmational
grammars, llike the ones presented above. The mo st notable succcess has
been to shoow that the form in which sentencess are understtood and
memorized corresponds closely to thee logical struccture underly ying them
(cf. Mehler aand Bever 19667).
Nevertheeless, idiomaatic ambiguity y is hardly explicable within
w the
transformatiional grammaar rules of underlying
u sttructures. Instead, the
idiomatic mmeaning seemss to be underrstood by com mbining severral words
into a compplex “idiom word”
w and find
ding the meanning of the ph hrase by a
search throough mental “idiom worrd” dictionary ry. Discovery y of the
idiomatic mmeaning of an idiom seems to result from m processing thet idiom
as a word. S Such a processsing strategyy differs from the one that has been
suggested ffor literal exppressions, wh herein each w word is perceeived, its
meanings diiscovered, andd then mappeed onto a sem mantic whole such that
the meaninggs of the indivvidual words relate to prodduce a meanin ng for the
sentence (cff. Bobrow 19773).
Furthermmore, Moon (11998: 178) reffers to idiom aambiguity witth a close
connection tto metaphoriccity, claiming that context can solve thee problem
36 Chapter One
of ambiguity that is evoked usually by idiom polysemy. But, due to

several potential metaphorical interpretations, even with the help of the
context, a language user may encounter obstacles to distinguish between
the literal and figurative level of idioms. Besides, Moon (1998: 185)
notices that idioms might be potentially ambiguous both in isolation, and
in context if they are unfamiliar to the language user. Then, the context
itself may produce false interpretations. Yet, if an idiom is well-known,
ambiguity occurs occasionally, even though some idioms have several
figurative meanings, e.g. look someone in the eye, sit on the fence, blow
the whistle on someone, to name a few.
Mäntylä (2004: 30), on the other hand, pays attention to different
degrees of people’s perception, even in the case when an idiom is familiar.
Consequently, for some language users, an idiom sounds more arbitrary
than for others; and there are totally arbitrary, and thus incomprehensible
idioms in a language, as well (Lakoff 1987: 451). What is more, Moon
(1998: 179) gives an example of some phrases that have a highly
improbable literal meaning, e.g. it is possible to literally bite the bullet, or
to have a bee in one’s bonnet, but this is rather not likely to happen in
today’s world. So, the fact that some literal interpretation is possible, does
not imply its real occurrence. Moon’s (1998) research results prove that
literal interpretations are hardly ever compared to figurative ones (Moon
1998: 180-181).
All in all, every creative language user may take advantage of the
figurative-literal relations surrounding idioms. Naturally, it can be noted
that meanings are never the same for different people. Taking into account
the fact that metaphoricity may help in predicting the meaning of an
unfamiliar idiom; nonetheless, a language user, mainly a non-native one,
may feel equally left in the dark when trying to bring to light the very
nature of an idiom, its figurativeness, literalness, or both characteristics.
This point adds to the complexity of idioms, since they are expressions
with, presumably no arbitrary meanings, but they behave like single,
arbitrary words. Finally, once deep-rooted idiom characteristics are
acknowledged and their figurativeness admitted, these expressions become
far more comprehensible.
1.3.2 Analysability vs. non-compositionality

Non-compositionality of idioms, understood as the feature in which “the
meaning of an idiom is not predictable just from the meanings of the
individual words that make it up” (Lakoff 1987: 448), has been questioned
first by linguists (Nunberg 1978; Wasow, Sag, and Nunherg 1983) and
then by psychologists (Cacciari and Glucksberg 1991; Gibbs et al. 1989b).

The main reason for the inadequacy of Lakoff’s (1987) definition of idiom
non-compositionality is its variability and possibility to be modified in
different ways, not only in terms of the entire idiom, but also in the case of
local modifications. Both Wasow et al’.s (1983) and Nunberg’s (1979)
statements about partial analysability of idioms have demonstrated that the
components of an idiom have distinguishable meanings that do contribute
to the total interpretation of the idiomatic expression. Thus, it is possible
to recognize a synchronic relation between the figurative and literal
meanings of idioms. The extent to which idiom constituent parts
contribute to the idiom overall interpretation has been termed semantic
decomposition (cf. Nunberg 1978) or their analysability (cf. Cacciari and
Tabossi 1988; Glucksberg 1991; and Gibbs 1994; among others).
Furthermore, Stock et al. (1993) notice that “there is a whole class of
idioms for which Non-compositionality is false,” and they observe that
with less metaphorical idioms, often “the apparent dissociation between
the literal and idiomatic meaning is simply due to the fact that the
connection is buried in the history of the language and the culture” (Stock
et al. 1993: 231). Gibbs et al (1989a: 578) comment that various studies
have shown that certain idioms are more decomposable, or semantically
analysable, than others. Thus, taking into account Nunberg’s (1978)
typology of idiom compositionality, Gibbs and Nayak (1989), and Gibbs
et al. (1989a) offered three types of compositionality – normally
decomposable idioms, abnormally decomposable idioms and non-
decomposable idioms – depending on the intensity with which individual
word meanings contribute to their idiomatic readings. Nunberg’s
taxonomy (1978), was proceded by Makkai’s (1972) taxonomy and
followed by Cacciari and Glucksberg’s (1991, 1993) divisions, then Sag et
al.’s classification (2002), and finally Yoshikawa’s taxonomy (2008). All
these typologies have been presented in section 1.3.1.2, as they seem to be
more detailed than those offered by different researchers on the basis of
idiom semantics, syntax, and functions (Jackendoff 1997; Horn 2003; and
Grant and Bauer 2004; among others).
Pulman (1993) names idioms analysable, provided they can be split
into such meaningful units that each of them corresponds to a part of the
metaphorical meaning of the whole. Analysability or decompositionality is
thus connected to figurativeness, and, as referred to by Gibbs (1993) and
Pulman (1993), also to variation. As argued by Pulman (1993), fully
opaque idioms hardly ever display variance, while figurative and
analysable ones do show it (cf. Stock et al. 1993: 234). Nevertheless, there
are idioms that are recognized as opaque now, but originally they used to
38 Chapter One
convey concrete meanings, analysable to people of that time. Consequently,

present-day idioms and their at-the-first-glance opacity may prove to be
completely apparent, and their intolerance of variance is, in fact, not
confirmed (Keysar and Bly 1999: 1575-1576).
Besides, Stock et al. (1993) state that a number of idioms is analysable
in such a way that each component can be understood, regardless of the
fact whether it has a particular metaphorical connotation or not. They call
this relationship between the words and the idiomatic meaning the
components carry in that particular expression, referent mapping (Stock et
al. 1993: 235). Moon (1992: 15) argues that referent mapping entails
“revitalising and foregrounding compositional meaning,” and she provides
further evidence for this statement by means of the following example: “A
television news reporter asking President Bush ‘Did this summit bury the
hatchet?’ Bush: ‘There is no hatchet’” (Moon 1992: 15-16).
In addition, as noticed by Gläser (1988), some types of fixed phrases
can be reduced and referred to the whole expression through one
constituent, e.g. it is possible to refer to a rolling stone without repeating
the whole expression a rolling stone gathers no moss (which has two
meanings: “people pay a price for being always on the move, in that they
have no roots in a specific place” (the original meaning); or “people who
keep moving avoid picking up responsibilities and cares”) (Gläser 1988:
274). The possibility to shorten some idiomatic phraseological units,
without missing their total figurative meaning, indicates that in some cases
idiom frozenness can be broken. Moreover, Moon (1998) adds that lexical
variation of idioms, applied to fine-tune them to the context and situation,
is taken as the “evidence of their compositionality” (Moon 1998: 170).
Compositionality is, in some way, related to the figurative usage of single
elements in general (Moon 1998: 201), e.g. in light a fire under someone,
“fire” is a component used both in a literal and a metaphorical sense.
Moon (1998: 201) calls such expressions “incorporated metaphors.”
What is more, CieĞlicka (2004: 99) emphasizes the role of internal
semantics in interpreting figurative meanings of idioms, which has been
confirmed while dealing with poorly known or unknown idiomatic
expressions. Flores d’Arcais’s (1993) experiment and its results reveal that
many people construct their paraphrases of unknown phrases, basing on
the semantic properties of idiom constituents, or on the literal meanings of
idiom words. In short, while processing less familiar idiomatic phrases,
language users apply a crucial technique of analysing semantic properties
of idiom components to obtain the overall figurative interpretation.
Therefore, CieĞlicka (2004: 99) draws a conclusion that, even in non-
compositional idiomatic phrases, both the semantics of idiom components
and the figurative notion they designate determine the comprehension and
use of the idiom. Thus, paying attention to semantic productivity of idioms
would be helpful to investigate the role that idiom constituents and their
meanings play in the idiom overall figurative interpretation. Glucksberg
(1993) describes semantic productivity as “the ability of people to create
new idiomatic meanings by changing relevant aspects of an idiom's
constituent elements” (Glucksberg 1993: 15). According to Glucksberg
(1993), semantically productive processes are motivated by communicative
intentions, and thus they enhance communicative functions, in
contradistinction to unmotivated synonym substitutions. Hence, semantic
productivity entails an interpretable relationship between original idiom
components and their substitutes, so that the speaker’s communicative
intention may be deduced (cf. CieĞlicka 2004: 100). To provide
Glucksberg’s (1991) example, replacing shatter the ice for break the ice
generates a new idiomatic sense that is based on the meaning of the
original idiom and the relation between the original constituents and their
substitutes. The difference between the meaning of break and its
substituted element shatter produces a new interpretation, defined by
Glucksberg (1991) as “to break down an uncomfortable and stiff social
situation flamboyantly in one fell swoop” (Glucksberg 1991: 149). Such
examples of semantically productive idiom variants, as observed by
Glucksberg (ibid.), can be encountered regularly in the media, literature
and everyday conversation; henceforth, being comprehended by language
users without difficulty.
Finally, Gibbs (1993) strongly argues that analysability, strictly related
to metaphoricity, is crucial in comprehending and learning idioms.
Besides, Gibbs (1994) sees semantic productivity as linked to the degree
of idiom analysability; which is quite interesting, due to the assumption
that the more analysable a given idiom is, the more probable it is for it to
undertake various semantic modifications. Nevertheless, Glucksberg
(1991) does not hold this belief, emphasising that analysability is “neither
a necessary nor a sufficient condition for an idiom to be varied
productively” (Glucksberg 1991: 151). Instead, what matters for
Glucksberg, (1991) is whether the semantics of the idiomatic phrase has a
direct functional relation to the idiom meaning. Idiom semantic
productivity is closely connected with idiom lexical flexibility, which is
discussed in the following section.
40 Chapter One
1.3.3 Fixedness of form and internal structure

Jackendoff (1997) recognizes idioms as linguistic units with an internal
linguistic structure constrained by syntax, semantics, morphology, and
phonology. This internal structure of idioms determines idiomatic
behaviour, and sheds light on idiom variability. Ifill (2002: 6) notices that
idioms as fixed phrasal expressions are not completely frozen forms.
Having examined how fixed idioms are, and in what ways, he refutes the
notion that idioms are atomic units which lack an internal structure.
Moreover, idiom variability is related to the lexical flexibility of idioms.
Even though in some idioms substituting their individual words with other
lexical items is permissible, without disrupting their overall figurative
meanings; other idioms appear to be so frozen lexically that changing any
of their individual components brings about losing their overall figurative
interpretation. Gibbs and Nayak (1989) exemplify this principle by means
of the idiomatic expression eat one’s words meaning “to swallow,” in
which replacing the word eat with swallow will result in a comprehendible
idiomatic phrase swallow one’s words. In contrast, altering the word
bucket into pail in the idiom kick the bucket will reduce the phrase to its
literal meaning, rather than the idiomatic one.
Furthermore, usually, idioms that do not undergo many alterations are
considered non-flexible or frozen, while idioms that admit most operations
are considered flexible. Fraser (1970: 22-42) proposes a six-point
Frozenness Hierarchy for idioms stretching from totally frozen forms that
permit no grammatical or lexical changes to idioms that tolerate
unrestricted variation. Fraser (1970) names the points in the Hierarchy as
depicted in (1.7), providing the representative examples to them (cf. Fraser
1970: 40-41; and Runosalo 2005: 15-16; among others):
(1.7) Fraser’s (1970) Frozenness Hierarchy:
L6 – Unrestricted: all transformation possible, i.e. no idiom can belong

to this level, due to the fact that a string of words that allows all the
possible transformations can only be a literal word cluster, e.g.
open a window
L5 – Reconstitution: nominalization of the verb phrase of an idiom
(thus it can function as a subject of the sentence), e.g. let the cat out
of the bag “to reveal a secret”ĺ Her letting the cat out of the bag...
L4 – Extraction: the particle movement rule (the particle is extracted
from the idiom), e.g. look up something “to admire” ĺ look
something up
– pre-posing prepositional phrases (preposition of an idiom is

extracted from the verb), e.g. depend on “to count on” ĺ on whom
we can depend
– passive transformation,6 where the extraction concerns the direct
object noun phrase. When passivized, it is placed outside the idiom,
e.g. hit the nail on the head “to be exactly right in one’s opinion”
ĺ The nail was hit on the head.
L3 – Permutation: particle movement / idiomatic indirect object
movement within the idiom itself, e.g. bring the house down “loud
claps and cheers of the audience to praise a good performance” ĺ
bring down the house
L2 – Insertion: indirect object movement, e.g. lend a hand “to help
someone”ĺ lend “Mary” a hand, “to help someone” ĺ lend a
“helping” hand
L1 – Adjunction: gerundive nominalization (-ing–form and the of–
genitive), e.g. burn the candle at both ends “to try to do too many
things in too short period of time” ĺ burning the candle at both
ends
L0 – Completely Frozen: no transformation possible, idioms cannot be
interpreted literally, e.g. trip the light fantastic “to dance”
Fraser (1970) has developed the Frozenness Hierarchy in accordance

with the transformational behaviour of idioms. That is why, idioms with
syntactic restrictions allow few transformations, whereas some idioms
allow a larger number of transformations, and others may be fully
restricted. L0 indicates that no operations whatsoever may affect an idiom.
Literally uninterpretable idioms, such as trip the light fantastic “to dance
or move to musical accompaniment,” belong to level L0. However, there
are no idioms which can be categorised under the uppermost level L6,
because this level assumes some operations such as topicalization, which,
according to Fraser (1970), cannot affect idioms. Yet, the most frozen
idioms, belonging to L0, permit no distortion, while the least frozen, L5,
allow a considerable variation.
Furthermore, Fraser (1970) makes a significant point about his
Frozenness Hierarchy by asserting that “any idiom marked as belonging to
one level is automatically marked as belonging to any lower level” (Fraser
1970: 39). Thus, if an idiom is marked, for example as level L3-
Permutation, it can undergo naturally all the operations stated for that
6
Fraser (1970) points out that in cases where the entire sentence is idiomatic (e.g.
proverbs) and when they permit passivization, the appropriate level for them in
this hierarchy would be L5-Reconstitution, not L4-Extraction.
42 Chapter One
level, but also the transformations included in the lower levels, levels L2-
Insertion and L1-Adjunction. For example, pass the buck to “to attribute to
another person or group one’s own responsibility” is analysed as
belonging to level L5. This means that any reconstitution operation is
permissible, including the action nominalization, but also any other
operations lower in the hierarchy are also tolerable for this idiom. But, the
idiom blow off some steam “to give vent to one’s repressed emotions”
belongs to level L1, and the only modification it allows is an adjunction
operation (the gerund nominalisation), with no other alterations possible.
Finally, a phrase keep watch over “to observe with continuous attention” is
marked as belonging to level L4, which predicts that extraction (the
passive and prepositional phrase pre-posing), insertion (adverbial
placement) and adjunction (gerundive nominalization) can be applied to it,
but that reconstitution (the action nominalization) cannot (Fraser 1970: 39-
40).
In short, Fraser’s (1970) hierarchy ranges from completely frozen
idioms to free collocations. Even though Fraser’s (1970) hierarchy has
never received empirical support, undoubtedly, it is a light in the tunnel to
understand the lexical flexibility of idioms. It is still valid today, being
often cited and referred to.
Then, a year before Fraser (1970), Weinreich (1969) attempted to set
out mathematical formulae to express the structures of idioms. Indeed,
Weinreich’s (1969) and Fraser’s (1970) work is respected and well-known
in the area of idiom studies. There were also other noteworthy studies on
idiom structure during the 1970’s, such as for instance, those of Makkai
(1972), and Newmeyer (1974). Makkai (1972) examines the question of
frozenness and restricted grammatical transformations by his
morphological bans. “The compulsory plural and compulsory singular
ban” means, in practice, that in idiomatic expressions the plural and
singular form cannot be changed freely (Makkai 1972: 122-123). Makkai
(1972) exemplifies the rule as follows: hammer and tongs “to argue, fight”
ĺ *hammer and tong; the skin of my teeth “miraculously” ĺ *the skin of
my tooth. Here, Makkai (1972) states that the second form of the two
idioms is not grammatical.
Interestingly, in Longman Dictionary of English Idioms (1979: xiv-
xix), it is noted that even though many idioms are so frozen that any
additional words cannot be put within the phrase, there are still some
exceptions, i.e. adjectives and adverbs are often allowed to be inserted
within many idioms, e.g. to go (all) to pieces “to break down.” Secondly,
but perhaps not preferably, impolite and swear words may be inserted in
frozen idioms, playing the role of intensifiers, e.g. he went the whole
bloody / damn hog “to do something thoroughly, or too well.” Moreover,

some idioms can be widely varied. Some give rise to other word forms
(e.g. to split hairs “to exaggerate, to see all the possible troubles around”
ĺ hair-splitting) and some permit wide variations, e.g. day in, day out ĺ
where day can be replaced by almost any word which denotes a period of
time, e.g. week, night, month, year, etc.
Additionally, in the more recent studies, Cutler (1982: 317) examines
whether syntactic frozenness could be correlated with the length of time
that the idiom has occurred in a language. Having compared 131 idioms,
used by Fraser with his different levels of frozenness, with the Oxford
English Dictionary, in order to find out the earliest citation marked for
each idiom, Cutler (1982: 319) confirmed that even though frozenness and
age do not correlate completely, there exists an unfailing tendency that the
older an idiom is, the more frozen it is. Cutler (ibid.) presents two reasons
why this should be so. Firstly, syntactic freezing seems to be a gradual
process which may last for (decades or) centuries. Secondly, an idiom
becomes syntactically frozen when its meaning is no longer apparent, due
to the fact that its original literal reference has become obscured. Thus, for
example, let off steam “lose one’s temper” belongs to Fraser’s (1970) level
0, since steam engines have nowadays been replaced by electric machines
(Cutler 1982: 319). The so called pseudo idioms in archaic forms are a
result of frozenness, they cannot be altered at all. In modern English, the
idiomatic meaning is the only meaning pseudo idioms have left. In earlier
times, the words in pseudo idioms also had a literal meaning. This can be
seen in the idiom hither and yon “here and there.” Since the idiom has an
archaic form, and neither “hither” nor “yon” are used alone in the modern
language, it is called an pseudo idiom. (cf. Runosalo 2005: 19).
Besides, Fernando (1996) treats fixedness of form as a basic attribute
of an idiom, more essential than any other distinctive feature, including
figurativeness. Gläser (1988: 266), on the other hand, proposes that,
instead of frozenness, semantic and syntactic stability should be used to
characterise idioms. In addition, Stock et al. (1993) claim that idioms
accept quite a lot of variation either in syntax (tense, third person singular,
negation, position of particles, part of speech; have an/no axe to grind “to
have a grievance, a resentment with a desire to get revenge or sort it out”
ĺ He had a political axe to grind) or vocabulary (a dog’s breakfast ĺ
dinner “something that has been done very badly”) (Stock et al. 1993:
234). Despite the fact that there are idioms which are completely frozen in
their form, meaning and context (e.g. kick the bucket), many of them can
employ alterations in their grammar, vocabulary, and context (Pulman
1993; and McGlone et al. 1994).
44 Chapter One
In addition, Gibbs and Nayak (1989) mention that in English,

grammatically correct operations affecting idioms, are as follows: adverb
insertion, adjective insertion, present participle, action nominalization, and
the passive. Moreover, Fernando (1996: 42-65, 124-152) lists several
instances of lexical transformations and states that transformable idioms
“may be modified by various transformational operations: addition,
permutation, substitution, and deletion. In each case, the conventional
meaning is varied adding to the interpretative effort on the part of the
addressee” (Fernando 1996: 151). All in all, the truth is that along with the
language changes, both idiom variations change, and linguists’
comprehension of idioms develop. Thus, even though kick the bucket is
usually referred to as an idiom that tolerates only variation in tense, Moon
(1998: 123) argues that also kick the pail and kick the can are possible.
Nonetheless, there exist some worries related both to the frequency and
intensity of lexical modification, as well as to the challenge a language
user has to face while choosing the words that can be replaced within the
idiomatic string (Stock et al. 1993: 233-234). Besides, such alterations
may change the meaning, or at least the tones the idiom denotes. The
context itself may influence the idiom, making it sound more idiomatic or
literal for the receiver: it is literally possible to kick the bucket or shoot
oneself in the foot (Ortony et al. 1978; and Cronk et al. 1993). As
mentioned by Mäntylä (2004: 33), this complexity, and the fact that
idioms are generally used both in speech and in a written language, can be
troublesome for a native speaker forming or interpreting such expressions.
But such a difficulty is even harder and frequent for non-native learners
who attempt to recognize and comprehend idioms. For example, two
idioms might get mixed up, or the overall idiomatic meaning can be
changed or made literal, by replacing an idiom component; consequently,
the whole meaning produced by the expression may become humorous or
unsuitable.
Regardless of these problems and doubts concerning any idiomatic
alterations, lexical variation does exist and language users can produce
new phrases creatively. What is more, Mäntylä (2004: 34) points out that
creativity applied to idiom modifications can be illustrated thanks to the
ambiguous meaning of some idioms and the relationship between their
figurative and literal interpretation, or lexical and syntactical variability.
Alteration of the idiom form and/or vocabulary, to suit the context and
situation while still retaining the features of an idiom, can also bring some
playful results. Thus, flexibility of idioms is strengthened even more due
to one’s creative production of these expressions.
Moreover, Gibbs et al. (1989b) suggest that lexical flexibility of

idioms is constrained by their semantic analysability. Their hypothesis is
that the lexical flexibility of idioms is governed by speakers’ assumptions
about the ways in which parts of idioms contribute to their figurative
interpretations as a whole. The results of their three experiments indicate
that idioms whose individual semantic constituents contribute to their
overall figurative meanings (e.g. go out on a limb “to enter a risky
situation because of having a different opinion, and try to defend it”) were
judged as less disturbed by changes in their lexical pieces (e.g. go out on a
branch) than the non-decomposable idioms (e.g. kick the bucket “to die”)
when their individual words were altered (e.g. punt the pail). Yet, instead
of stressing the significance of semantic analysability in constraining the
lexical flexibility of an idiom, Glucksberg (2001) offers recognising a
speaker’s communicative intentions that considerably influence lexical
flexibility.
Finally, Swinney and Cutler (1979: 531), who analyse a possible link
between frozenness and the lexical status of idioms, argue that the more
frozen an idiom is, the stronger its lexical status appears to be, viz. the
more it is perceived as one word, not as a sentence-like expression
containing separate words and meanings. This statement seems to be quite
logical since the more often the expression appears in the same form, the
easier it is to recognize and memorise it. On the contrary, the more
variable a phrase can be, the more difficulties a language user encounters
to identify it in its altered forms.
1.3.4 Literalness, familiarity and predictability of idioms

Comprehending idioms is inevitably related to processing and understanding
them. There are several factors, such as idiom familiarity, transparency,
and context of idiom usage, that are found to influence idiom
comprehension. However, the literalness and predictability of idiomatic
phrases also belong to crucial dimensions of idiomaticity, playing a
significant part in idiom comprehension (Liu 2008).
To begin with, as noted by CieĞlicka (2004: 106), literalness denotes
the degree to which an idiomatic string can be interpreted in a literal mode
(cf. Popiel and McRae 1988). Thus, while the idioms break the ice or have
cold feet are literal, in that they can be taken to mean both literally and
figuratively, the non-literal idiom make a clean sweep (which means “to
eliminate an unwanted person or thing”) has only an idiomatic reading.
Likewise, Tabossi et al. (2011: 113) notice that both go bananas “to
become wildly irrational” and shoot the breeze “to converse aimlessly; to
46 Chapter One
chat” have no literal interpretation, although for two different reasons.

While the former is syntactically ill-formed as “go” is an intransitive verb
that cannot usually take a direct object; the latter is syntactically
acceptable, but semantically atypical as “the breeze” is not the kind of
object that can normally be shot. Besides, traditionally, English idioms
without literal corresponding items are less syntactically flexible than
idioms that have a literal counterpart (Fraser 1970). Contrary to this
viewpoint, Gibbs and Nayak (1989) argue that non-literal idioms are
recognized as more decomposable, thus more flexible than literal idioms.
The aspect of literalness is also studied by Mueller and Gibbs (1987)
(cited by CieĞlicka 2004: 106). In their research they first hypothesize that
not all idioms are equally represented in the mental lexicon, but depending
on the number of their possible meanings, they will have multiple entries
in the lexical storage, instead. These assumptions are confirmed by the
data which prove shorter processing times for idioms with distinct literal
and figurative meanings than for idioms with either isomorphic literal and
figurative meanings, or with only one figurative interpretation. In short,
Mueller and Gibbs (1987) demonstrate that every meaning of an idiom is a
separate entry in the lexicon and that idioms themselves do not belong to a
homogenous class in terms of their storage, access and processing (cf.
CieĞlicka 2004: 107). Mueller and Gibbs’ (1987) suggestion about idiom
variety as regards their storage and processing has been accepted by
CieĞlicka (2004: 107), while their idea of “multiple entry” sounds
ambiguous, due to the lack of solid empirical verification. Consequently,
“the obtained faster processing times for idioms with more meanings
might have been caused by such factors, uncontrolled in Mueller and
Gibbs’ (1987) study, as idiom familiarity, predictability, transparency
among others” (CieĞlicka 2004: 107).
What is more, Mueller and Gibbs’ (1987) results, confirming faster
processing of idioms with both literal and figurative meanings when
compared to those with only a figurative interpretation, may be contrasted
with the effects obtained by Brannon (1975, cited in Swinney and Cutler
1979). Brannon (1975) claims that it takes longer to classify a sentence
when it comprises a literal idiom than when the sentence includes an
idiom with only a figurative meaning. Thus, for Brannon (1975) figurative
idioms are processed faster than the literal ones.
Additionally, literalness of an idiom is sometimes called idiom
transparency, defined as the degree to which the meaning of an idiom can
be derived from the constituents of an idiom. Thus, as has been mentioned
in the previous section, transparent idioms will be those whose figurative
meaning can be deduced from the literal interpretation of their components
(e.g. play with fire “to do something dangerous or risky”). On the other
hand, opaque idioms are those whose meaning cannot be derived from the
analysis of their constituents (e.g. kick the bucket “to die”). Idioms also
vary as regards their well-formedness, that is their syntactic structure can
be well-formed, as in paddle your own canoe “to handle one’s own
affairs,” or ill-formed, as in go bananas “to become wildly irrational.”
Lastly, idioms can be literal (ambiguous), when they offer both the literal
and figurative interpretations, e.g. break the ice, which if literally taken,
means “to cut the frozen surface,” and if figuratively treated, it means “to
do or say something to relieve tension or get conversation going.”
However, idioms can also be non-literal (unambiguous), when only the
figurative interpretation is probable, e.g. drink somebody under the table
“to be able to drink more alcohol than someone else” (cf. CieĞlicka et al.
2008).
Yet, the impact of literalness on the comprehension of idiomatic
phrases is also shown in the studies of Titone and Connine (1994a),
supporting the compositional approach to idioms. They confirm the
significant contribution of literalness to the activation of idioms literal and
figurative interpretation in the course of idiom processing.
So far, the study of the influence of literalness on idiom processing has
yielded inconsistent results. Brannon (1975), and Popiel and McRae
(1988) find longer classification times for literal idioms, compared to
idioms with only a figurative meaning. Reverse outcomes are obtained by
other researchers in their multiple tasks (Swinney and Cutler 1979; Estill
and Kemper 1982; Schweigert 1986; Mueller and Gibbs 1987; Schweigert
and Moates 1988; and Cronk and Schweigert 1992; among others).
Tabossi et al. (2011: 113) add that in the incidental-memory experiment
conducted by Botelho da Silva and Cutler (1993), the participants
remembered idioms much better than control strings, but with no
difference in their ability to remember literal and non-literal idioms.
Titone and Connine (1994a) investigate the impact of literalness on the
processing of idiomatic expressions, and they confirm that literalness
determines the activation of the literal and figurative meaning of idioms
during their processing. In brief, literal idioms, unlike non-literal ones,
reveal less activation of the figurative meaning and more activation of
their literal interpretation of the constituent words of the idiom string.
In tum, familiarity, as the other key factor in idiom comprehension, is
identified by Titone and Connine (1994b) as the “frequency with which a
listener or reader encounters a word in its written or spoken form and the
degree to which the meaning of a word is well known or easily
understood” (Titone and Connine 1994b: 250). In addition, Cronk and
48 Chapter One
Schweigert (1992) study the effects of familiarity, literalness and usage on

the comprehension process. They recognize familiarity as the subjective
frequency of the figurative meaning of an idiom. Subjective ratings of
idiom familiarity are usually provided by the language users themselves. A
more familiar English idiom can be exemplified by the phrase pain in the
neck “someone / something annoying,” while a less familiar idiom would
be, e.g. paddle his own canoe “to handle one’s own affairs.” Cronk and
Schweigert (1992) examine the relationship between familiarity and the
idiom figurative and literal meaning. In their study, they measured with
computer the reading times as the participants were reading the sentences
with idiomatic expressions. Their results confirm that the idiom
comprehension is reliant both on literalness as well as on familiarity of
idioms. Nevertheless, Cronk and Schweigert (1992: 138) express surprise
that familiarity has only an effect on idiom figurative use, and the higher
literalness an idiom has, the faster it is understood. Hence, idioms that are
unfamiliar and have low ratings in literalness cause most problems in
interpretation. Cronk and Schweigert’s (1992: 139) results verify that
figurative meanings of idioms are processed more quickly than literal
ones. However, Cronk and Schweigert (1992: 140) realize that their results
are not consistent with Gibbs’ (1980) theory, since figurative meanings are
not understood more rapidly in the case of less familiar idioms.
On the other hand, Popiel and McRae’s (1988) study revealed different
frequency and familiarity ratings for literal and figurative senses of
idioms, which indicates, according to the authors, that these variables
should be carefully controlled in the future analyses, and that not having
controlled them properly might have resulted in inconsistency in the
previous idiom studies (cf. CieĞlicka 2004: 107). In addition, Cronk and
Schweigert (1992), who examined the effects of literalness and familiarity
on the processing of idioms inserted in sentences, refer to the dimension of
idiom literality as literalness. Their results demonstrate, as mentioned by
CieĞlicka (2004: 108), that sentences with idioms having likely literal
interpretations (high-literalness idioms), e.g. he had his hands full, have
shorter reading times than sentences containing idioms unlikely to be
interpreted literally (low-literalness idioms), e.g. a thousand dollars is
chicken feed “extremely little, insufficient.” Besides, it takes longer to
read the sentences with less familiar and low-literalness idioms than the
sentences with familiar figurative meanings and low literalness. In their
further research, Cronk, Lima, and Schweigert (1993) prove that an idiom
is processed fastest when it is both highly frequent and highly familiar.
Reagan’s (1987) study, on the other hand, reveals a statistically
remarkable correlation between flexibility and familiarity. The more
familiar an idiom is assessed, the more probably it is regarded to be

acceptable in various syntactic transformations. In brief, a much larger
number of syntactic operations is allowed for familiar idioms than for
unfamiliar ones.
Furthermore, the predominance of familiarity over unfamiliarity in
processing idiomatic sentences is also confirmed by Flores d’Arcais
(1993), whose analyses and obtained reading time data present unfamiliar
idioms as those requiring significantly longer inspection time. Yet, another
valuable result is obtained by Forrester (1995), who examines the extent to
which comprehending idiomatic phrases in context is governed by the
words which constitute such phrases. By applying reading time as a
dependent measure, and by substituting idiomatic expressions with phrases
which retain the equivalent semantic meaning in context, the significance
of familiarity in comprehending idiomatic expressions is confirmed. As
noticed by CieĞlicka (2004: 111), this finding causes a serious problem for
Bobrow and Bell’s (1973) Idiom List Hypothesis, as well as for both
Swinney and Cutler’s (1979) and Gibbs’ (1980) models (to be discussed in
section 1.4 of the chapter), presuming faster access of idiom figurative
interpretations.
In brief, familiarity as variable is well-known to have an impact on
word recognition, and it is a strong predictor of speed and accuracy in
several experimental tasks such as lexical decision and reading aloud
(Gernsbacher 1984: 275). Familiarity has also been explored in idiom
processing. Much research has demonstrated that familiar idioms are
processed quicker and more accurately than unfamiliar ones. Processing
unfamiliar idioms, on the other head, requires contextual information and
common-sense knowledge (Schweigert 1986). Cronk and Schweigert
(1992) examine the connection between familiarity and literalness,
confirming that idiomatic expressions in their figurative interpretation are
read quicker when they have high rather than low familiarity. Schraw et al.
(1988) conclude that both lexicalization and familiarity contribute to the
chance of idiomatic preferences, while only lexicalization contributes
considerably to the comprehension of idiomatic meanings.
Furthermore, the context in which an idiom appears is the next crucial
factor in idiom comprehension. As mentioned by Mäntylä (2004: 35),
idioms are hardly ever neutral, as their constituents do usually carry a
certain connotation or style. Consequently, when using idioms, their style
and context ought to be taken into consideration, which has also been
highlighted in the field of idiom research (e.g. Fernando 1996: 101).
Subtle nuances in idiom meaning are sometimes abandoned; thus, Stock et
al. (1993: 231-233) emphasise that although two idioms might be close to
50 Chapter One
each other in meaning, there is some degree of slightly distinguishing

difference that may be recognized only if idioms are not oversimplified.
For instance, kick the bucket is usually understood as “to die,” while it
actually means “to die by natural causes, and relatively suddenly too”
(Stock et al. 1993: 233). Briefly speaking, idioms are often highly context
and register determined, which makes them even more complex.
Moreover, Liontas (2003) conducted an experiment with twenty-eight
adult third-year learners of Spanish in order to investigate the effect which
context (or the lack thereof) has on idiom understanding, among others.
His findings indicate that the use of context is of great importance in the
construction of idiomatic meaning. Additionally, as assumed by Liontas
(2003), the results show that the lack of context has a negative effect upon
the accuracy of idiom interpretation by L2 learners. Thus, as proved by
Liontas (2003), “context has a significant main effect on learners’ success
in accurately comprehending and interpreting VP idioms” (Liontas 2003:
299).
What is more, CieĞlicka et al. (2009) establish that context does play a
crucial role in language processing, although the opinions about the exact
point during processing at which context affects language comprehension
vary. In the broad spectrum of psycholinguistic literature, two radically
different views concerning the issue of how context influences lexical
access have been offered: the modular (context-independent) view and the
direct access (context-dependent) view; while a hybrid model of Giora’s
(1997, 1999, 2002, 2003) graded salience hypothesis, has gained the
greatest popularity. Giora’s (1999) model emphasises the priority of the
so-called salient meaning of words / expressions defined as their
“lexicalized meaning, i.e. the meaning retrievable from the mental lexicon
rather than from the context, e.g. the literal meaning of novel metaphors
but not their intended, non-literal meaning made available by context”
(Giora 1999: 919). Following Giora’s (1999) approach to the role of
salience and context in the processing of potentially ambiguous multi-
word phrases in both native language (L1) and foreign language (L2),
CieĞlicka et al. (2009) conducted an experiment. Their results prove that in
the neutral context, when no clue exists as to the interpretation of the
approaching ambiguous multiword expression, it is the literal meaning that
is activated automatically and more strongly in the bilingual mode. Longer
reading times for the disambiguating regions following non-salient,
figurative meanings of phrases are hence “compatible with the graded
salience view, under which the clash between the automatically activated
salient meaning and the disambiguating non-salient interpretation requires
extra processing time” (CieĞlicka et al. 2009: 302).
Further studies have also highlighted the significance of context in

idiom processing. Context has been revealed to play an essential role in
suppressing irrelevant meanings, but its effects were modulated by
salience (prominence) of idioms (CieĞlicka 2011). Besides, CieĞlicka and
Heredia (2011) indicate that context and salience effects are considerably
modulated by the language (native vs. non-native) of the stimulus
materials being presented to each hemisphere. But no significant
differences between the right and left hemispheres are found in terms of
their sensitivity to contextual constraints. Moreover, CieĞlicka (2013)
explores possible cerebral asymmetries in the processing of decomposable
and non-decomposable idioms by fluent non-native speakers of English. In
her experiment, native language (Polish) and foreign language (English)
decomposable and non-decomposable idioms were embedded in
ambiguous (neutral) and unambiguous (biasing figurative meaning)
context and presented centrally, followed by laterally presented target
words associated with the figurative meaning of the idiom or literal
meaning of the last word of the idiom. Consequently, the obtained results
suggest that a number of factors, such as language status (native vs. non-
native), salience, or context, instead of compositionality per se, emerge as
decisive in determining idiom processing. In short, CieĞlicka (2013)
concludes that no matter how persuasive idiom compositionality appears,
“lack of rigorous procedures and inconsistent classifications of idioms into
one or another category cast doubt on the idea that idioms varying in
compositionality are stored and processed differently in the course of their
immediate on-line comprehension” (CieĞlicka 2013: 484).
What is more, CieĞlicka et al.’s (2014) recent study reports an eye
movement factor and the impact of salience, context, and language
dominance on the processing of idiomatic expressions. Having recorded
eye movements of Spanish-English bilinguals, while reading ambiguous
(literally plausible, such as kick the bucket) English idioms, they
investigate whether the degree of literal and figurative activation in
bilingual idiom processing may be determined by language dominance
(i.e. dominant vs. non-dominant). Each idiom is used either in its
figurative or literal meaning, and put in a sentence with a neutral
preceding context, when its figurative or literal meaning becomes clear,
because of the subsequent disambiguating information, or the preceding
supportive context, evidently biasing one of the meanings. The data
obtained from this study provide convincing evidence that the effects of
salience and context on eye movement patterns are controlled by language
dominance.
52 Chapter One
Finally, numerous research with children has highlighted the important

impact of context on idiom comprehension (e.g. Levorato et al. 2007). It
has been found out that children understand idiomatic expressions more
precisely when they are exposed in informative contexts than when they
are offered in isolation. When encountered out of context, idioms tend to
be interpreted literally. The ability to use contextual information in
language processing has also been found to influence children’s
performance in idiom comprehension. Adults, however, are more affected
by the familiarity of the idiom.
The next factor, which has a significant influence on idiom
comprehension, is its predictability. As mentioned by Tabossi et al. (2011:
112), as far as idioms are concerned, predictability is defined as the
likelihood of completing an incomplete string in an idiomatic style, e.g.
“Mary is in seventh…” is usually completed by speakers with “heaven” to
form an idiomatic expression, even though some other options are possible
to make the phrase literal (e.g. row / place). Generally speaking,
predictability is a characteristics of sentences or expressions that may
influence the lexical processing of a forthcoming word, which may be
facilitated in a predictable context compared with a less predictable one
(Schwanenfluegel and Shoben 1985; and Rayner and Pollatsek 1989;
among others). Moreover, the dimension of idiom predictability
significantly affects the time course of activation of literal and idiomatic
meanings of the idiom during its comprehension, as confirmed by Cacciari
and Tabossi (1988), who prove that, when processing predictable idioms,
only their figurative connotations are activated, by the time the last word
of an idiomatic string is met. Instead, with unpredictable idioms, whose
figurative meanings do not bring any associations until after the whole
string has been processed, only the literal meaning is active at the end of
the idiomatic phrase. Similarly, Titone and Connine (1994a) find
predictability to influence the activation of figurative and literal senses of
an idiom, during its comprehension. In addition, unquestionably,
predictability is to be proved a helpful factor while discussing the
Configuration Hypothesis of idiom recognition (Cacciari and Tabossi
1988) in the subsequent section, 1.4. In this view, idioms are not
represented as individual items, but as configurations of words. The words
that are stored in the lexicon are the same and processed during literal
language comprehension. Nevertheless, as Tabossi et al. (2011: 112)
explain, speakers know that some arrangements of words, e.g. kick
brought together with the bucket, have a figurative meaning (e.g. “to die”),
and while recognising a string of this type, the figurative meaning
connected with it is retrieved from memory. Tabossi et al. (2011: 112)
note that an idiomatic sequence during speech comprehension is processed

word by word, just like any other piece of language, until enough
information is gathered to render the sequence of words distinguishable as
a memorised idiom. The string idiomatic meaning can be activated without
a context biasing the figurative interpretation, only when it becomes
predictable, and is known as an idiom.
Furthermore, it is worth mentioning another characteristics of idioms,
the well-formedness or ill-formedness of the literal meaning of idioms,
emphasised by CieĞlicka (2004: 113). If an idiom has a well-formed literal
meaning, well-formed syntactic constituents are expected, in accordance
with the rules of grammar (e.g. hit the books, or kick the bucket). While
others, ill-formed expressions, consist of components which violate the
general grammar rules (e.g. pop the question). Disrespecting selection
restriction rules among idiom constituents makes an idiom semantically
ill-formed, whereas syntactic ill-formedness, usually violates
subcategorization restrictions (e.g. in the idiom to be in the know, the verb
‘know’ becomes a noun). Besides, Gibbs and Nayak’s (1989) research
reveals that syntactic flexibility of idioms is not firmly determined by their
literal well-formedness, and that syntactically flexible idioms do not need
to have well-formed literal meanings. Moreover, Mueller and Gibbs
(1987) confirm that processing literally well-formed idioms takes less time
than comprehending ill-formed idiomatic expressions (cf. CieĞlicka 2004:
114-115).
Finally, Mäntylä (2004: 34) notices that, interestingly, idioms are
predicted to be encountered more often in informal, spoken language
rather than in written or more formal language. This assumption may
appear to be partly true, as it is newspapers that mostly have drawn the
attention of phraseologists (e.g. Fernando 1996; and Moon 1998: 69-71),
due to numerous examples of idioms and their different variations
identified in newspaper headlines. Whether in formal or informal, oral or
written styles, undoubtedly, idioms effectively arise interest of the viewers
or listeners; thus, they are frequently used by both writers, presenters,
journalists, sports and TV reporters, among others. Moreover, Strassler
(1982) emphasises that idioms are most often employed when talking
about a third person or an object, hardly ever referring to the speaker or
receiver him/herself. So, there may exist some restrictions with regard to
their use in the presence of speech participants.
To recap, different researchers agree that the level of difficulty in
idiom comprehension differs across the different dimensions along which
idiomatic expressions vary. The dimensions analysed in this section
included literalness, familiarity and predictability of idioms. Certainly, a
54 Chapter One
lot of attention has been paid to these characteristics, which concern the
role of context, well-formedness of idioms, and the level of their
formality. So far, in the study of idiom syntax, knowing the syntactic
behaviour of individual idiomatic expressions has been a precious tool that
can be applied to manipulate the syntactic variable experimentally
(Tabossi et al. 2011: 113). Undoubtedly, both the syntactic productivity
and the lexical creativity of idioms are matters of degree, depending on the
idiom compositional properties. This conclusion indicates that idioms do
not form a unique class of linguistic items, but share many of the
properties with more literal language. Indeed, understanding the syntactic
behaviour of idioms is a fundamental mission for any theory of idiom
representation and processing, which is to be analysed in section 1.4, and
several alternatives are still under debate (Gibbs and Nayak 1989; and
Sprenger et al. 2006).
1.4 Hypotheses and models of idiom representation

and processing
The figurative language literature can faithfully mirror the development of
idiom representation and processing models, which parallels the evolution
of theoretical approaches related to idiomatic expressions. In general,
theoretical accounts of idiom representation and processing can be divided
into two main classes: non-compositional theories and compositional ones.
However, the strict bipolar division has proved to be insufficient in the
course of time, on account of the results of the psycholinguistic research
into phraseological units. Consequently, hybrid approaches and the model
of dual representation of idioms have emerged, as a result of trying to
combine the two previous models.
1.4.1 Non-compositional Models

In the past few decades, both psychological and neuropsychological
approaches have begun to examine thoroughly the nature of idioms in
various languages, giving rise to a series of competing models with respect
to idiom lexical representation and processing. The non-compositional
view of idioms (Weinreich 1969; Fraser 1970; Katz 1973; and Chomsky
1980; among others) treats idiomatic phrases as non-compositional strings
whose figurative meanings are not directly related to the literal meanings
of their individual parts. Referring to non-compositional models of idiom
processing as direct look-up models, Glucksberg (1993: 4) means the fact
that all of them are specified arbitrarily and comprehended by retrieving
T
Towards Definiing an Idiom 55
the meaningg of an idiom as a a whole, raather than by llinguistic proccessing of

their constituuent parts. Thhe Idiom List Hypothesis
H (B
Bobrow and Beell 1973),
the Lexical RRepresentatioon Hypothesis (Swinney andd Cutler 1979 9) and the
Direct Accesss Hypothesiss (Gibbs 1980 0, 1984; Schw weigert 1986) constitute
c
the three oof the most recognized versions of the traditional non-
compositionnal model of idiom analyssis, which aree to be scruttinised in
sections 1.4..1.1, 1.4.1.2, and
a 1.4.1.3, reespectively.
1.4.1.1 The Literal First Hypothesis (B

Bobrow and B
Bell 1973)
The literal pprocessing model,

m developped by Bobroow and Bell (1973),
( is
referred to iin the literatuure as the Idioom List Hypoothesis (cf. Gllucksberg
1993), the L Literal First Hypothesis
H (ccf. Cronk andd Schweigert 1992), or
the Serial PProcess Model (cf. Gibbs and a Nayak 19989). Bobrow and Bell
(1973) proppose that undeerstanding idiioms occurs iin three stagees. In the
first stage, the languagee comprehen nsion means figuring out a literal
interpretatioon of the strinng. But if th his linguistic analysis failss and the
literal meanning is excludeed, idiomatic meanings em merge from thee lexicon.
This model,, then, assum mes that literall meanings off ambiguous idiomatic
expressions should be proocessed faster than their figgurative interp pretations.
In other woords, the literral first hypothesis proposses that, only y after an
appropriate literal meaninng is not foun nd, a mental iddiom list is acccessed to
retrieve the idiomatic meaaning. Besides, Bobrow andd Bell (1973) state that
idioms are sstored in a sppecial list of idiomatic exppressions, or an idiom
lexicon, sepparate from the word leexicon. Figuure 1-3 below is an
illustration oof Bobrow andd Bell’s (1973 3) Literal Firsst Hypothesis.
Figure 1-3. Literal First Hypothesis

H accorrding to Bobrow
w and Bell (197
73)
(own source)
In brief,, Bobrow and Bell (1973 3) propose twwo separate modes

m of
processing, a literal and a figurative one. o The literaal meaning is accessed
first, and onnly after its rejection is the idiomatic mean
aning retrieved
d.
56 Chapter One
1.4.1.2 The Lexical Representation Hypothesis

(Swinney and Cutler 1979)
In contradistinction to the previous model, the Lexical Representation

Hypothesis posits that idioms are stored along with other words in
memory. Swinney and Cutler (1979) implicate parallel activation of both
the literal and the figurative meaning. Then, the idiomatic meaning is
processed first, as it is fixed and stored in a separate list. Figure 1-4
illustrates Swinney and Cutler’s (1979) Lexical Representation
Hypothesis.
Figure 1-4. Swinney and Cutler’s (1979) Lexical Representation Hypothesis

(own source)
Research conducted to support Swinney and Cutler’s (1979)

hypothesis shows that people recognize grammatical idioms, presented out
of context, as meaningful expressions more quickly than non-idiomatic
phrases. Follow-up studies have produced either supporting findings
(Estill and Kemper 1982; Glass 1983; and Botelho da Silva and Cutler
1993; among others) or contradictory evidence (Burt 1992).
1.4.1.3 The Direct Access Hypothesis (Gibbs 1980, 2002)
The third option, the Direct Access Hypothesis (Gibbs 1980, 2002),
illustrated in Figure 1-5 below, assumes that an idiom’s figurative
meaning can be activated without the literal meaning being processed first.
Gibbs’s studies show that, given a suitable context, the idiomatic meaning
is processed sooner than the literal meaning (cf. Gibbs 1985; 1986;
Schweigert 1986; Schweigert and Moates 1988; and Needham 1992;
among others). It is possible that people completely bypass such mode
when faced with a highly familiar idiom, or when they have a sufficient
context to infer an idiomatic interpretation. In these cases, the idiomatic
meaning is directly accessed.
Figure 1-5. The Direct Access Hypothesis according to Gibbs (1980, 2002)
(own source)
To conclude, the basis for all the three hypotheses fundamentally

implies the very idea that the meaning of an idiom is stored in a separate
mental idiom list (Weinreich 1969). On the other hand, there are several
pieces of criticism against the non-compositional models, presented above.
First, the studies of Swinney and Cutler (1979), or Titone and Connine
(1999), among others, demonstrate that idiomatic expressions are not
processed more slowly than literal expressions. In fact, the opposite is
often the case, which goes against the prediction of the Literal First
Hypothesis. Second, idioms have been found to be more than just frozen
phrases or long words. For example, some idioms can be transformed to
some extent, and still be recognized and understood, e.g. the idiom spill
the beans can be used as “the beans were spilt by Mary.” This is possible
because spill the beans can be mapped on the meaning “to reveal the
secret,” i.e. spill (reveal) and beans (secret). Such an idiom shows that the
internal structure of the word strings matters during comprehension. The
meaning of some idioms, like play with fire “to do something dangerous or
risky,” can also be inferred from the literal interpretation of their
components (to do something dangerous). These findings reveal that
idioms are not a homogeneous, distinct group, and thus may not involve
different processing strategies from those valid for literal expressions
(Titone and Connine 1999).
1.4.2 Compositional Models

In contradistinction to the non-compositional models, compositional
theories propose that idioms vary with respect to their compositionality,
that is, the degree to which the literal meanings of their constituent words
contribute to their overall figurative interpretation varies. Several lines of
research have convincingly shown that idiom processing cannot be
exclusively reduced to the holistic retrieval of a lexicalized meaning, and
that it involves an obligatory semantic and syntactic analysis of its
constituent words (cf. Cacciari and Tabossi 1988; Glucksberg 1993;
Titone and Connine 1994a; and Tabossi and Zardon 1995; among others).
Most of the research undertaken within the compositional framework has
58 Chapter One
thus attempted to investigate the varying degree to which literal meanings

of idiom constituents contribute to their overall figurative interpretation.
Major compositional models of idiom storage and their comprehension
that developed in the literature include the Configuration Model (Cacciari
and Tabossi 1988), the Conceptual Metaphor Hypothesis, and the Idiom
Decomposition Hypothesis (Gibbs and Nayak 1989; Gibbs, Nayak, and
Cutting 1989). Within a compositional account, individual meanings of
idiom components play a significant role in constructing the overall
figurative interpretation of an idiom.
1.4.2.1 The Configuration Hypothesis (Cacciari and Tabossi 1988)
In the Configuration Model, proposed by Cacciari and Tabossi (1988) and

illustrated in Figure 1-6 below, the idiomatic configuration is recognized
via the activation of the idiom key.
Figure 1-6. Cacciari and Tabossi’s (1988) Configuration Hypothesis

(own source)
As seen in Figure 1-6, in the Configuration Model, the configuration takes

its specific idiomatic meaning, while the literal meanings are still being
activated. The processing time may vary, depending on the position of the
idiomatic key within the configuration. This fact could not be explained by
the ‘first generation’ hypotheses (discussed in section 1.4.1), and thus
makes Cacciari and Tabossi’s (1988) model superior to them.
1.4.2.2 The Conceptual Metaphor Hypothesis (Gibbs, Bogdanovich,

Sykes, and Barr 1997)
In the Conceptual Metaphor Hypothesis, Gibbs et al. (1997) propose that

metaphors are fundamental to human thought, and they influence the
comprehension of many aspects of language, including idioms. An
example of a conceptual metaphor is love is a journey.7 This metaphor is
embedded in the idiom-containing sentences like their marriage is on the
rocks and our relationship is at a cross-road. This hypothesis, illustrated
7
Cf. Footnote 5.
in Figure 1-7 below, suggests that conceptual metaphors facilitate

understanding idiomatic expressions.
Figure 1-7. The Conceptual Metaphor Hypothesis (Gibbs, Bogdanovich, Sykes,

and Barr 1997) (own source)
In fact, conceptual metaphors are activated during idiom comprehension.

Individual words in the idiom can metaphorically contribute to its
figurative meaning. The words associated with the metaphor (e.g. journey)
were more quickly identified as meaningful, after the presentation of the
idiom.
1.4.2.3 The Idiom Decomposition Hypothesis (Gibbs and Nayak 1989;

Gibbs, Nayak, and Cutting 1989)
Gibbs and his colleagues (Gibbs and Nayak 1989; and Gibbs et al. 1989a)
introduce the Idiom Decomposition Hypothesis, within the compositional
framework. Although often cited as such, the Idiom Decomposition
Hypothesis is not an assumption about the processing of idioms;
decomposability is an influencing variable with regard to comprehension
or representation of idioms (Gibbs et al. 1989a).
To begin with, the Idiom Decomposition Hypothesis is about the
analysability of idioms, viz. the “speaker’s assumptions about how the
meaning of the parts contribute to the figurative meanings of the whole”
(Gibbs and Nayak 1989: 104). A decomposable idiom is an idiom whose
individual components contribute to its figurative meaning (e.g. play with
fire “to do something dangerous or risky”), while idioms whose individual
elements do not make such a contribution are non-decomposable (e.g. kick
the bucket “to die”). But true decomposability is a feature of idioms that is
relevant from a psycholinguistic point of view and it is based on speakers’
judgements; whereas compositionality is a theoretical assumption about
the combination of syntactic constituents and their phrasal or sentential
meanings, which is important within linguistic theories such as generative
grammar.8 From the generative, syntactic point of view, only the literal
8
The compositionality principle, as a basic assumption of generative grammar,
goes back to Frege (1884 / 1980), and holds that the “meaning of an expression is a
function of the meanings of its parts and of the way they are syntactically
60 Chapter One
meaning of an idiom is compositional, while the figurative meaning is

always non-compositional. Therefore, decomposable idioms can be partly
compositional, whereas non-decomposable idioms are usually truly non-
compositional (cf. Hamblin and Gibbs 1999).
Furthermore, the results of CieĞlicka’s (2013) study, referring to
possible cerebral asymmetries in the processing of decomposable and non-
decomposable idioms by fluent non-native speakers of English (cf. section
1.3.4), prove to be inconsistent with the Idiom Decomposition Hypothesis
(Gibbs et al. 1989a; 1989b), and only partially consistent with the idea of
the differential cerebral involvement in processing (non-)decomposable
idioms (cf. Beeman’s 1998 Fine-Coarse Coding Theory). That is why, as
noted by CieĞlicka (2013), a number of factors, rather than
compositionality by itself, emerge as essential in determining idiom
processing, such as language status (native vs. non-native), salience, or
context.
In a nutshell, compositional models assume that idiom comprehension
uses ordinary language processing. When an idiomatic expression is
encountered, it is processed gradually like a normal expression. The
components of an idiomatic word string contribute to a figurative meaning
in either a literal or metaphorical way.
1.4.3 Hybrid Approaches

Hybrid accounts of idiom comprehension, processing, and production are
expected to offer the best solution to the problem that any theoretical
approach to idioms necessarily encounters, namely the simultaneously
compositional and non-compositional nature of idiomatic expressions. The
three stances, to be outlined below, are likely to combine the traditional
non-compositional and compositional models, or would constitute a
certain solution to the problems that the previous models have to deal
with. These are respectively: the Model of Idiom Comprehension by
Titone and Connine (1999), the Model of Dual Idiom Representation
(Titone and Connine 1999), and the Hybrid Model of Idiom Production
(Cutting and Bock 1997).
combined” (Partee 1984: 153; and cf. also Dowty 2007: 25). It is responsible for
the great problems that idioms pose within this framework and is the reason why
generative grammarians have proposed various, complex assumptions for the
description of idiomatic expressions. These assumptions were intended to make
idioms fit into the overall compositional pattern (e.g. Weinreich 1969; Fraser 1970;
Katz 1973; and Everaert 1993; among others).
1.4.3.1 The Hybrid Model of Idiom Comprehension

(Titone and Connine 1999)
In the Hybrid Model of Idiom Comprehension, proposed by Titone and

Connine (1999), the insights gained from the Idiom Decomposition
Hypothesis are used, although Titone and Connine do not strictly
distinguish between decomposability and compositionality. The authors
focus on “determining the degree to which idiomatic and literal meanings
are initially computed during idiom processing” (Titone and Connine
1999: 1668). With a relatively tiny sample of 24 participants and 32
idioms, they conducted an eye-tracking study. The results support their
hypothesis that automatically both meanings, i.e. the literal and the
figurative ones, are activated. For non-decomposable (in their terminology
non-compositional) idioms, it takes longer to integrate the correct meaning
into the idiomatic context, because in this case the two meanings are
semantically distinct. The Hybrid Model is superior to the other processing
hypotheses because it controls for the decomposability of idioms. It
allows, as the Configuration Model (Cacciari and Tabossi 1988) does (cf.
section 1.4.2.1), for both the literal and the figurative meaning, to be
activated during idiom processing.
Generally, the present study agrees with the fundamental assumptions
and findings of both the Configuration Model and the Hybrid Model.
However, the comprehension hypotheses discussed above suffer from
certain shortcomings. The studies of the ‘first generation’ (cf. section
1.4.1) are too simple due to their either–or characterization of an idiom
literal or figurative meaning. All these hypotheses restrict themselves to
the native mental lexicon only. Furthermore, they are limited to the
lexical, namely, linguistic level, whereas conceptual aspects during idiom
comprehension are not measured. Studies that deal with conceptual
aspects in idiom comprehension (e.g. Gibbs 1995; and Glucksberg et al.
1993; among others) chiefly concentrate on the conceptual or metaphorical
motivation for the meaning of an idiom, but this aspect has not yet been
integrated into comprehension models. Regardless of an idiom status as
being decomposable or non-decomposable, its figurative meaning has to
be learned and stored separately. However, some studies conducted with
non-native speakers suggest (Abel 2003) that if lexical information, i.e. the
literal meaning of the constituents, is not sufficient, conceptual knowledge
can play a role in the idiomatic comprehension process. In the present
study, the notion conceptual refers to a non-lexical, that is, non-linguistic
aspect of cognition and should not be confused with semantic knowledge
(Pavlenko 1999; and Roelofs 2000). In the context of the L1 and the L2
lexicon, Kroll and Sholl (1992), Kroll (1993), and de Groot (2002), among
62 Chapter One
others, assume that there are language specific, separate lexical

representations but only one conceptual representation, which is
independent of languages or modalities.
1.4.3.2 The Model of Dual Idiom Representation

(Titone and Connine 1999)
Even though over the last five decades, linguists and psycholinguists have
developed a number of hypotheses to describe the distinctive grammatical
characteristics of idioms, and to provide some explanation for their
processing and representation; there are three aspects that have been
neglected. According to Abel (2003), these are: first, the studies deal
exclusively with the native mental lexicon, and do not try to integrate the
second language (L2) lexicon. Second, the studies concentrate either on
lexical representations or on conceptual aspects, but do not try to combine the
two into one theoretical model. Lastly, most of the studies do not allow for
frequency effects to play a role in the representation or processing of idioms.
The Model of Dual Idiom Representation (from now on the DIR
Model), offered by Titone and Connine (1999), is a psycholinguistic
model which includes the above-mentioned aspects. Not only does it
combine the lexical and the conceptual level but it also integrates the
representation of idioms in the first language (L1) and the L2 lexicon.
Abel (2003) presents some supporting evidence for the DIR Model based
on empirical studies on the decomposability of idioms with native and
non-native speakers of English. The insights and results of experimental
morphological studies are also used to confirm the assumptions of this
model. Thus, the Model of Dual Idiom Representation (DIR Model)
assumes that non-decomposable idioms require an idiom entry (at the
conceptual level), whereas decomposable idioms can be represented via
constituent entries, and can additionally develop an idiom entry (both at
the lexical level), as illustrated in Figure 1-8 below.
Figure 1-8. Titone and Connine’s (1999) Model of Dual Idiom Representation
(own source)
For decomposable idioms, the idiom entries are regarded as supplementary

pieces of information about frequently occurring linguistic entities, and not as
a compulsory prerequisite to idiom processing. Moreover, the more
frequently an idiom occurs as an idiomatic configuration, the more
probable the development of an idiom entry is, irrespective of whether the
idiom is decomposable or non-decomposable. Frequency plays a key role
in language processing and should consequently be a part of every model
of idiom representation.
All in all, the Model of Dual Idiom Representation (DIR Model) is a
model that attempts to compensate for the issues neglected by the models
discussed earlier in the literature. It assumes that there is not only a lexical,
but also a conceptual level of representation, and that constituent and
idiom entries co-exist at the lexical level. If decomposable idioms have no
idiom entry at the lexical level, conceptual representations are accessed
during idiom comprehension. The supporting evidence for the dual
representations is gathered from native and non-native judgements on the
decomposability of idioms (cf. Abel 2003).
1.4.3.3 The Hybrid Model of Idiom Production (Cutting and Bock 1997)
Cutting and Bock (1997) represent the first attempt to address questions
concerning the storage and retrieval of idiomatic phrases during language
production. They ran a series of error-elicitation studies in which
participants were briefly presented with two idioms and, after a short
pause, were asked to produce one of them as quickly as possible. The
dependent measures are production latencies and blending errors, that is,
unconventional combinations of two idioms. In the first experiment,
Cutting and Bock (1997) employed pairs of idioms with similar (kick the
bucket, or meet the Maker “to die”) or different (shoot the breeze “to
converse aimlessly; to chat,” or raise the roof “to show great enthusiasm”)
idiomatic meanings, and with the same syntactic form (chew the fat “to
gossip or make a small talk,” or raise the roof “to show great enthusiasm”)
or different (chew the fat “to gossip or make a small talk,” or nip and tuck
“inconclusive as to outcome; close or just even in a race or comparison”)
syntactic forms. What follows is an assumption that if idioms are stored as
unitary forms, then their syntactic structure should have no effect on the
resulting idiom blends. In contrast, if idiomatic expressions do submit to
syntactic analysis in the course of their production, then idioms with a
similar structure should produce more blends than those with differing
structures. Cutting and Bock’s (1997) experiment reveals that same-syntax
idioms, with similar figurative meanings, are more likely to blend than
different-syntax idioms, with different figurative meanings. This evidence
64 Chapter One
is taken by the authors as an argument against a lexicalized view of idiom

storage.
Furthermore, Cutting and Bock (1997) investigate the differential
lexicalization status of decomposable and non-decomposable idioms
hypothesized by Gibbs and Nayak (1989). They thus offered the
participants the pairs of idioms that are either decomposable (hold your
tongue or button your lip “not to speak”) or non-decomposable (shoot the
breeze “to converse aimlessly; to chat,” or chew the fat “to gossip or make
a small talk”). The reasoning adopted was that if lexical representation of
non-decomposable idioms is more unitary, such idioms should be less
susceptible to the production of idiom blends (e.g. shoot the fat) in the
error elicitation task than decomposable idioms, in line with the idiom
decomposition model. On the other hand, decomposable idioms, with
individual components mapping directly onto the idiomatic senses, should
elicit a substantially bigger number of idiom blends (e.g. hold your lip).
The analysis of participants’ responses shows that both decomposable and
non-decomposable idioms elicit a comparable proportion of idiom blends.
Consequently, this result implies, according to the authors, that the lexical
representations of both idiom types are identical, especially as far as the
production process is concerned.
Accordingly, Cutting and Bock (1997) propose a Hybrid Model of
Idiom Production, whose architecture is compatible with the models of
language production, suggested in the psycholinguistic literature (Dell
1986; Levelt 1989; and MacWhinney 2008; among others). The hybrid
model assumes that idioms are stored as whole units at the lexical-
conceptual level of the lexicon. The lexical-conceptual nodes representing
them are connected to the syntactic component of the system retrieving the
phrasal frames and specifying grammatical slots in idiom phrases. Besides,
the lexical-conceptual node representing an idiom is connected in the
lexicon with lemmas9 for individual words constituting the idiomatic
phrase. The model easily explains the increase in a number of blending
errors as a function of structural and meaning similarity, which is
demonstrated in Experiment 1, described above. Idioms with the same
syntactic form share the same syntactic frames, and idioms with similar
meanings activate similar conceptual representations, which results in
more competition than in the case of syntactically or semantically
dissimilar idioms. Since one concept can activate multiple lexical concept
nodes, including those representing idioms, similar meaning or similar
9
The very term lemma refers to a representation of a lexical item grammatical
class information plus a pointer to the word forms (Roelofs 1992; and Levelt and
Meyer 2000; among others).
structure idioms, such as meet your Maker and kick the bucket “to die,” are
activated simultaneously and will compete in the course of language
production, which might lead to blend errors such as ‘meet the bucket’.
Moreover, with the purpose of testing the hybrid account of idiom
production, Sprenger, Levelt and Kempen (2006) examined the production
of idioms in a series of studies employing reaction time paradigms. In their
first experiment, Sprenger et al. (2006) tested the claim that idiom
constituents are the same lemmas which get activated in the course of
producing literal phrases and that idiom phrases have their unitary
representations in the mental lexicon. They, thus, employed a cued-recall
task, in which native speakers of Dutch produced either idiomatic or literal
phrases they had learned earlier, as an answer to a visually displayed
prompt word. While looking at the prompt word, participants heard a
prime, which was either a word identical to the noun of the phrase to be
produced, or a word unrelated to any of the phrase’s elements. The results
showed that both idiomatic and literal phrases are produced faster when
clued-up by one of their content words. This result, according to Sprenger
et al. (2006), supports the view that idioms are compositional phrases,
made up of the same simple units which are activated in the course of
literal language production. In addition, the comparison of the priming
effects found for idiomatic and literal phrases reveals that idioms are
primed significantly stronger than literal utterances. Sprenger et al. (2006)
explain this difference in the strength of priming by postulating a separate
meaning representation for idioms in the mental lexicon. Lemmas which
belong to an idiomatic phrase are, in this way, incorporated into a unitary
lexical entry. Therefore, priming one of the lemmas, belonging to this
common idiomatic representation, results in the spreading of activation to
the remaining lemmas of the idiom entry, and makes them more available
for retrieval.
These conclusions, confirmed in the remaining experiments (cf.
CieĞlicka 2010), explicitly support the Hybrid Account of Idiom
Representation, in which idioms are both unitary and compositional
phrases. Besides, in a post hoc analysis of the results obtained in all the
three experiments, Sprenger et al. (2006) find that idiom decomposability
does not influence the extent to which literal meanings of idiom elements
become activated in the course of idiom production. Taking everything
into account, the study conducted by Sprenger et al. (2006) confirms the
validity of the Hybrid Account of Idiom Representation, proposed by
Cutting and Bock (1997). In an effort to make the hybrid model applicable
to production and comprehension, Sprenger et al. (2006) propose an
extension and modification of the original hybrid model, which they call
66 Chapter One
the superlemma theory.10 One of the advantages of Sprenger et al.’s

(2006) model is that it postulates a simplified mechanism of idiom
production, proposing that it is identical to the mechanism of processing
single words. This is so because idioms are represented with their own
superlemma. This superlemma (for example, skate on thin ice “to be in a
risky situation”) gets activated along with other words and phrase lemmas
(such as risk or gamble) in the course of language production, and
competes with them in the selection process. Another advantage of the
superlemma theory over the hybrid model is, according to Sprenger et al.
(2006), that it can easily account for the differing syntactic flexibility of
various idiomatic expressions. Since syntactic information about idiomatic
expressions is stored in the superlemma, all the constraints operating on a
particular idiom, as well as the actual grammatical relations holding
between its constituents, are coded at the superlemma level. Such a
solution is much simpler than the phrasal frames with open slots proposed
in the hybrid model, which cannot straightforwardly account for syntactic
idiosyncrasies of idioms (cf. CieĞlicka 2010).
To sum up, on account of idiom heterogeneity and a wide diversity of
approaches offered for phraseological units, there is a constant necessity to
provide some rules and theories that would both explain and categorise
idiomatic phrases. In section 1.4, an attempt has been made to overview
some hypotheses and models of idiom representation and processing. The
traditional non-compositional models of idiom representation and
processing differ in terms of how and when idiom meanings are thought to
be retrieved, nevertheless, they share the supposition that idiomatic
meaning is semantically distinct from the meanings of the constituents of
an idiom. Although the compositional hypotheses differ in the ways that
idiomatic meanings are activated, they imply that some relationships can
exist between an overall idiomatic meaning and the individual component
meanings of an idiom. Finally, the hybrid approaches towards idioms
assume that all the previous traditional hypotheses restrict themselves to
the native mental lexicon only. Regardless of the status of an idiom, as
either being decomposable or non-decomposable, its figurative meaning has
to be learned and stored separately. However, if lexical information, i.e., the
literal meaning of the constituents, is not sufficient, conceptual knowledge
can play a profound role in the idiomatic comprehension process.
10
A superlemma is a separate all-inclusive representation of an idiom, introduced
into the hybrid model at the lexical-syntactic processing level in order to account
for the evidence that the syntactic properties of an idiom are in some way
independently represented. This superlemma is linked with the individual lemmas
which make up the idiomatic phrase.
1.5 The working definition of an idiom

The chapter has presented a wide spectrum of definitions of idioms, and a
number of dimensions along which idioms can show their heterogeneous
nature. Some of these idiom characteristics, e.g. analysability, the
fixedness of form and internal structure, predictability and familiarity,
among others, have been shown to affect significantly idiom comprehension.
Others, like formality, and syntactic and semantic well-formedness, still
need to be investigated in connection with their influence upon the access
to and comprehension of idiomatic expressions. Making a contributory
conclusion, CieĞlicka (2004: 115) states that the various dimensions of
idiom variability have also been discovered to correlate with one another,
although the picture of an idiom they form is tremendously varied.
Consequently, it is improbable for idioms to be stored, accessed and
processed in an identical manner. Conversely, they might be represented
in the mental lexicon in a different way, while the activation of their literal
and figurative interpretations during their comprehension depends on a
sum of factors.
Taking everything into account, a working definition should be
introduced to be employed in this book. The literature presents a great
number of attempts that have been made to define idiomatic expressions;
nevertheless, providing a brief and detailed definition of an idiom in an
indisputable way has been proved a mission hardly possible. Similarly, it
has been equally unmanageable to classify any particular expression as an
idiom, collocation, phrase, etc., since idioms constitute only a “subset of
the fixed expressions in a language community” (Glucksberg 2001: 68),
not to mention all the entities that should or should not be subsumed under
this label. As a result, neither the unified scientific approach nor a
linguistic clear view have been offered so far to create the all-
encompassing term of an idiom. Instead, on the account of the widely
heterogeneous nature of idioms, as noticed by Mäntylä (2004: 36), the
emphasis should be put on the meaning of the whole expression, and on its
figurativeness. Besides, since idiomatic expressions differ extremely in
form and structure, the acceptance of idiomatic variability and
heterogeneity seems to be the best way out.
Thus, for the purpose of this study, the basis for a definition of an
idiom is the assumption, based on common linguistic definitions, that,
firstly, in some cases an idiom is a combination of words that is associated
with a meaning that cannot be understood on the basis of the literal
definitions of the individual words it contains (Liu 2008). Hence, an idiom
is recognized as a set phrase semantically opaque, whose meaning cannot
68 Chapterr One
be deduced from the meeanings of its constituents. Even if it sh hows “an

expressive aand particularr value, and sometimes (aa lexical or syntactic)
s
matchless sttructure, it beehaves as a single semantiic unit” (Cariine 2005:
495). On thee other hand, in i contradistinnction to opaqque idioms, traansparent
idioms show w a close relattionship betweeen a metaphoorical and literal sense.
Yet, idioms vary as regards compositio onality, whichh refers to the degree to
which the phhrasal meaninng, once know wn, can contribbute to the idiiom parts,
e.g. in spill the beans thhere is a clearr correspondeence between spill and
beans and the relevantt parts of itss figurative meaning “to o divulge
information..” Besides, idioms also vary in the extennt to which theey can be
syntacticallyy transformedd, still retaininng their idiom matic meanin ng (Gibbs
and Gonzales 1985). Finnally, some id dioms do havve both figuraative and
mbiguous (e.gg. break the icce “to cut
literal interppretation, and are called am
the frozen ssurface” vs. “tto do or say something to relieve tensio on or get
conversationn going”) (Pappagno and Romero Lauro 22010: 21-22).
Secondlyy, an idiom, asa a multi-worrd phrase, shouuld be treated d as a unit
with a combbination of chharacteristics, rather than w with a high inttensity of
one specificc characteristicc. Neither liteeralness alonee nor figurativ
veness, or
fixedness off form, or thee degree of analysability
a aare sufficient when an
idiom appeaars individuallly (cf. Pulman n 1993: 250).
Thirdly, the book wiill focus on th he idioms ideentified as multi-word
m
phraseologiccal units contaaining more th han one wordd, with the excclusion of
proverbs, saayings and connversational phrases,
p even though it is incredibly
difficult to ddraw the bordder lines betw ween the unitss themselves. Mäntylä
(2004: 38), in her Figurre 1-9, displaays both the multitude of different
multi-word expressions and a their muttual correlatioon. As can bee noticed,
since they overlap withh one anotheer, it is harddly possible for these
expressions to be separateed.
Figure 1-9. The rough fieldd of multi-word

d expressions (cf. Mäntylä 2004:
2 38)
As shown inn Figure 1-9, even though all

a the multi-wword expressiions often
connote diff
fferent interprretations from
m their constittuents, they belong
b to
specific categories each. Proverbs with their purpose to draw a moral and
teach a lesson are closely related to cultural aspects; sayings are generally
not figurative; and conversational phrases, at last, with their special
interactional function are more literal than figurative. In a nutshell, Figure
1-9 illustrates the overlap between the boundaries of various multi-word
expressions, making them nearly inseparable from one another.
Fourthly, the terms that are expected to occur in the book, i.e.,
“idioms,” “phraseological units,” “idiomatic expressions,” “multi-word
expressions” or “fixed expressions” seem to be more widely known in
English. They have been chosen to make the subject matter more
straightforwardly accessible to scholars not directly working in the field of
phraseology. Gläser’s (1998) definition of “phraseological units,”
reformulated in (1.8) below, appears to be the most suitable one, as it
comprises virtually all the possible notions, characteristics and entities that
should be encompassed by the term.
(1.8) A “phraseological unit” is a lexicalized, reproducible bilexemic or

polylexemic word group in common use, which has relative syntactic
and semantic stability, may be idiomatized, may carry connotations,
and may have an emphatic or intensifying function in a text. (Gläser
1998: 125)
It is significant, as referred by Knappe (2004: 8), that in Gläser’s

(1998) definition above, the indicators “relative” and “may” suggest a
gradation of the presence of these features, while “lexicalized,” “bilexemic
or polylexemic word group,” “reproducible” and “in common use” are
invariable. The features “reproducible” and “in common use” of
phraseological units seem to directly resemble lexemes. Nevertheless,
phraseological units are not produced as such, but like lexemes they are
reproduced and are regularly used entities of the language. Hence, they
are, to a large extent, prefabricated units (cf. also Cowie 1998a: 1).
Besides, the other features in Gläser’s (1998) definition, “syntactic and
semantic stability” are relative, and “idiomaticity, connotations, and the
emphatic or intensifying” force of phraseological units are non-
compulsory. Connotations of phraseological units, which “enrich [the]
cognitive content [of a word or phrase] by means of emotive and/or
attitudinal semantic markers” (Gläser 1998: 128) are analogous to simple
and complex lexemes. Importantly, Gläser’s (1998) definition correlates
with the recent one presented by O’Dell and McCarthy (2010). The latter
identify idioms as “fixed combinations of words whose meaning is often
difficult to guess from the meaning of each individual word” (O’Dell and
McCarthy 2010: 6).
70 Chapter One
Finally, indeed there have been proposed different typologies of

idiomatic expressions, i.e. the one of Makkai (1972); these of Nunberg
(1978) and his followers (Gibbs and Nayak 1989: 104; Titone and
Connine 1999); the one of Cacciari and Glucksberg (1991), the one of Sag
et al. (2002), to list just a few. However, for the sake of this book, I adopt
the clear-cut bipolar taxonomy of idioms, offered by Nunberg et al.
(1994), and followed by Harwood et al. (2016); thus, there are two types
of idioms, i.e. (i) idiomatically combining expressions (e.g. pull strings “to
use connections”), and (ii) idiomatic phrases (e.g. kick the bucket “to
die”). Idiomatically combining expressions (ICEs) have meanings, even
conventional ones, distributed among their constituent elements. In
idiomatic phrases (IdPs), instead, their components do not contribute to
the overall idiomatic meaning, but form a whole unit which is mapped
onto the figurative interpretation. In short, idiomatically combining
expressions include idioms whose overall idiomatic interpretation is
derivable (normally or abnormally, literally or figuratively), and they are
generally known as decomposable/compositional or analysable. Idiomatic
phrases, in turn, comprise idioms, whose overall idiomatic interpretation
is not derived from the constituent parts, and they are referred to as non-
decomposable / non-compositional, frozen, opaque, or unanalysable.
All in all, the heterogeneity of idiomatic expressions and a huge
diversity of definitions offered for an idiom imply the necessity of some
rules and theories that would both explain and categorise at least some
notions and irregularities related to idioms. This task will be undertaken in
the subsequent chapters.
1.6 Concluding remarks

This chapter has opened with a detailed presentation of several definitions
of an idiom, taken first from dictionaries and encyclopaedias as a starting
point. These sources of the definitions of an idiom have provided the first
clues as to the perspectives from which language scholars might have
looked at the notion of an idiom. The definitions put forward by linguists
and scientists have also been analysed. Although the term idiom is
frequently used in the literature, the substance of what it refers to varies.
Just as with defining and classifying formulaic language in general, it
appears impossible to reach agreement as regards the definition of an
idiom. Despite all potential arrays of idiom diversity, some general
introductory principles that most of the definitions share have been
outlined: (i) an idiom is recognized as an expression that contains more
than one word, and whose meaning is different from the sum of the literal
meanings of its components; (ii) some subset of idioms has a fixed word
order, which implies they have a restricted set of variants and should be
treated as entities since the link between their form and meaning has not
been recognized; (iii) recently, with a great input of psycholinguistic
studies (cf. Fernando 1996; and Moon 1998; among others), a great
number of idioms far from being dead or frozen has been found, but
instead, they may be marked with possible alterations, metaphoricity and
perceptible origins of their meanings.
Besides, some crucial characteristics of idioms have been established.
Firstly, it is idiomatic diversity, which results in various taxonomies of
idioms, classified in multiple ways, based on idiom semantics, syntax, and
function, e.g. the taxonomy of Makkei (1972), the one of Nunberg (1978)
and of his followers (Gibbs and Nayak 1989: 104; Titone and Connine
1999), Cacciari and Glucksberg (1991), Nunberg et al. (1994), the
taxonomy of Sag et al. (2002), and the classification of Yoshikawa (2008),
to name just a few. Yet, the kind of typology chosen for the purpose of this
book is the bipolar classification offered by Nunberg et al. (1994), who
divide idioms into (i) idiomatically combining expressions (e.g. pull
strings “to use connections”), and (ii) idiomatic phrases (e.g. kick the
bucket “to die”).
Additionally, the feature that most idiomatic phrases share, even
though the boundaries are sometimes overextended, is the metaphorical or
figurative nature (Mäntylä 2004: 28-29). What is more, the ambiguity of
many idiomatic expressions has been pointed out, as some idioms have
one interpretation (the literal meaning) derived from the meanings of the
words involved and/or the other–the idiomatic meaning. Subsequently,
some attention has also been paid to the further characteristics of idioms,
i.e., idiom analysability, the issue of non-compositionality, and idiom
semantic decomposition; all of which refer to the extent to which idiom
constituent parts contribute to the idiom overall interpretation (cf. Nunberg
1978; Cacciari and Tabossi 1988; Glucksberg 1991; and Gibbs 1994;
among others). As the third characteristics of idioms, the fixedness of form
and internal structure of idioms have been examined. Jackendoff’s (1997)
approach to idioms as linguistic units with an internal linguistic structure
constrained by syntax, semantics, morphology, and phonology, has shed
light on idiom variability. Fraser’s (1970) six-point Frozenness Hierarchy
places idioms on the scale, grouping them from totally frozen forms that
permit no grammatical or lexical changes to idioms that tolerate
unrestricted variation. Undoubtedly, both the syntactic productivity and
the lexical creativity of idioms are matters of degree, depending on the
idiom compositional properties. Moreover, the literalness of idioms, their
72 Chapter One
familiarity and predictability, investigated with the focus put on the role of
context, well-formedness of idioms, and the level of their formality, have
been discussed. Furthermore, some space has been devoted to the models
of idiom representation and processing most commonly cited in the
literature; namely, the non-compositional models (e.g. Swinney and Cutler
1979; Gibbs 1980), the compositional hypotheses (e.g. Cacciari and
Tabossi 1988; Gibbs, Nayak, and Cutting 1989), and the hybrid
approaches with the Model of Dual Idiom Representation (e.g. Titone and
Connine 1999).
In brief, in Chapter One some main properties and models of idioms
representation and processing have been provided as an essential
background to understand the syntactic and semantic variability of idioms,
which is to be studied in Chapter Four of the book. Moreover, the working
definition of an idiom as a multi-word phrase and phraseological unit has
been established, to be adopted for the purposes of this book. Gläser’s
(1998) definition of “phraseological units” has been chosen as the one
comprising all the possible notions and characteristics that should be
encompassed by the term under consideration.
Since the book is to focus on idioms referring to psychological states in
English, it is important to learn first the basic syntactic, semantic, and
aspectual characteristics of psychological verbs. These characteristics and
the structure of psychological verbs, as well as an overview of syntactic
approaches to psych-verbs, are the main focus of the subsequent Chapter
Two.
CHAPTER TWO
SYNTACTIC AND SEMANTIC

CHARACTERISTICS OF PSYCHOLOGICAL
VERBS
2.1 Introduction
The name psych-verbs (psychological verbs) is commonly assigned to

verbs denoting mental or emotional states, such as fear, love, worry,
frighten, or surprise. Such verbs select a participant / an individual who
experiences an emotional or mental state, usually referred to as an
Experiencer, and a non-Experiencer argument, sometimes called stimulus,
trigger of emotion, causer or target/subject matter, or simply subsumed
under the label of “theme” (Landau 2010: 5). What distinguishes psych-
verbs from other predicates is the fact that (i) at least one of their
arguments refers to a sentient, usually human, Experiencer, who is able to
feel the emotion that the verb describes (Grafmiller 2013: 10), and that (ii)
psych-verbs display unique syntactic properties, so-called “psych
effects.”1 Indeed, for several decades psychological predicates have been a
subject of debate in theoretical syntax. Members of this class have become
a fertile ground for examining both the verb meaning, and the connection
the meaning has with grammatical structure (Belletti and Rizzi 1988; Biaáy
2005; Evans 2009; Verhoeven 2010, among others). Even though there is
little consensus regarding mapping between particular elements of verb
meaning and the syntactic structure of verbs, some components of
meaning, i.e. stativity, agentivity and causativity, have been extensively
1
Arad (1998) makes a suggestion that psych-verbs are neither lexically nor
syntactically distinctive, since they share the same structure as their main predicate
(locative, dative etc.). However, their unique syntactic properties rely more on
their stativity rather than on their being assigned a specific structure (cf. Grafmiller
2013).
74 Chapter Two
discussed in the literature on the example of psych-verb behaviour

(Grafmiller 2013: 11).
This chapter of the book is not to offer a comprehensive analysis of
psychological predicates. Instead, the aims of Chapter Two are: (i) to
present the crucial syntactic, semantic, and aspectual characteristics of
psychological verbs, and (ii) to offer a brief overview of the syntactic
approaches to psych-verbs, available in the literature. Due to space
limitations, only the most influential proposals concerning psych-
predicates are to be focused on.
Chapter Two comprises five sections, and it is organised as follows:
section 2.2 opens up with a working definition of psych-verbs, adopted for
the sake of the book, followed by the syntactic classification of
psychological predicates offered in the literature. In section 2.3 the lexical-
semantic representation of a verb is discussed. Section 2.4 presents the
aspectual typology of class I-III psych-verbs at the Lexicon-Syntax
Interface, and the syntactic tests to make the distinction between stative,
eventive non-agentive, and eventive agentive readings of class II Oject
Experiencer (OE) psych-verbs clear. Section 2.5 is an overview of the
main syntactic analyses of psych-verbs, starting with Belletti and Rizzi’s
(1988) unaccusative approach to OE psych-verbs, through Landau’s
(2005, 2010) locative approach, and Fábregas and Marín‘s (2015) layer
theory, up to Grafmiller’s (2013) recent account of psych-verbs. While
presenting the approaches to psychological predicates, a special focus is
laid on OE psych-verbs, which, in contrast to Subject Experiencer (SE)
psych-verbs, are syntactically more complex, exhibiting a number of
seemingly conflicting properties (cf. Landau 2010: 5). Besides, an outline
of the crucial syntactic properties of English OE psych-verbs is offered,
which is relevant for an analysis undertaken in the subsequent chapters of
the book. Finally, the chapter closes with a summary, provided in section
2.6.
2.2 Syntactic typology of psych-verbs

The working definition of psych-verbs adopted for the purpose of this
analysis is the one provided by Landau (2010: 4n2), according to whom
psychological verbs carry “psychological entailments involving an
individual being in a certain mental state.” Thus, frighten is a psych verb
in (2.1a), since it means that Nina is in a certain mental state (i.e. fright)
caused by the science fiction film; whereas visit, in (2.1b), is not a psych
Syntactic and Semantic Characteristics of Psychological Verbs 75
verb, on account of the fact that the sentence involves no state of mind
either of Charles or of Nina.2
(2.1) a. This science fiction film frightens Nina.

b. Charles visits Nina.
Psych verbs can be divided into two classes, depending on the syntactic
position of the Experiencer (e.g. Croft 1993; Pesetsky 1995; and Iwata
1995; among others), such as Subject Experiencer (SE) verbs and Object
Experiencer (OE) verbs, as illustrated in (2.2a-b):
(2.2)
As can be seen in (2.2a-b), psych-verbs show different syntactic

realizations of the Experiencer argument, which can surface either as a
subject or as an object.
Moreover, when taking into account the relationship between the
lexical properties of psych-verbs and their syntactic structure, i.e. within
the lexicon-syntax interface, Belletti and Rizzi’s (1988: 291-292)
classification of psych-verbs is frequently referred to in the literature.
Belletti and Rizzi’s (1988) tripartite taxonomy, originally created for
Italian psych-verbs, is displayed in (2.3):
(2.3) Belletti and Rizzi’s (1988) classification of psych-verbs:
a. Class I: The temere class

(Nominative Experiencer, accusative Theme)
Gianni teme questo.
Gianni fears this.
2
According to Klein and Kutscher (2005: 2), from the semantic point of view,
psych-verbs can be classified into verbs denoting emotions (love, frighten, etc.),
perception verbs (see, taste, etc.), cognitive verbs (think, assume, muse, etc.), and
evaluating verbs (respect, appreciate, etc.). However, some of the verbs listed here
do not satisfy Landau’s (2010) definition of psych-verbs.
76 Chapter Two
b. Class II: The preoccupare class

(Nominative Theme, accusative Experiencer)
Questo preoccupa Gianni.
This worries Gianni.
c. Class III: The piacere class

(Nominative Theme, dative Experiencer)
(i) A Gianni piace questo.
To Gianni pleases this.
(ii) Questo piace a Gianni.
This pleases to Gianni.
(Belletti and Rizzi 1988: 291-292)
Class I comprises SE psych-verbs, illustrated in (2.2a), (2.3a) and (2.4),

for such verbs as, e.g. hate, love, or adore.
(2.4) Paul hates / detests / loves classical music.

(Experiencer as the subject)
SE psych-verbs feature a nominative Experiencer and an accusative

Theme. These verbs are generally regarded to be similar to other transitive
stative verbs, e.g. know.
Class II and III comprise OE psych-verbs, which, on account of their
specific psych-properties (cf. section 2.5.2 for more discussion), have
received a lot of attention in the literature so far. Thus, in Class II, as in
(2.2b) and (2.3b), with verbs like frighten, worry or distress, the subject is
associated with the role of the Theme, and the Experiencer appears as an
accusative object. Class III, in turn, illustrated for Italian in (2.3c), and
represented in English by verbs like appeal to or matter to, includes
psych-verbs with a nominative Theme in the subject position, and a dative
Experiencer, occupying the object positions, and both permutations
(2.3c)(i) and (2.3c)(ii) are acceptable. Accordingly, cross-linguistically
psych-verbs are classified in accordance with the typology offered by
Belletti and Rizzi (1988), and this taxonomy is adopted for the sake of this
book.
Since the Experiencer argument can be realized in psychological
predicates either as a subject or as an object, psych-verbs pose a problem
for linking, viz. the mapping of thematic roles to arguments in the syntax3
3
The concept of semantic roles relates to the notion of thematic relations, i.e. the
relations which are semantic in nature. The thematic role is, thus, a semantic
(cf. Belletti and Rizzi 1988; Grimshaw 1990; Pesetsky 1995; Arad 1998;
2000; Anagnostopoulou 1999; Primus 1999; Pylkkänen 2000; Verhoeven
2010; and Landau 2010; among others). This challenge that psych-verbs
pose is to be discussed in sections 2.3 and 2.5 in more detail.
In a nutshell, psych-verbs are expressed in different ways across
languages. In languages like German, Italian, or English, psych-verbs
permit word orders or binding relations that are impossible for non-psych-
verbs. A comprehensive investigation of psych-verbs in and across a
variety of world languages is provided by Landau (2010). In addition,
Bennis (2000), and Drijkoningen (2000) offer some detailed analysis of
psych-verbs in Dutch. Pesetsky (1987) analyses psych-predicates in
English; Anagnostopoulou (1999, 2008) in Greek; while Ruwet (1972,
1993), Legendre (1989, 1993), and Bouchard (1992, 1995) in French.
Moreover, Arad (1998), Reinhart (2002), and Anagnostopolou (2008)
discuss psych-verbs in Hebrew; Klein and Kutscher (2005), and Temme
(2014) in German; whereas Belletti and Rizzi (1988), Cresti (1990), and
Arad (1998) in Italian. Besides, Biaáy (2005) and Rozwadowska (1992,
2012, 2014) work on Polish psych-verbs; Franco (1990), and Jiménez-
Fernández (2014) analyse Spanish psych-verbs; while Kim and Larson
(1989) study Korean psych-predicates. Indeed, psych-verbs have become a
much debated issue in the literature.
2.3 The lexical-semantic representation of a verb

The main goal of this section of the chapter is to introduce linking rules in
the light of the Lexicon-Syntax Interface. While analysing the lexical
semantics of a verb and the syntactic structures a verb can occur in, at least
three different levels of representation of a verb can be distinguished, as
illustrated in Figure 2-1 below:
function of an argument in a given sentence, such as the role of an Agent, Theme,

or Instrument. For Chomsky (1981), theta roles are semantic roles that are assigned
in syntax. In sentence John likes hot coffee, we have two obligatory arguments, i.e.
John is an external argument, and coffee is an internal argument. These arguments
have thematic roles of an Agent / Experiencer, and Theme / Object, respectively
(cf. Kiparsky 1987; Baker 1989; and O’Grady 1998; among others).
78 Chapter Two
Figure 2-1. The lexical-semantic representation of a verb

(Alexiadou et al. 2004: 11)
As shown in Figure 2-1, a verb can be represented at the following levels:

(i) a lexical-semantic representation (also called a lexical conceptual
structure (LCS)), (ii) a lexical-syntactic representation, commonly known
as a predicate-argument structure or an argument structure (AS), and (iii) a
syntactic structure representation.
The lexical-semantic representation of a predicate, is a “deep”
semantic description unique for a single predicate, which divides a verb
meaning into variable structures and meta-predicates (such as cause, be,
etc.). This semantic configuration is mapped onto an argument structure
(AS), which is responsible for the unaccusative-unergative distinction, and
it determines the number and syntactic roles of arguments a verb entails,
e.g. by assigning external and internal theta roles (Williams 1981;
Stalmaszczyk 1996; and Belletti and Rizzi 1988). The argument-structure
representation may be the same for different verbs. Although essentially
different, the lexical conceptual structure and the argument structure are
part of the lexical representation of a predicate and thus part of the lexicon,
which is distinct from syntax. In other words, lexical semantic properties
are directly reflected in the argument structure, which is linked with
syntax. This assumption is adopted for the sake of this book, and the study
of idioms, since idioms cannot be analysed only in terms of syntactic rules
(cf. Chapter Four). Besides, as maintained by van Valin (1990) and Levin
and Rappaport Hovav (1995), lexical semantic concepts are, theoretically,
accessible and directly related to syntax.4
4
Interestingly, Tenny (1987), Rappaport Hovav and Levin (1988), and Grimshaw
(1990), on the other hand, hold the view that there seems to be no direct relation
between syntax and the lexical semantics of predicates (LCS) but only between
syntax and the AS.
Notably, the correspondence between semantic properties and syntactic

behaviour of verbs, between the LCS and the AS, and between the AS and
the syntax is addressed, first, in Perlmutter and Postal’s (1984) Universal
Alignment Hypothesis (henceforth, UAH), reproduced in (2.5a). Second,
the assumption concerning the semantics-syntax connection underlies the
Uniformity of Theta-Assignment Hypothesis (henceforth, UTAH), put
forward by Baker (1988), as formulated in (2.5b).
(2.5) a. Universal Alignment Hypothesis (UAH):

There exist principles of universal grammar which predict the
initial relation [= syntactic encoding], borne by each nominal
in a given clause from the meaning of the clause
(Perlmutter and Postal 1984: 97)
b. Uniformity of Theta Assignment Hypothesis (UTAH)

Identical thematic relationships between items are represented
by identical structural relationships between those items at the
level of D-Structure
(Baker 1988: 46)
The UTAH, formulated in (2.5b), states that a lexical item bearing a

certain thematic role (e.g. Theme), will always be mapped onto a
particular, fixed, structural position (e.g. direct object).
It can be observed that according to the UTAH, the correlation

between the syntax and the lexicon seems to be determined, since once we
have the set of roles the verb assigns, the syntactic structure is fixed by
them. Besides, as noted by Pesetsky (1995: 12), the UAH, stated in (2.5a),
is weaker than the UTAH, reproduced in (2.5b), in that it does not entail
identical syntactic linking patterns in cases of semantic identity, but only
involves predictable linking patterns.
By capturing a correlation between the lexicon and the syntax, the
UTAH makes it possible to map two distinct roles onto the same position
(e.g. Agent, Causer and Experiencer may all appear in the subject
position), as noted by Arad (1996). However, the UTAH is violated if an
explicit role appears in two different syntactic positions. This occurs in the
case of (i) the dative alternation (the alternation exhibited by verbs such as
give, whose Goal argument may either be case-marked by the verb or by
the preposition to); (ii) the locative alternation, i.e. load-type verbs (e.g.
spray, hammer, load), which allow either their Goal argument or their
Theme argument to occupy the direct object position; (iii) experiencer
80 Chapter Two
verbs, i.e. pairs such as fear/frighten, in which the argument bearing the
Experiencer role is mapped, firstly, onto the subject position and secondly,
onto the object position; (iv) variable behaviour verbs, i.e. intransitive
verbs which may appear as both unaccusatives and unergatives, with the
existence of a locative PP (in Dutch and Italian) or without it (in Hebrew).
In short, psych-verbs are among these items which defy the UTAH.
Accordingly, to save Baker’s (1988) hypothesis, alternative views on
psychological predicates have been proposed in the literature, as discussed
in section 2.5.
Furthermore, for Grimshaw (1990), the lexical-syntactic representa-
tion, viz. the argument structure (AS) of a verb is related to the event
structure of a verb. For instance, an accomplishment verb like x constructs
y is analysed as an activity in which x engages in construction plus a
resulting state in which existence is predicated of y (Grimshaw 1990: 26),
as represented in (2.6):
(2.6)
(Grimshaw 1990: 26)
As assumed by Grimshaw (1990), the AS contains an aspectual dimension

since argument relations are determined by the thematic properties of the
predicate (i.e. the thematic hierarchy) and by the event structure of the
predicate (i.e. its aspectual properties). In short, if a predicate lacks an
event structure, it also lacks the AS and takes no grammatical arguments at
all. With this in mind, the subsequent section (section 2.4) of the chapter is
devoted to the event structure and aspectual classification of psych-verbs.
2.4 Aspectual classification of psych-verbs

Even though there have not been very many studies of the lexical aspect of
psychological predicates, in comparison with the research done into the
argument structure of psych-verbs, certain conclusions regarding the
lexical aspect of the different classes of psych-verbs have been reached so
far. Indeed, psych-verbs are frequently ambiguous between states and
events (non-states), exhibiting subtle aspectual distinctions (e.g. Grimshaw
1990; van Voorst 1992; Tenny 1992; Marín and McNally 2011; and
Rozwadowska 2012). To make the discussion clear, first, some

fundamental semantic and aspectual concepts have to be introduced.
However, a more detailed analysis of the aspectual properties of verbs lies
outside the scope of this book.
2.4.1 Basic event categories

On the basis of Ryle’s (1949) and Kenny’s (1963) assumptions, Vendler
(1957, 1967) reconsidered aspectual properties of verbs, and first proposed
a four-way classification of events, dividing them into states, activities,
accomplishments, and achievements. Reviewing Vendler’s typology of
events, Dowty (1979) offered the following examples of the four event
types, reproduced in (2.7).
(2.7) States Activities Accomplishments Achievements

know run paint a picture recognize
believe walk make a chair find
have swim deliver a sermon lose
desire push a cart draw a circle reach
love drive a car recover from illness die
(Dowty 1979: 54)
To classify events, Vendler (1967) uses the aspectual properties of

verbs referring to lexical aspect, called Aktionsart. Under Vendler’s
classification, activities and states both denote situations that are
inherently temporally unbounded (atelic); states depict static conditions,
whereas activities denote on-going dynamic aspects. Activities and
accomplishments differ from achievements and states in that the former
comply well with continuous and progressive aspects, while both
accomplishments and achievements express a change of state, and are
henceforth temporally bounded (telic). Activities and accomplishments
extend over a period of time, but accomplishments are punctual. In other
words, accomplishments approach an endpoint gradually (as in paint a
picture or recover from illness), whereas achievements take place
immediately (as in lose, recognize or find) (Vendler 1967; cf. Dowty 1979;
and PiĔón 1997).
Accordingly, in terms of telicity, activities resemble states, whereas
accomplishments are similar to achievements. However, it has also been
noticed that states can be grouped with achievements and activities with
accomplishments, since the former pair lacks the progressive aspect, while
the latter pair allows it (cf. Lakoff 1966, Shi 1988). To make this
distinction between the different types of events clear, Smith (1991: 30),
82 Chapter Two
and van Valin and LaPolla (1997: 91-102), among others, decompose
Vendler’s classes in terms of independent features, as reproduced in (2.8).
(2.8) a. States: [–telic, –durative, –dynamic]

b. Activities: [–telic, +durative, +dynamic]
c. Achievements: [+telic, –durative, +dynamic]
d. Accomplishments: [+telic, +durative, +dynamic]
(cf. van Valin and LaPolla 1997: 91-102)
The properties of Vendler’s (1967) event types can be captured by the

abovementioned features. Nonetheless, the linguistic status of
achievements has been questioned by many linguists so far. They are
characterised as punctual and telic, but the question whether duration is a
property inherent to the verb or not, remains unclear (Lin 2004: 19-20).5 In
short, achievements are associated with the general term “change of state”
verbs, which is assumed to be directly encoded in the meaning of a verb.
Undoubtedly, the type of event makes a big difference for the lexicon-
syntax interface. However, following the literature, Bach’s (1981) term
“eventuality” is frequently adopted to cover all four event types. Bach
(1986) offers the following classification of predicates (cf. Carlson 1981),
which is reproduced in (2.9) below.
5
For example, for Pustejovsky (1991) accomplishments and achievements fall into
the so-called “transitions,” for Tenny (1987: 20) achievements and
accomplishments are dissimilar in terms of the duration of an event, while Verkuyl
(1993: 48) concludes that the distinction between achievements and
accomplishments, viz. the duration of events, is a matter of real-world knowledge.
(2.9) Bach’s (1986) classification of predicates:6
(Bach 1986: 6)
Besides, Bach (1986) postulates the so-called psych-eventualities,

adopted by Rozwadowska (2012),7 in which agentivity, closely related to
dynamic events, is expected to disappear.8
Undoubtedly, it is Vendler’s (1967) classification of event types that
has served as a cornerstone for other lexical semantic representations and
6
Typical examples are: (a) sit, stand, lie + LOC (b) be drunk, be in New York, own
x, love x, resemble x (c) walk, push a cart, be mean (Agentive) (d) build x, walk to
Boston (e) recognize, notice, flash once (f) die, reach the top.
7
Rozwadowska (2012: 535) uses the term “psych-eventualities,” to identify mental
and emotional states and their beginnings. Developing further the ideas of Arad
(1998) and Pylkkänen (1997), she argues that psych-eventualities are what
Rappaport Hovav and Levin (1988) treat as temporally dependent co-existing sub-
events with one event variable, which is identified through one structure
participant only.
8
Cf. Piñón (1997), who distinguishes between two fundamentally different types
of eventualities: (i) those with some duration, which he calls happenings; and (ii)
boundary happenings, which are the initial or final boundaries of some happening.
Boundary happenings are absolutely instantaneous. Piñón (1997) argues that
predicates such as begin or arrive denote boundary happenings, whereas a change
of state predicate like cool does not.
84 Chapter Two
theories of verbal argument structure. Among them there is Dowty’s

(1979) seminal work, in which he makes an attempt to decompose states,
activities, accomplishments, and achievements in terms of the primitives
DO, CAUSE, and BECOME. This is replicated in the form of patterns in
(2.10), and illustrated by means of exemplary English sentences in (2.11).
(2.10) a. state: ʌn (Į1,…, Įn)

b. activity: DO(Į1,[ ʌn (Į1,…, Įn)])
c. achievement: BECOME[ʌn (Į1,…, Įn)]
d. accomplishment:
[[ DO (Į1,[ʌn (Į1,…, Įn)])] CAUSE [ BECOME [ ʌn (Į1,…, Įn)]]]
(2.11) a. He sweeps the floor clean.

[[DO (he, sweeps(the floor))] CAUSE [BECOME[clean(the floor)]]]
b. John walks.
[DO(John, walk)]
(Dowty 1979: 123-124)
Importantly, Dowty (1979) divides causative structures into two

subevents: a causing subevent and a result subevent. Many linguists have
adopted this division in their analyses. The representation of the resultative
sentence (2.11a), consists of the causing subevent “he sweeps the floor”
and the result subevent “the floor is clean.” In turn, unergative verbs, as in
(2.11b), are composed of a single subevent with the primitive DO.
Two decades later, Rappaport Hovav and Levin (1998) also defined a
basic inventory of event building blocks in terms of Vendler’s (1967)
event types, as reproduced in (2.12).
(2.12) a. [ x ACT<MANNER> ] (activity)

b. [ x <STATE> ] (state)
c. [ BECOME [ x <STATE> ] ] (achievement)
d. [ x CAUSE [ BECOME [ y <STATE> ] ] ]
(accomplishment)
e. [[x ACT<MANNER>] CAUSE [BECOME [y<STATE> ]]]
(accomplishment)
(Rappaport Hovav and Levin 1998: 108)
On the basis of (2.12), the meaning of a verb is expected to comprise

an association between a constant and an event template from the
inventory given above. Constants are defined as open-class items
originating from a fixed ontology (e.g. manner, instrument, state, etc.),
represented within the angle brackets of the event template.9 Rappaport

Hovav and Levin (1998) claim that in this theory verbs directly project
(encode, or lexicalize) complex event structures.
What is more, Dowty (1979) argues that stative predicates are the
smallest and simplest building-blocks of event structure. Besides, Dowty
(1979: 180) introduces a fundamental distinction within stative verbs:
verbs of position (e.g. sit and lie) which take the progressive form (X is
sitting) in English, while other statives (e.g. know and love) do not have
such a form. Bach (1986) refers to this idea and distinguishes two types of
statives, dynamic (e.g. sit, stand, and lie in combination with a location
modifier) and static statives (e.g. be drunk, be in New York, own, love, and
resemble).
The split within the class of statives is based, to a great extent, on the
ideas of Davidson (1967)10, and was developed further in Maienborn
(2003, 2005, 2007)11 and Rothmayr (2009). Examining statives, Rothmayr
9
Besides, for Rappaport Hovav and Levin (1998), each constant is associated with
a name (i.e. a phonological string). A set of “canonical realization rules” governs
the compatibility of different constant types with different event types (i-vi):
i. manner ĺ [ x ACT<MANNER> ]
(e.g., jog, run, creak, whistle, etc.)
ii. instrument ĺ [ x ACT<INSTRUMENT> ]
(e.g., brush, hammer, saw, shovel, etc.)
iii. placeable object ĺ [ x CAUSE [ BECOME [ x WITH <THING> ]
]]
(e.g., butter, oil, paper, tile, wax, etc.)
iv. place ĺ [ x CAUSE [ BECOME [ x <PLACE> ] ] ]
(e.g., bag, box, cage, crate, garage, pocket, etc.)
v. internally caused state ĺ [ x <STATE> ] (state)
(e.g., bloom, blossom, decay, °ower, rot, rust, sprout, etc.)
vi. externally caused state ĺ [ [ x ACT ] CAUSE [ BECOME [ y
<STATE> ] ] ]
(e.g., break, dry, harden, melt, open, etc.)
(Rappaport Hovav and Levin 1998: 109)
10
The Davidsonian (1967) account of adverbial modification, initially used with
action verbs, has been used to separate the ontological properties of situation types.
Davidson argues that adverbials are attached to an event argument that must
therefore be present in the structure of (eventive) verbs. Since statives do not allow
for such adverbials, they are taken to lack this argument, referred to as the
Davidsonian argument.
11
Maienborn (2003, 2005, 2007) offers a theory of nondynamic expressions,
which distinguishes between state verbs (sleep, sit, stand, lie and wait), consistent
with the criteria for the so-called Davidsonian eventualities, and stative verbs
(know, weigh, own and resemble), which refer to the Kimian (1969) state criteria.
86 Chapter Two
(2009) observes that the nature of stative verbs has not yet been
completely explained, but at least four suppositions about statives can be
made. First, stative predicates are recognised as the smallest and simplest
building blocks of event structure. Second, stative verbs have a less
complex structure compared to eventive verbs. Third, stative verbs are
very similar to copular constructions. Finally, stative verbs contain only a
single (structural) argument, which is the holder of the state (Rothmayr
2009: 39). Taking the distinction between Kimian (1969) and Davidsonian
(1967) statives (cf. Maienborn 2003) as a preliminary assumption, the
results of Rothmayr’s (2009) study of different types of stative verbs
reveal that these predicates cannot be seen as the smallest building-blocks
of event structure. Indeed, it is possible for a verb to express both a state
and an event (either by referring to an intentional activity by an agent, or
by expressing a change of state). Therefore, the event structure must be a
property of the lexical-semantic structure of stative predicates in a verb.
Finally, Rothmayr (2009) offers a two-group division of stative verbs into:
i) verbs that allow stative reading only (SE verbs, such as love, and verbs
of position); and ii) verbs that display systematic stative-eventive
ambiguity (e.g. verbs with both a causative and stative reading, such as
obstruct, dispositional verbs, such as help, verbs with a modal operator
and a feature of action, such as threaten, and verbs of body posture).
According to Rothmayr (2009), the stative/eventive ambiguity can be
explained by a cause-operator in a verb’s semantic structure that relates
two sub-eventualities to one another; the stative interpretation arises when
both sub-eventualities are stative (Rothmayr 2009: 80-82).
Unquestionably, Vendler’s (1967) event classification may be used as
a guide for lexical semantic representations, but it may not be the final
determining factor of event-based theories of argument structure. Apart
from the work of Dowty (1979), Rappaport Hovav and Levin (1998), and
Rothmayr (2009), many theories of lexical semantic representations of a
verb have been developed so far, e.g. Carter (1976), Jackendoff (1983),
Pinker (1989), Parsons (1990), Pustejovsky (1991), Croft (1998), and
Rothstein (2004, 2008), among many others. In general, these alternative
frameworks concern lexical semantic representations of a verb that
decompose events into more primitive predicates.
The latter group comprises copular constructions, regardless of whether they are
stage-level predicates (denoting a temporary property) or individual-level
predicates (carrying a more or less constant property). Kimian states, to be precise,
reflect Kim’s (1969) notion of temporally bound property exemplifications–they
are not discernible, and they cannot be modified by event-related adverbials
(Maienborn 2007), while Davidsonian states can.
2.4.2 Aspectual classes of psych-verbs

The Vendler-Dowty classes of Aktionsart, especially the Aktionsart of OE
psych-verbs in English, have been a subject of debate in the literature.
Indeed, across different languages psychological predicates are ambiguous
between agentive, eventive and stative interpretations (e.g. Arad 1998,
1999; Marín and McNally 2005: 212; 2011: 468; Landau 2010: 129; and
Alexiadou and Iordăchioaia 2014; among others). For instance, psych-
verbs are assumed by van Voorst (1992) to be achievements (which are
telic, but cf. Landau 2010: 150), while Filip (1996) states that causative
psych-verbs are atelic.
2.4.2.1 The aspectual status of Subject-Experiencer psych-verbs
In the literature, SE psychological predicates are uniformly defined as

stative. The fact that they occur felicitously in the simple present tense
without a habitual reading, as in (2.13a), proves their stativity. SE psych-
verbs make an implication of temporal persistence into the past and future,
similarly to individual-level predicates (Carlson 1977; and Condoravdi
1992). Consequently, SE psych-verbs, treated as individual-level verbs,
are assigned a stative reading since they do not appear felicitously, in the
usual context, with temporal modifiers such as yesterday (Spanish ayer),
as illustrated in (2.13b) (cf. Marín 2001 for Spanish data).
(2.13) a. {Odia / Aborrece / Teme} su libertad.

hates / loathes / fears his freedom
He {hates / loathes / fears} his freedom.
b. ??Ayer {odió / aborreció / temió} su repentina libertad.

yesterday hated3SG / loathed3SG / feared3SG his sudden
freedom
*Yesterday he {hated / loathed / feared} his sudden freedom.
(Fábregas et al. 2012: 164)
Thus, SE psychological predicates are generally assumed to denote states

(Grimshaw 1990; and Pustejovsky 1991; among others).12
12
An interested reader is referred to Alexiadou and Iordăchioaia (2014), who note
some ambiguity among SE psych-verbs in Romanian (between eventive and stative
readings), and in Greek (either only eventive or only stative).
88 Chapter Two
2.4.2.2 Aspectual classes of Object-Experiencer psych-verbs
All class III OE psych-verbs are believed cross-linguistically to be stative.

These predicates are never used agentively; hence, they do not occur with
agentive adverbs on purpose / intentionally, as exemplified in (2.14a-b)
below (cf. (2.3c)).
(2.14) The stative reading of class III psych-verbs:

a. The issue of starving children appeals to Maria (*on purpose /
*intentionally)
b. These children matter to Maria (*on purpose / *intentionally)
Maria is an Experiencer, placed in the Object position, while “the issue

of starving children” / “these children” represent the Stimulus / Theme,
located in the Subject position.
In comparison with SE and class III OE psych-verbs, there is much less
agreement regarding the aspectual value to be assigned to OE psych-verbs
of the frighten type (class II). To be precise, English frighten verbs are
sometimes treated as (telic) achievement predicates (van Voorst 1992), in
a way analogous to accomplishments (Tenny 1994). Contrary to those
views, Filip (1996) argues that class II psych-verbs are not telic, and both
Pylkkänen (2000) and Arad (1998) have claimed that, at least, in some
interpretations certain members of the frighten class in Finnish
(Pylkännen) and English and Romance (Arad) are stative.
For the sake of this book, Arad’s (1998, 1999) aspectual typology of
class II OE psych-verbs is adopted, with her claim that most of these verbs
are ambiguous between the three readings, i.e. (i) eventive agentive
(regular transitives); (ii) eventive non-agentive (stative / transitive); and
(iii) stative. The variety of the readings, available even for one psych-verb,
such as frighten, is exemplified in (2.15).
(2.15) The three readings of class II psych-verbs:

a. Nina frightened Laura deliberately (intentionally / on purpose)
/ to make her go away.
(eventive agentive reading)
b. Nina frightened Laura unintentionally / accidentally.

(eventive non-agentive reading)
The explosion / the thunderstorm frightened Laura.
(eventive non-agentive reading)
c. Dogs frighten Laura. (stative reading)
(Arad 1998: 3,6)
According to Arad’s (1998) event-based approach, in the eventive

agentive reading in (2.15a), we have an intentional Agent (Nina) that
brings about a change of state in Laura, the Experiencer, who gets
frightened. In Arad’s (1998) analysis, the eventive non-agentive reading,
exemplified in (2.15b), emerges when someone (Nina) or something (the
explosion / the thunderstorm) causes some change of mental state in the
Experiencer (Laura) unintentionally. Thus, on this reading there is a
change of state in the Experiencer, but no intentional agent. Thus, on the
non-stative readings (agentive or eventive ones) the Agent/Causer “have
done their job” as soon as the change of state is achieved (Arad 1998: 5).
Researchers further disagree as regards the specific aspectual value of
eventive OE verbs and its potential reference to agentivity (achievements,
e.g. van Voorst 1992, vs. achievements or accomplishments, depending on
agentivity, e.g. Landau 2010). More recently, Grafmiller (2013) has
advocated for OE verbs, as well. According to him, any of OE verbs can
be used to describe a dynamic event with an agent, and no systematic
aspectual classification can be made for this type of psych-verbs.
The stative reading, as in (2.15c), is the generally assumed psych
reading, with neither an Agent nor any change of mental state in the
object. Instead, as argued by Pylkkänen (1997), it involves perception of
some stimulus (the subject) by the Experiencer (the object). This
perception activates some mental state in the Experiencer. There is, thus,
“triggering of a state, but no change of state” (Arad 1998: 6), as shown in
(2.16), where Nina is the Experiencer.
(2.16) The stative reading of class II OE psych-verbs:

a. John / John’s haircut annoys Nina.
b. John / John’s behaviour / nuclear war frightened Nina.
c. This problem concerned Nina.
d. Blood sausage disgusts Nina.
(Arad 1998: 4 (4))
What is more, Arad (1998: 4) distinguishes several characteristics

which make the stative reading different from the other two. First, there is
“no Agent” in the stative reading, viz. neither the activation of the mental
state by the stimulus nor the perception of the stimulus by the Experiencer
is under the control of the Agent. It is something inherent to the stimulus
that generates a particular mental state in the Experiencer. Consequently,
the Experiencer cannot control the mental state which the stimulus triggers
in it. Second, on the stative reading there is “no change of state” in the
Experiencer, as is the case, e.g. with the psych verbs concern or worry,
90 Chapter Two
which are treated as fundamentally stative. They entail no single point of

change of state, in which the Experiencer turns from “unconcerned” /
“unworried” into “concerned” / “worried.”13 In brief, according to Arad
(1998, 1999), the stative reading involves triggering of a mental state
which holds as long as the Experiencer perceives the Stimulus; the non-
stative interpretation, in turn, involves an agent which brings about a
change of state.14 Therefore, a key issue in recognising these aspectual
values is the role of the subject: while states do not allow agents, many OE
verbs can appear with both agent subjects (Nina frightened Laura
deliberately / to make her go away) and non-agent subjects (Nina
frightened Laura *deliberately / unintentionally / accidentally) (cf. Arad
1998: 3, 6; Cheung and Larson 2015: 166 (104)).
Furthermore, for Arad (1999), it is stativity that makes the class of
psych-verbs differ from prototypical transitive verbs. Thus, she concludes
that “the ‘psych’ category does not give us a cohesive syntactic-semantic
class” (Arad 1999: 15, cf. Sonnenhauser 2010).15 To conclude, Verhoeven
(2010), similarly to Rozwadowska (2005), notes that the availability of the
alternation between an agentive / stative and non-agentive / non-stative
reading is subject to typological variation.
To conclude, SE predicates may be taken to be stative. For some
researchers (cf. Arad 1999), SE verbs may refer to simple events (as
intransitive activities do), for other linguists, SE verbs as statives comprise
two subevents (cf. Biaáy 2005; and Rozwadowska 2012; among others).
Moreover, all class III OE psych-verbs are believed cross-linguistically to
be stative (unaccusative), as they are never used agentively (cf. (2.3c), and
e.g. The issue of starving children matters to Maria). The controversy
opens up for class II OE predicates, which are more complex and can have
either a stative or eventive reading.
13
Arad (1998) notes that both the stative reading and the non-stative reading are
causatives (as is evident by the causative morphology on OE verbs in Finnish,
Hebrew and Japanese). The type of causation is different in each case: for non-
stative readings it is an active causation, causing a change of state, the other is
stative causation, or triggering a concomitant state (Arad 1998: 6).
14
Cf. Pylkkänen (2000: 431-432) and the correlation between “affectedness” and
“eventiveness” in Parodi and Luján (2000).
15
Likewise, Pylkkänen (2000) identifies a class of causative psych-verbs in
Finnish, which represent temporary ‘stage-level’ states rather than permanent
“individual-level” states (e.g. inho-tta “disgust,” saali-tta “cause to pity,” sure-tta
“cause to be sad;” see also Marín and McNally 2005). Pylkkänen (2000)
recognizes as well another set of causative psych-verbs which are non-stative and
involve the inchoative morpheme (e.g. raivo-stu-tta “cause to become furious,”
kauhi-stu-tta “cause to become terrified,” viha-stu-tta “cause to become angry”).
2.4.2.3 Syntactic tests for stativity / non-stativity of class II OE psych-

verbs
Landau (2010), following Pesetsky (1995), claims that especially class II

OE psych verbs, are not aspectually identical. Some verbs (e.g. scare,
startle) seem to be restricted to a non-stative reading; others (e.g. amuse,
embarrass) remain neutral, i.e. aspectually ambiguous between eventive
and stative readings; whereas other psych-verbs, though they are
“relatively few” (e.g. concern, depress) are “strictly stative” (Landau
2010: 129). Landau’s (2010) assumption is followed by Cheung and
Larson (2015: 136-137), and Guidi (2011: ex. 37), among many others.
Guidi (2011) also observes that OE psych-verbs in Old English, where the
Experiencer was accusative, were aspectually ambiguous, just as they are
in Present-day English.16
Consequently, some tests to distinguish stative from non-stative class
II verbs are of much significance. The stative reading is achieved most
easily with bare plurals, as in (2.15c), and imperfective aspect, such as the
present tense, as in (2.16c) and (2.16a,d). The verb refers to a long-lived
state thanks to the simple present in which the predicate occurs. English
appears to show preference for simple present tense verbs to be interpreted
as generic statements (Carlson and Tanenhaus 1988). Stative Experiencer
verbs cannot be naturally used in the progressive form; that is why, “*This
problem is concerning Nina” is ungrammatical. Indeed, Landau (2010: 49)
mentions that in English the progressive form is a standard test for non-
statives. However, Arad (1998: 6) maintains that in some cases the stative
reading cannot be limited only to this syntactic diagnostic. To be specific,
on the basis of the sentences, reproduced in (2.17), Landau (2010) draws
the conclusion that the verb depress cannot be treated as stative in the
active (2.17a) version, since it can appear in the progressive. Nevertheless,
its passive form, as in (2.17b) is stative, and cannot occur in the
progressive; even though it is well-known that verbal passivization does
not change verbs from stative to non-stative or the other way round.
(2.17) Adjectival passive with stative class II OE psych-verbs:

a. The situation is depressing Mary.
b. * Mary is being depressed by the situation.
(cf. Grimshaw 1990:114, ex. 13; Landau 2010: 49)
16
Guidi (2011: 42) argues that this problem of ambiguous aspectual interpretation
can occur for verbs with dative Experiencers too, but he does not discuss this issue
further.
92 Chapter Two
Landau (2010: 99) argues that the passive in the ungrammatical

sentence (2.17b) is adjectival, what justifies the stativity of the verb
depress (cf. Grimshaw 1990: 114; Pesetsky 1995; and Grafmiller 2013).
Pesetsky (1995) claims that class II verbs vary in their stative behaviour,
as illustrated in (2.18).
(2.18) Passive with stative psych-verbs:

a. Sue was continually being scared by odd noises.
b. Harry is clearly fearing an outbreak of the flu.
c. *An outbreak of the flu is clearly being feared by Harry.
d. An outbreak of the flu is feared by Harry.
(Pesetsky 1995 (73a, 75e, 76e, 77e))
In contradistinction to depress (which is strongly stative), the verbs

scare, terrify, shock and surprise reveal an eventive reading both in the
active and the passive, as in (2.18a). Pesetsky (1995) further claims that
(2.17a) has a special (“judgmental”) non-iterative meaning which (for
some reason) is unavailable with passives. This restriction applies to other
statives, such as class I SE psych verbs, exemplified in (2.18b-d).
Additionally, as noted by Landau (2010: 50-51), English is expected to
have eventive verbal psych passives since their verbal status is proved by
the fact that these passives in the progressive are incompatible with special
idiosyncratic prepositions, as in (2.19).
(2.19) Verbal passive with eventive (non-stative) psych-verbs:

a. Bill was enraged by/at totally innocent remarks.
b. Bill was often being enraged by / *at totally innocent remarks.
(Landau 2010: 57-58; cf. Pesetsky 1995: ex. 81)
Landau (2010: 57) claims that such idiosyncratic prepositions are a

proof of adjectival passives, which are lexically derived; thus (2.19a)
sounds grammatical. On the other hand, these prepositions are disallowed
in contexts that force the choice of a verbal passive, like the progressive
aspect, given in (2.19b).
Nevertheless, Pesetsky (1995) argues that some class II psych-verbs do
not passivize at all, as illustrated in (2.20a-b), which makes them similar to
class III OE psych-verbs that never form passives, as in (2.20c-e) (cf.
Perlmutter and Postal 1984). These facts make Pesetsky (1995) suggest
that all these verbs are unaccusative as they do not form passives. 17
(2.20) Psych-verbs which do not form passive forms:

a. * We were escaped by Smith’s name.
b. * Panini was eluded by the correct generalization.
c. * Mary wasn’t appealed to by the play.
d. * John was mattered to by this.
e. * Mary was occurred to by the same idea.
(Pesetsky 1995, ex. 153b, 154b, 155b, 156b, 157b)
Since some OE psych-verbs do not form passives and fail the

progressive test, as exemplified in (2.20), a pseudocleft test can be used to
distinguish stative from eventive readings (cf Landau 2010: 101). Stative
verbs fail the pseudocleft test, whereas eventive class II verbs pass it, as
shown in (2.21a) and (2.21b), respectively.
(2.21) A Pseudocleft test to distinguish stative from eventive readings of

OE verbs:
a. * What that solution did was escape/elude/concern Mary.
(stative reading)
b. What that noise did was scare/surprise/startle Mary.
(eventive reading)
(Landau 2010: 50)
In addition, Grafmiller (2013) confirms that the semantic distinction

between the stative (adjectival) and eventive (verbal) forms of passive
participles is subtle, and over the years various grammatical diagnostics
have been proposed for distinguishing between them syntactically (cf.
Grafmiller 2013: 76; and Wasow 1977: 338-341, who provide the criteria
for identifying adjectival character of passive participles). Grafmiller
(2013: 87-96) claims that at least some OE verbs can form verbal passives
providing they satisfy the following criteria, listed in (2.22) below:
17
Further evidence for the unaccusativity of escape and elude comes from the fact
that they do not form middles or –er nominals (similarly to concern and interest),
as in (i) (cf. Pesetsky 1995; Levin 1986).
(i) a. * Great ideas elude/escape/concern/interest easily.
b. * an eluder, *an escaper, *a concerner, *an interested.
(Landau 2010:50)
94 Chapter Two
(2.22) The criteria for identifying verbal passive with eventive (non-
stative) psych-verbs:
a. use in the iterative progressive, e.g.

(i) Odd noises were continually scaring Sue.
(eventive psych-verb)
(ii) ??Odd noises were continually depressing Sue.
(stative psych-verb)
(iii) If you turn on the TV and are continually being bored by
the programming, it’s likely you have the wrong type of
cable package.
b. punctual past (by means of the adverb suddenly), e.g.

(i) Suddenly he was scared by an unexpected groan from the
next room.
(ii) ??An unexpected groan from the next room suddenly
depressed him.
(iii) ??Suddenly he was depressed by an unexpected groan from
the next room.
c. needs V-ed construction more frequent than with statives,

e.g.
(i) Nobody needs angered / upset by the truth.
(ii) Young people shouldn’t need depressed / concerned by life.
(cf. Pesetsky 1995: 29-30, ex. 71, 73;
Grafmiller 2013: 88-94)
All the instances in (2.22) require supposedly eventive interpretations of

the predicate, and therefore they work as diagnostics of verbal passives. In
fact, Grafmiller (2013) proves that any OE verb can be used in the
progressive passive with an iterative interpretation – even those that are
most frequently claimed to denote states, e.g. bore, concern, depress and
worry, as in (2.22a)(iii).
Moreover, some verbs like depress are said to be unacceptable when
modified by adverbs like suddenly in the past tense, while other verbs, e.g.
scare, sound perfectly fine modified this way, as shown in (2.22b)(i)-(iii).
Grafmiller (2013: 91) explains that the sentences modified by the adverb
suddenly describe a punctual change in the Experiencer, hence the
predicate is required to be interpreted as an event, not as a state.
Consequently, only those verbs which are acceptable with such
interpretations allow verbal passives.
Finally, the needs V-ed construction, as in (2.22c)(i)-(ii), provides
another piece of evidence for the existence of verbal passives with some
OE verbs. The sentence in (22c)(ii) is fine to Grafmiller (2013), even
though it involves two verbs that are normally listed as stative OE psych-
verbs, viz. concern and depress. Indeed, Grafmiller’s (2013) analysis of
the corpus data shows that eventive and stative uses are available for all
OE verbs in both the active and passive, which runs counter to many
authors’ claims (Arad 1998; Bouchard 1995; and Landau 2010; among
others).
Additionally, the needs V-ed construction can be mainly found in
dialects of western Pennsylvania, central Ohio, and other parts of the Great
Lakes area of the U.S. Landau (2010: 51) points out that the Pittsburghese
dialect of English provides further evidence for eventiveness, rather than
agentivity of verbal psych passives (cf. Tenny 1998; Grafmiller 2013: 91-
95).18 Indeed, the more eventive the verb, the more felicitous verbal
passives are. However, Tenny (1998) notes that “a complex of factors
influences the degree of eventiveness, including not only agentivity but
also volitionality, punctuality, and the affectedness of change of state in
the experiencer. ... Individual speakers vary in how strict they are with this
scale in making verbal passives” (Tenny 1998: 595). Therefore, relying on
Tenny’s (1998) analysis, Landau (2010: 51) draws the conclusion that
English provides evidence from independent sources for the possibility of
verbal passive to justify the non-stativity of class II verbs.
What is more, Verhoeven (2010: 18-19, 42-44) carries out some
diagnostic tests for agentivity and stativity, in order to identify semantic
properties of particular verbs of different psych-verb classes in five
different languages. Especially in those languages which display a
18
In her analysis of this dialect, Tenny (1998) explains that the verbal passive
participle construction is well-matched to eventive adverbials, progressive aspect
and idiom chunk passives, and mismatched with the adjectival un-passive, as in
(i)a-(i)d, respectively.
(i) a. The dog needs scratched hard.
b. The car has been needing washed for a long time now.
c. Tabs need kept on the suspect.
d. * The house needs unpainted.
(Landau 2010: 51 (102))
96 Chapter Two
grammaticalized expression of the progressive aspect, the verbs were

tested within the corresponding constructions. Thus, three standard
diagnostic tests that were implemented in this study comprise: (a) the
VOLITIONALITY TEST examines the compatibility of the verb with an
adverb denoting the volitional involvement of the actor, e.g. the adverb
intentionally, (b) the IMPERATIVE TEST examines whether an order can
be expressed by using the imperative form of the verb and provides further
evidence for the possibility of an agent to have volitional control over the
event, and (c) the STATIVITY TEST examines whether the verb can be
used in a form or context that implies a dynamic internal temporal
structure of the event.
All these aspectual properties of psych-verbs and the valid syntactic
tests used to distinguish the different readings which have been mentioned
earlier in this section, are summarised in (2.23). In all these exemplary
sentences in (2.23), Mary is an Experiencer, while ‘dogs’ or ‘children’ are
the Theme.
(2.23) Aspectual properties of psych-verbs:
a. class I: SE psych-verbs
Mary loves / fears dogs. (stative reading)
Test 1: not possible in the progressive tenses
*Mary is loving / fearing dogs.
Test 2: not used in imperatives
Love / *fear dogs!19
b. class II: OE psych-verbs

(i) eventive agentive (for regular transitive verbs)
These children frightened / annoyed Mary.
Test 1: with an adverb on purpose / deliberately / intentionally

These children frightened / annoyed Mary on purpose /
deliberately / intentionally
Test 2: not with an adverb unintentionally
These children frightened / annoyed Mary *unintentionally.
Test 3: used in imperatives
Frighten / annoy Mary!
19
The SE psych-verb “to love” is stative, but shares some properties with eventive
verbs, e.g. the imperative Love children! sounds good.
Test 4: with for-adverbial

These children were annoying / amusing Mary for an hour / *in
an hour.
(ii) eventive non-agentive (for stative or transitive verbs)
These children frightened / annoyed Mary.
Test 1: possible with an adverb unintentionally
These children frightened / annoyed Mary unintentionally.
Test 2: not possible with an adverb on purpose / deliberately /
intentionally
These children frightened/annoyed Mary *on purpose/
*deliberately/ *intentionally
Test 3: used in imperatives
Frighten / annoy Mary!
Test 4: with in-adverbial
These children frightened/annoyed Mary in an hour / *for an
hour.
(iii) stative reading (for unaccusative verbs20)
Dogs depress / concern Mary.
Test 1: not possible with progressive tenses

* Dogs are depressing / concerning Mary.
* Depress / concern Mary!
These children depressed / concerned Mary for an hour / *in
an hour.
c. class III: OE psych-verbs
stative reading (for unaccusative verbs)
Dogs matter to / appeal to Mary.
Test 1: not possible in the progressive tenses
* Dogs are mattering / appealing to Mary.
* Matter / appeal to Mary!
These dogs mattered to / appealed to Mary for two years / *in
an hour.
20
On the unaccusative status of stative OE verbs, cf. section 2.5.
98 Chapter Two
Undoubtedly, one of the reasons for the existence of such a diversity of

analyses for the Aktionsart of psychological verbs relates to the inherent
syntactic complexity they manifest. As is well known and generally
assumed, psych-verbs can participate in various alternations in different
languages (see e.g. Belletti and Rizzi 1988). However, only a stative
reading is associated with the so-called “psych” properties. Indeed, as first
noted by Postal (1971) and Dowty (1991), among others, OE psych-verbs
display peculiar characteristics when they are stative. Therefore, the most
essential distinction is the one between stative and non-stative (eventive or
agentive) readings of OE psych-verbs. As shown in (2.23b)(i)-(iii), stative
OE can be distinguished from non-stative, i.e. eventive and agentive OE
psych-verbs, by means of some syntactic diagnostics, viz. progressive
aspect and forming imperatives. Whereas adverbs unintentionally / on
purpose / deliberately may serve as diagnostics to separate eventive from
agentive readings of OE psych-verbs, as illustrated in (2.23b)(ii)-(iii).
Besides, the in-adverbial triggers a telic interpretation in the eventive
reading, while the for-adverbial forces an atelic reading in the stative
reading. With the for-adverbial there is usually an ambiguity between an
activity and a state, but, given that the activity reading is only available
with agents, it is excluded with the non-agentive subject of eventive OE
verbs. (cf. Arad 2002; Grafmiller 2013; and Alexiadou and Iordăchioaia
2014; among others).
In a nutshell, it is commonly presumed that across different languages,
all class III verbs are stative; class I predicates include verbs with stative
readings; whereas most class II verbs have stative or eventive readings.
Widely recognized is the fact that the peculiar psych properties of OE
verbs occur only on their non-agentive readings. OE verbs are ambiguous
between: stative, causative eventive (non-agentive) and agentive readings.
Even the difference between eventive and agentive is not very clear.
However, psych-verbs on their agentive reading behave in a way similar to
all other transitive Agent-Patient predicates. In other words, when the
arguments of agentive verbs are canonical event participants (with both an
agent and a change of state), the verbs also have an external argument, a
canonical object, and display no psych effects. Nonetheless, when a
predicate has neither an agent nor expresses a change of state, it does not
have a canonical subject and object either (Arad 1998: 9). Therefore, as
regards psych-verbs, there is a correlation between semantic / aspectual
properties of the predicate and its syntactic realisation.
2.5 Syntactic structures and characteristics of psych-verbs

In general, psych-verbs are recognised to have a special status within the
grammatical system of a language. Landau (2010) emphasizes that
Experiencers are “grammatically” special, since they show a great
structural variety and properties which distinguish them from non-
psychological structures. On account of their distinctive properties, i.e.
their “misbehaving” in numerous respects, psychological predicates have
given rise to various approaches to the lexicon-syntax interface.
This section of the chapter is to review briefly the main accounts which
concern syntactic structures (sections 2.5.1-2.5.4), and characteristics of
psych-verbs (sections 2.5.2.1-2.5.2.3). What all these approaches have in
common is that they offer alternative solutions to the problem that
psychological predicates pose to linking. Importantly, the licensing of
arguments has been maintained to be correlated with templatic
information, i.e. with a fixed number of slots for different elements, which
are fixed in their position and order relative to each other (Mattissen 2003:
286). This information, in turn, determines the event structure of a
predicate. In this respect, any proposal about the licensing of a predicate’s
arguments is likewise a proposal about the predicate’s event structure.
Nevertheless, due to the space limitation, only the most crucial accounts
are to be presented, while an interested reader is asked to refer to the
literature for more details.
Psych-verbs were first analysed by Postal (1971), and since then, two
opposing approaches have emerged to explain the unexpected behaviour
of psych predicates: (i) syntactically motivated that implies syntactic
movement; and (ii) semantically based that appeals to thematic relations.
These two stances have run parallel to each other up till now, trying to
provide an answer to some problems posed by psych-verbs. The core
problematic issue refers to linking, i.e. mapping of the arguments of
psych-verbs from lexicon to syntactic position. Since psych-verbs show
different syntactic realizations of the Experiencer argument, which
surfaces either as a subject or as an object, they pose a problem for the
U(T)AH, reproduced in (2.5a-b). The problem refers to the pairs like those
in (2.24)-(2.35), replicated after Pesetsky (1995).
(2.24) a. Bill was very angry at the article in the Times.

b. The article in the Times angered/enraged Bill.
(2.25) a. The paleontologist liked/loved/adored the fossil.
b. The fossil pleased/delighted/overjoyed the paleontologist.
100 Chapter Two
(2.26) a. Bill disliked/hated/detested John’s house.

b. John's house displeased/irritated/infuriated Bill.
(2.27) a. Bill was satisfied/content with the Chinese dinner.
b. The Chinese dinner satisfied/contented Bill.
(2.28) a. Sue resented Bill’s remarks.
b. Bill’s remarks embittered Sue.
(2.29) a. Mary rejoiced at the French victory.
b. The French victory cheered/exhilarated Mary.
(2.30) a. John worried about the television set.
b. The television set worried John.
(2.31) a. Bill was furious about/fumed about the article in the Times.
b. The article in the Times infuriated Bill.
(2.32) a. Sue’s remarks puzzled us.
b. We puzzled over Sue’s remarks.
(2.33) a. Sue grieved over/at the court decision,
b. The court decision grieved Sue.
(2.34) a. John is bored with the problem of lexical entries.
b. The problem of lexical entries bores John.
(2.35) a. Bill fears/is afraid of ghosts.
b. Ghosts frighten Bill.
(Pesetsky 1995: 18)
In the above-mentioned pairs, in the (a) examples of (2.24)-(2.35), there

are SE psych-verbs, i.e. the Experiencer is the subject and the Theme is the
object; whereas the (b) examples of the Experiencer functions as the
object. To be precise, e.g. the Experiencer Bill in (2.24a), (2.26a)-(2.27a),
(2.31a), (2.35a) occupies the subject position. In (2.24b), (2.26b)-(2.27b),
(2.31b), (2.35b), Bill fills the object position, then it should be associated
with the thematic role of Theme. Bill still acts as an Experiencer, even
though the syntactic position is switched from the subject to the object,
which contradicts the UTAH.
In other words, psych-verbs pose a problem for Baker’s (1988)
Uniformity of Theta Assignment Hypothesis (UTAH). According to the
UTAH, there is a systematic relation between thematic information and
syntactic projections. Thus, an argument that bears a particular thematic
role is consistently mapped onto the same syntactic position at D-structure,
e.g. an Agent is always projected as a subject. However, psych-verbs
exhibiting irregular mapping, contradict the UTAH. Following the
examples given in (2.24a,b)-(2.35a,b), it can be observed that both (a) and
(b) examples share the same thematic roles of Experiencer and Theme, that
are realised in different syntactic positions (cf. Belletti and Rizzi 1988;
Grimshaw 1990; Pesetsky 1995; Arad 1998, 2002; Anagnostopoulou

1999; Pylkkänen 2000; Reinhart 1996, 2002; Verhoeven 2008; Isse 2008;
Landau 2010; and Alexiadou and Iordăchioaia 2014; among others)
Thus, one of the puzzles concerning the analysis of psych-verbs
regards the explanation of how apparently equivalent thematic relations
can be realized in different positions. A considerable number of proposals
has been made in the literature so far to solve the puzzle psychological
predicates pose for linking. What follows is an overview of the latest and
the most crucial theories regarding psych-verbs.
2.5.1 Belletti and Rizzi’s (1988) unaccusative approach to OE

psych-verbs
One of the most popular solutions to the problem posed for linking,
formulated within the transformational framework, takes the unifying
thematic factor as its starting point. It is Belletti and Rizzi (1988) who
propose that the thematic correspondence between different kinds of
psychological verbs can be explained by means of an equivalent or, at
least, a similar deep structure for all of them. To save the U(T)AH, Belletti
and Rizzi (1988) postulate treating (OE) psych-verbs as unaccusative
verbs, which lack an external argument and are not expected to assign case
to D-structure objects.
Belletti and Rizzi’s (1988) classification of psych-verbs into three
classes, as illustrated in (2.3), and repeated for convenience in (2.36),
identifies psych-verbs as sharing identical theta grids. This assumption
goes against Baker’s (1988) UTAH.
(2.36) a. Class I: The temere class

(Nominative Experiencer, accusative Theme)
Gianni teme questo
Gianni fears this
b. Class II: The preoccupare class

(Nominative Theme, accusative Experiencer)
Questo preoccupa Gianni
This worries Gianni
c. Class III: The piacere class

(Nominative Theme, dative Experiencer)
(i) A Gianni piace questo
To Gianni pleases this
102 Chapter Two
(ii) Questo piace a Gianni

This pleases to Gianni.
For Belletti and Rizzi (1988), the only aspect in which the three verb
classes differ from one another is the specification of case-grid, which
governs diverse derivations from deep to surface structure. They claim that
the Experiencer is a deep structure subject with verbs belonging to class I
(Nominative Experiencer, accusative Theme). Thus, the D-structure
configuration of sentence (2.36a), which comprises an SE psych-verb
(class I), is illustrated in (2.37):
(2.37) D-structure configuration of SE (class I) psych-verbs:
(cf. Belletti and Rizzi 1988: 293)
Moreover, the D-structure of psych-verbs belonging to class II

(Nominative Theme, accusative Experiencer) and class III (Nominative
Theme, dative Experiencer), exemplified in (2.36b) and (2.36c)
respectively, is represented in (2.38). The verbs from these classes form,
according to Belletti and Rizzi (1988), a double object construction with a
nonthematic subject position.
(2.38) D-structure configuration of OE (class II and class III) psych-verbs:
(Belletti and Rizzi 1988: 293)
Belletti and Rizzi (1988: 293-294) explain that sentences such as

(2.36b) and (2.36c) may be derived from (2.38) via NP-movement to the
subject position. In the tree diagram in (2.38), the Theme originates as an
internal argument and moves to the subject position. The structure in
(2.38) contains the Experiencer in a higher position than a Theme. The
verbs of both class II and class III are treated by Belletti and Rizzi (1988)
as double object unaccusatives. Significantly, they argue that the Theme
argument in both classes of verbs is internal, i.e. is a sister of the V head.
In class III the Experiencer is assigned the case of an indirect object
(dative).
Furthermore, despite the obvious differences in linear order, in both
(2.37) and (2.38), the verb directly ș-marks the Theme, and the constituent
Verb + Theme compositionally ș-mark the Experiencer. Consequently,
Belletti and Rizzi (1988: 344) form a hypothesis, as in (2.39):
(2.39) Assumptions about the Mapping:

Theta hierarchy: Agent > Experiencer >…….. > Theme
Hypothesis:
syntactic configurations projected from a given ș-grid should
reflect the hierarchy, so that for every pair of ș-roles in the ș-
104 Chapter Two
grid, the higher role in the hierarchy is projected to a higher

structural position.
(Belletti and Rizzi 1988: 344, fn. 36)
In addition, on the basis of the hierarchy and the hypothesis in (2.39), the
following assumption can be made:
(2.40) Given a ș-grid [Experiencer, Theme], the Experiencer is

projected to a higher position than the Theme.
(Belletti and Rizzi 1988: 344)
Accordingly, Belletti and Rizzi (1988: 344) propose for the three
psych-verb classes the Lexical Representations generating the required
syntactic configurations, as in (2.41):
(2.41) a. Class I - temere [fears]: ș-grid [Experiencer, Theme]

Case grid [ _ _ ]
b. Class II - preoccupare [worries]:
ș-grid [Experiencer, Theme]
Case grid [ ACC _ ]
c. Class III - piacere [pleases]:
ș-grid [Experiencer, Theme]
Case grid [ DAT _ ]
As shown in (2.41), within class I of psych-verbs, which represent the

uncontroversial transitive structure, the Experiencer has the external ș-
role, and no inherent Case is assigned. In class II and class III there is no
external ș-role, while the Experiencer is associated with an inherent Case
(accusative or dative). Since it has no external ș-role, the Experiencer must
be generated in a VP-internal position where it can be governed. In
accordance with (2.40), the Experiencer must be higher than the Theme.21
In addition, psych-verbs of class II and III must have an unaccusative
structure (with no external ș-role). The Theme, on the other hand, not
being assigned case in its original position, has to move to the subject
position to be assigned structural nominative case there. The Theme, thus,
may also satisfy the Extended Projection Principle (EPP).22 Likewise,
Grafmiller (2013) notes that this is basically what happens with
21
Cf. Pesetsky’s (1987) comments on Belletti and Rizzi (1988)’s assumption about
the identical ș-grids of all the three classes of psych-verbs.
22
Cf. Chomsky (1982) for the Extended Projection Principle (EPP), which
concerns the obligatoriness of subjects.
intransitive unaccusatives (e.g. The vase broke), whose single arguments

are internal arguments, which must move to the subject position to satisfy
the EPP (Grafmiller 2013: 14).
What is more, Belletti and Rizzi (1988) put forward a number of
arguments in favour of the unaccusative analysis of class II psych-verbs.
The first one is based on the subject, which is said not to be a deep subject.
They justify this standpoint by referring to some syntactic phenomena, a
few of which, summarised in (2.42) below, are valid cross-linguistically.
(2.42) Arguments for an unaccusative analysis of Class II - preoccupare

verbs:
a. Passives:
Structures with non-thematic subjects cannot undergo
passivization.23 Even though some verbs of the preoccupare
class allow passives, these are not true passives but rather
adjectival ones, e.g.:
Gianni è disgustato dalla corruzione di questo paese
Gianni is disgusted by the corruption of this country.
b. Binding:
The Experiencer in the object position can bind an anaphor
in the subject position, e.g.
Questi pettegolezzi su di sé preoccupano Gianni piú di ogni
altra cosa
These gossips about himself worry Gianni more than
anything else
However, Belletti and Rizzi (1988: 312-313) argue that these examples
involve D-structure binding. Instead, they propose that Principle A is an
‘anywhere principle’, thus it can be satisfied at D-structure, or at S-
Structure (or LF), e.g.
They seem to each other [t to be intelligent]
Principles B and C must apply at S-Structure (unlike Principle A, which is

an anywhere principle) in order to account for the ungrammaticality of:
23
Cf. Belletti and Rizzi (1988: 306-308) for a complication with fare + Infinitival
PPs.
106 Chapter Two
*Himselfi worries Johni/himi.24

The further arguments for the unaccusative status of class II psych-

verbs, provided by Belletti and Rizzi (1988: 324-334), concern the
properties of the object of preoccupare verbs. They claim that this object
is not a canonical object, for the following syntactic reasons summarised
in (2.43).
(2.43) Properties of the object of preoccupare-verbs:
a. Lack of island properties:
The object of preoccupare-verbs is not a canonical object, but it

is the sister of V’, immediately dominated by VP. Therefore, if
this assumption is correct, the Experiencer must lack typical
properties of canonical objects. One of these characteristics is
transparency to extraction processes.
Objects of temere-verbs are transparent to wh-extraction (cf. (i)
below), while objects of preoccupare-verbs are not (a similar
violation appears with ne-cliticization, though the violation is
weaker than with wh-phrases), cf. (ii) below.
(i) La compagnia di cui tutti ammirano il president

The company of which everyone admires the president
(ii) *La compagnia di cui questo spaventa il president
The company of which this frightens the president.
b. The Accusative Case of the Experiencer of preoccupare-verbs:
The Accusative Case overtly manifested under cliticization is a

canonical object property:
Questo lo preoccupa
This him worries.
However, this is not a structural Accusative, but inherent
Accusative Case. Otherwise, Burzio’s Generalization (Burzio
24
An interested reader is referred to Belletti and Rizzi (1988: 295-324) for more
details and other arguments in favour of the unaccusativity of class II psych-verbs,
such as: anaphoric cliticization, focus and agentivity effects, arbitrary pro, and
causatives.
1989: 178-186) (V is a structural Case assigner iff it has an

external argument) would be violated, which Belletti and Rizzi
(1988) take to be a generalization about Structural Case.
c. Selection of auxiliary avere “have”:
Belletti and Rizzi (1988: 333) suggest that auxiliary selection is

not an unaccusativity diagnostic, but instead, avere “have” not
essere “be,” is chosen when the verb is able to assign accusative
case (structural or inherent).
What is more, Belletti and Rizzi (1988: 334-342) provide a detailed

analysis of the properties of the object of piacere-verbs, recapitulated in (2.44).
(2.44) Properties of the object of piacere-verbs:

a. The experiencer bears dative Case:
Being unaccusative, the verbs cannot assign structural Case.

They assign inherent dative, unlike preoccupare-verbs that
assign inherent accusative.
b. The auxiliary selected is essere “be”:

Since they assign dative, they select essere “be,” in
contradistinction to preoccupare-verbs, which select avere
“have” because they assign accusative. This property classifies
the verbs as unaccusatives.
c. The orders EXP-V-THEME and THEME-V-EXP are both

equally possible:
Belletti and Rizzi (1988: 336-342) note that this property refers
to the fact that the Experiencer bears dative case, which is
assigned by the preposition/case marker ‘a’. Both dative and
accusative realization at S-structure must be in the government
domain of an appropriate case marker, which is either the verb
or an inserted preposition. At S-structure the dative realization
of the NP is assigned by the governing preposition, and the
a+NP dative Experiencer is permitted to move around freely.25
25
For an explanation why in the order EXP-V-THEME the Experiencer takes a
Subject rather than a Topic position, cf. Belletti and Rizzi (1988: 339 fn. 32).
108 Chapter Two
Belletti and Rizzi’s (1988) arguments in favour of unaccusativity of

class II psych-verbs have been discussed in the literature by many linguists
(cf. Grimshaw 1990; and Pesetsky 1995; among many others). Indeed, as
recognised by Levin and Rappaport Hovav (2005: 142-144), Belletti and
Rizzi’s (1988) study of psych-verbs has its place in the class of prominent
hypotheses. Larson (1990: 601) observes that Belletti and Rizzi’s
unaccusative approach to OE psych-verbs leads to a Relativized Uniformity
of Theta Assignment Hypothesis (RUTAH, i.e. relativized UTAH), since
an Experiencer is always higher than the Theme (cf. Baker 1997). As a
consequence of A-movement of the Theme into the subject position with
OE verbs, the arguments reveal the different surface realization.
To sum up, Belletti and Rizzi (1988) argue that OE verbs, in spite of
being apparently identical to transitive verbs, differ from them
substantially in their syntactic behaviour (e.g. reflexive si, causativiztion
and object extraction, among others). That is why, OE verbs receive a
special position in theories of the syntax-lexicon interface. Their lexical
uniqueness is marked with the fact that one of the arguments of psych-
verbs is specified as an Experiencer. Besides, having different syntactic
structure from usual transitive verbs, makes OE psych-verbs syntactically
distinct. According to Belletti and Rizzi (1988), OE verbs form a unique
class both semantically and syntactically, and therefore they should be
assigned a unique structure, the so-called “psych structure,” based on the
structure of unaccusative verbs.
2.5.2 Landau’s (2005, 2010) locative approach

In his recent book on psych-verbs, called The Locative Syntax of
Experiencers (LSE), Landau (2005, 2010) presents evidence, from a
variety of languages and sources accumulated over the years, that
Experiencers are conceptually encoded as “mental locations–containers or
destinations of mental states/effects” (Landau 2005: 7). Consequently,
Experiencers are essentially locatives, in the sense of receivers of
experience (cf. Guidi 2011: 32)
Similarly to Belletti and Rizzi (1988), Landau (2010) treats Class I SE
psych-verbs as regular transitive verbs. Landau (2010) follows, e.g. Arad
(1998, 1999), in stating that all class III verbs are stative (unaccusative),
which are never used agentively, while most class II verbs are ambiguous
between the three readings, i.e. (i) stative (unaccusatives), (ii) eventive
agentive (regular transitives), (iii) eventive non-agentive (stative /
transitive).
2.5.2.1 Subject-Experiencer verbs
Landau (2010) adopts Arad’s (1998) argument that even stative SE verbs
can denote locative relations. Thus, the Experiencer (Monica, he) is either
conceived of as the “substance” contained in the mental state or the
container in which the mental state resides, as illustrated in (2.45).
(2.45) a. Monica is in love (with Paul).

b. There is in him a great appreciation for artists.
(cf. Arad 1998: 228 (83))
Landau (2010: 11) argues that his contention that the Experiencer
denotes a mental location holds true, even when the Experiencer occurs as
a bare nominal, as in the case of SE verbs in Hebrew, French and Navajo.
He emphasises that in those languages and in many others, there occur
frequently periphrastic constructions, comprising the verbs be / have, a
psych noun and an Experiencer location, as illustrated in (2.45a)-(2.46a).
Besides, in Irish and Scottish Gaelic, Experiencers are solely introduced
by locative prepositions, as exemplified in (2.46b).
(2.46) a. Il ne pouvait plus contenir sa rage.

he not could more to-contain his rage
He could no longer contain his (own) rage
(Bouchard 1995: 266, ex. 13g)
b. Tá eagla roimh Y ar X.
is fear before Y on X
X is afraid of Y
(McCloskey and Sells 1988, ex. 77a)
In addition, Landau (2010) emphasises the fact that SEs, which are
stative transitives, behave in a unique way when they co-occur with a
locative preposition (cf. Doron 2003). Accordingly, Landau (2010: 12)
proposes that SE psych-verbs can also be treated as having an oblique
Experiencer, even in languages like English, where they always take the
nominative (nonoblique) form. To justify this assumption, he evokes
Speas’ (1990) arguments concerning SE verbs, which are said to introduce
a path, either as a goal or a source, unlike non-Experiencer Subjects, as
illustrated in (2.47) and (2.48).
(2.47) a. I got angry but it went away.

b. ?? I laughed but it went away.
110 Chapter Two
(2.48) a. I tried to remember his name, but it wouldn’t come to me.

b. ?? I tried to write his name, but it wouldn’t come to me.
(Speas 1990, ex. 3,7)
Another argument in favour of the locative character of SE emerges in

Hebrew, where there exists a paradigm of adjectival passives in non-psych
predicates, beynoni pa’ul, which expresses the original (verbal) external
argument in a by-phrase, as reproduced in (2.49a).
(2.49) a. ha-sefer arux al-yedey orex mikco’i.

the-book edited by editor professional
The book is edited by a professional editor
b. ha-šir ha-ze a’huv/mu’adaf al/*al-yedey harbe ma’azinim.

the-song the-this loved/hated on/*by many listeners
This song is loved/hated(Adj) by many listeners
(Landau 2010: 13-14 (25a), (27a))
Landau (2010: 14) argues that only in the case of SE verbs the
preposition al-yedey “by” is supplanted by a different preposition. In the
beynoni pa’ul of these verbs, the original external argument (the
Experiencer) surfaces with the locative preposition al “on,” as illustrated
in (2.49b). What is more, the same preposition al “on” also occurs with SE
transitive verbs in lexical causativization. There the original object
remains accusative, and the original subject (the Experiencer) becomes
oblique, with the preposition al “on,” as exemplified in (2.50).
(2.50) a. Gil sana/xibev et beyt-ha-sefer.

Gil hated/like ACC the-school
Gil hated/liked school
b. Rina hisni’a/xibeva al Gil et beyt-ha-sefer.

Rina made Gil hate/like school
(Landau 2010: 15 (29a,b))
Consequently, as discussed above, Landau (2010) distinguishes SE

psych-verbs from all other subjects, providing some evidence for a
locative preposition present in the SE structures. Landau (2010)
emphasises that it is the syntactic behaviour of Experiencer objects which
deviates from that of canonical objects in various languages.26 For

decades, these so-called psych effects have been studied in the theoretical
as well as recent experimental research. The distinct properties of OE
psych-verbs will be of main interest in the subsequent section.
2.5.2.2 Object-Experiencer verbs and their syntactic properties
Limiting his analysis to the VP-structure, and following an extensive

discussion in Pesetsky (1995) and Iwata (1995), Landau (2010) claims that
a psychological predicate is identified in the grammar by the presence of a
specific structure. The structure proposed by Landau (2010: 8) for OE
psych-verbs is the one presented in (2.51) below.
(2.51)
(Landau 2010: 8)
The verbal structure of OE psych-verbs in (2.51) shows an Experiencer

licensed within a prepositional phrase. Besides, with this concrete
structural representation for OE psych-verbs in mind, Landau (2010: 8)
offers more detailed structures in (2.52) and (2.53) below, for both class II
and class III verbs. He notes that psych-verbs are special due to the
oblique nature of their Experiencers.
With class III verbs, regarded as unaccusative, for most languages the
Experiencer is either encoded by means of an oblique case (often dative)
or by means of a PP, following Belletti and Rizzi (1988), Pesetsky (1995),
and Arad (1998), among others. This assumption is also made by Landau
(2010: 19-20), who points out that object Experiencers universally bear an
inherent case and that the inherent case is universally assigned by a P. In
the case of class III verbs, the Theme argument of these verbs is not a
Causer but rather a Target/Subject Matter, T/SM (Pesetsky 1995). Besides,
in languages where the dative marker is not an independent preposition,
class III Experiencers are governed either by a lexical preposition
(English) or a null preposition Øȥ (in languages with morphological case),
26
As stated by Landau (2010), the specific syntactic behaviour of OE is best
visible in peculiarities concerning binding, extraction/islandhood, reflexivization
and argument linearization, etc. (cf. Mohanan and Mohanan 1990)
112 Chapter Two
which assigns the dative case. The VP structure of class III psych-verbs is
then as in (2.52).
(2.52) Class III verbs – unaccusatives
[VP [PP P DP] [V’ V DP]]

Experiencer Theme
(Landau 2010: 8, ex. (12b))
In (2.52), the Experiencer may move overtly to the subject position

(depending on the language), resulting in the so-called “quirky” subjects,
or covertly, forming the “second” subject (Landau 2010: 88). The latter
case is valid for languages like English, which prohibit inherent case-
marked Determiner Phrases (DPs) in the specifier of Tense (Spec,TP)). In
English, the Theme argument raises to [Spec,TP] overtly, and the
Experiencer raises to a second [Spec,TP] at LF. This effect not only
creates a multiple-specifier structure, but it also is referred to as LF-
quirkiness by Landau (2010: 87). Besides, class III OE verbs are stative
unaccusatives, which means that they can never be used agentively or
eventively.
Undeniably, the oblique nature of Experiencers is much less obvious
with class II verbs, which in many languages apparently occur with
nominal (accusative) object Experiencers. Dividing class II verbs into
three groups, Landau (2010) distinguishes: (i) stative verbs; (ii) eventive
non-agentive verbs; and (iii) eventive agentive psych-verbs (cf. also Arad
1998, 1999).
The first group of stative class II psych-verbs comprises verbs like
interest, concern, depress, which have the same unaccusative structure as
class III verbs in (2.52), with a difference that the Experiencer in class III
verbs is governed by a lexical P for English, but in class II, it is governed
by a null P. The syntactic structure of class II stative verbs is shown in
(2.53a).
The second and the major group of class II psych-verbs consists of

eventive non-agentive verbs like frighten, startle, anger, surprise, which
may work as statives or transitives. Accordingly, even though some class
II verbs are only either stative (concern) or eventive (startle), the border
line is not clear-cut as some verbs in class II are ambiguous, since they
allow for both stative and eventive readings (frighten). Landau (2010, 55-
56) underlines that, in fact, most class II verbs are ambiguous, to varying
degrees, between stative and eventive readings. The unaccusative status is
exhibited empirically only by those verbs (like concern, interest) which
are unambiguously stative. Thus, he assumes that only stative class II
verbs lack, in their thematic grid, a causer argument, which, in turn, is the
source of eventive interpretation for non-stative verbs.27
Generally speaking, these special psych properties, restricted to non-
agentive OE verbs, as argued by Landau (2010: 127-128), are related to
the presence of a (possibly null) locative preposition with a [loc] feature,
governing the OE. Accordingly, the absence of these properties must be a
sign of the absence of the preposition; thus, agentive contexts (with no
psych properties) exclude the psych-preposition. Besides, due to locative
inversion an Experiencer object in class II verbs is raised to the subject
position, while all Experiencers become LF-subjects, landing in [Spec,
TP], as shown in (2.53) for stative psych-verbs, and in (2.54) for eventive
verbs.
27
To recall briefly, Belletti and Rizzi (1988) identify for Italian Class I of
psychological verbs as the one with the uncontroversial transitive structure, since
the Experiencer has the external ș-role, and no inherent Case is assigned. Class II
and class III of psych-verbs, with no external ș-role, and the Experiencer
associated with an inherent Case (accusative or dative) have, according to Belletti
and Rizzi (1988), an unaccusative structure.
114 Chapter Two
(2.53) Stative psych-verbs: LF
(Landau 2010: 87)

(2.54) Eventive non-agentive psych-verbs: LF
(Landau 2010: 88)
The third group of class II verbs comprises eventive agentive verbs,

considered to be regular transitive verbs that take a direct object. This
class differs from non-agentive predicates since it does not show the
special syntax of psych-verbs. According to Landau (2010), some eventive
psych-verbs in the transitive use have a Causer as an external argument,
projecting a light v (cf. Arad 1988; Pesetsky 1995; and Iwata 1995, among
others), and the Experiencer as an oblique object. The structure for
agentive class II transitives is represented in (2.55).
(2.55) Agentive class II verbs – transitives
[VP DP [V’ V [VP V [PP P DP]]]]

Causer Experiencer
(Landau 2010: 8, ex. (12b))
What is more, refuting Belletti and Rizzi’s (1988) assumption about

the unaccusative status of psych-verbs of eventive class II,28 Landau
(2010) claims (following Pesetsky 1995) that most class II verbs are not
unaccusatives. Besides, he proposes that, commonly, non-nominative
Experiencers bear inherent case, which is assigned by a P (cf. Emonds
1985). This implies that nominals marked for inherent case are always
dominated by a PP node. This PP may be headed by a lexical P (as in
English obliques) or a null P (as in Latin obliques), but both cases are
structurally distinct from bare DPs (Landau 2010: 21-22).29
Additionally, providing a wide range of syntactic properties of OE
verbs in different languages, from a cross-linguistic perspective, Landau
(2010) makes an attempt to prove that Experiencers behave like
datives/PPs, i.e. locatives. Such psych properties are divided, according to
28
Landau (2010: 19-20) deals with Belletti and Rizzi’s (1988) claim concerning
the unaccusativity of II and III class psych-verbs. He says that it is unproblematic
for class III verbs, which assign dative case to the Experiencer and select the
auxiliary essere. Nonetheless, class II verbs with these two properties are never
found: they select the auxiliary avere, and assign the accusative case, violating
Burzio’s (1986) generalization.
29
The trees by Landau (2010) and Harwood et al. (2017) (e.g. (4.54) in Chapter
Four) are syntactic trees containing reference to some semantic roles.
116 Chapter Two
Landau (2010: 75), into core and peripheral ones, with the information
which languages they are typical of, listed in (2.56):
(2.56) A classification of Psych Properties
(I) Core Properties
(a) All Class II Verbs (Non-agentive)

1. Overt obliqueness of Experiencer (Navajo, Irish, Scottish Gaelic).
2. Accusative / Dative alternations (Italian, Spanish).
3. Islandhood of Experiencer (Italian, English).
4. PP-behavior in wh-islands (English, Hebrew).
5. No synthetic compounds (English).
6. No Heavy NP Shift (English).
7. No Genitive of Negation (Russian).
8. Obligatory clitic-doubling (Greek).
9. Obligatory resumption in relative clauses (Greek, Hebrew).
10. No si/ se-reflexivization (Italian, French).
11. No periphrastic causatives (Italian, French).
12. No verbal passive in type B languages (Italian, French, Hebrew).
(b) Class III and Stative Class II (Unaccusatives)
1. No verbal passive (English, Dutch, Finnish).
2. No periphrastic causatives (French, Italian dialects).
3. No forward binding.
(II) Peripheral Properties

1. The T/SM restriction.
2. No causative nominalizations.
3. Backward binding.
(Landau 2010: 75)
As can be seen in (2.56), Landau (2010) distinguishes core psych

properties from non-core ones, i.e. properties that psych verbs share with
other verbs. To recall, all the core psych properties can be encountered
only in non-agentive contexts, while if an agentive context appears, a class
II verb behaves like any ordinary transitive verb (cf. Belletti and Rizzi
1988; Grimshaw 1990; Bouchard 1995; Arad 1998, 2000; and Landau
2010; among others). While Belletti and Rizzi (1988) associate all the
special psych properties with the unaccusative nature of class II verbs,
actually the single issue of unaccusativity cannot distinguish agentive from
non-agentive class II verbs in the general case (cf. Pesetsky 1995). Instead,
Landau (2010) assumes that “the agentivity puzzle” should be resolved

with the meaning shift from a non-agentive to an agentive reading of a
class II verb, which is complemented with an aspectual shift, as stated in
(2.57).
(2.57) a. Agentive class II verbs are change-of-state verbs (i.e.

accomplishments).
b. Non-agentive class II verbs are states or achievements.
(Landau 2010: 129)
Accordingly, OE verbs on the agentive interpretation are change-of-

state verbs, i.e. accomplishments. In the agentive context, the Experiencer
which undergoes a change of state, becomes an affected argument, i.e. a
direct object (Dowty 1991), a bare DP Experiencer. As a result,
Experiencer objects of agentive class II verbs cannot raise to the subject
position, since they are bare nominals receiving the structural accusative
case. In turn, non-agentive class II verbs are states or achievements
(Landau 2010: 129-131).30 The special behaviour of psych-verbs, i.e. their
genuine psych-effects, are restricted to non-agentive contexts (subject as a
Theme), whereas when a psych predicate is used agentively, the subject is
a volitional agent, while the Experiencer plays the role of a Patient.
Furthermore, Landau (2010: 18-19) claims that an oblique construction
forces a non-agentive reading, but a transitive construction does not force
an agentive reading. Thus, the oblique Experiencer correlates with a non-
agentive subject. Indeed, non-agentive OE constructions are universally
oblique. For most languages, the preposition governing the Experiencer is
null (Øȥ), i.e. it involves a PP headed by Øȥ (as in English). The “psych”
prepositions have different versions across languages; in Irish psych
predicates are special in that the preposition introducing an Experiencer is
overt (usually, ar “on”).
In short, Landau (2010: 131) argues that it is the Experiencer that
undergoes the change of state in the agentive context, contrary to the
Experiencer in non-agentive (class II) contexts, which does not undergo a
change of state in the aspectually relevant sense. Instead, the Experiencer
is either a locus where a mental state resides (statives) or appears
(achievements). In these so-called “locative” contexts, Øȥ is a crucial
interpretive ingredient.
30
This approach is compatible with Marín and McNally’s (2011) account and an
earlier analysis of psych-verbs offered by van Voorst (1992). For a further
discussion concerning the aspectual properties of psych-verbs cf. Grimshaw
(1990), Pesetsky (1995) and Pylkkänen (2000), among others.
118 Chapter Two
2.5.2.3 Syntactic characteristics of English psych-verbs
Based on the core and peripheral properties, reproduced in (2.56), which

Landau (2010) lists as typical of psych-verbs, syntactic diagnostics can be
set to distinguish psych-verbs from other verbs, and non-agentive from
unaccusative psych-verbs (cf. section 2.4.2.3, in which syntactic tests for
stativity / non-stativity of class II OE psych-verbs are analysed). Five of
the core properties are applicable to English psych-verbs, as illustrated in
(2.58).
(2.58) a. All Class II Verbs (Non-agentive)

i) Islandhood of Experiencer
ii) PP-behaviour in wh-islands
iii) No synthetic compounds
iv) No Heavy NP Shift
b. Class III and Stative Class II (Unaccusatives)
i) No verbal passive (English, Dutch, Finnish).
(Landau 2010: 75)
Even though prepositional objects in English are not strong islands,

some acceptable instances of prepositional objects may be found, as in
(2.59b).
(2.59) a. *Which film was Dirk amusing to the director of?

b. Which film did Sam entrust Marilyn to the director of?
(Roberts 1991 (43a,c))
Landau (2010: 29) points out, citing Roberts’ (1991) examples in

(2.59), that the Experiencer shows islandhood as an object of a non-
agentive psych predicate, as in (2.59b), but not as an object of an agentive
predicate, as in (2.59a) (cf. Johnson 1992 and Stowell 1986; and Pesetsky
1982).
More to the point, the PP-like behaviour of class II non-agentive
psych-verbs in wh-islands is illustrated in (2.60). The data in (2.60) allow
us to conclude that English treats accusative Experiencers as PPs in certain
contexts. Landau (2010: 29-30) makes a claim that Experiencer objects
behave like adjuncts since they are more resistant to extraction from wh-
islands than other direct objects. Nevertheless, (2.60b) gives the
impression that its ill-formedness, even though greater than that of 2.60a),
is not as strong as that of standard adjunct extraction out of a wh-island
(2.60c), but it still seems to have just the status of PP-extraction (2.60d).
(2.60) a. ?? Who did you wonder whether Sam knew?

b. ?* Who did you wonder whether the book bothered?
c. * Why1 did you wonder whether the book appealed to Sam t1?
(Johnson 1992 (25a, 26a))
d. ?* To whom did you wonder whether the book appealed t1?
(Landau 2010 (60))
Landau (2010: 29-30) argues that the kind of violation exhibited in PP-
extraction in (2.60d) is as unacceptable as extraction of genuine
Experiencer direct objects, which he takes to support his analysis of
Experiencers as arguments of null prepositions. Landau offers a syntactic
analysis of English OE verbs in which they do not take complement NPs
(or DPs) as do canonical transitive verbs, but instead select for PP
complements headed by a null preposition (Øȥ). Since objects of OE
verbs are arguments of null prepositions, extraction from within these null-
headed PPs should reveal the same degree of unacceptability as extraction
from overt-headed PP complements found with other verbs.31
Furthermore, Landau (2010) faces the controversial question of
whether class II verbs have a verbal passive. This is where the subdivision
of class II into three groups becomes vital. Hence, stative class II verbs
(and all class III verbs) generally fail to passivize since they are
unaccusative. Agentive eventive class II verbs are usual transitive verbs,
and therefore universally allow passivization. Non-agentive eventive class
II verbs, which are not unaccusative, allow pseudopassives (i.e.
prepositional passive, a form of English passive voice in which the object
of a preposition becomes the subject of a clause). However, the verbs with
the oblique nature of Experiencers are expected not to passivize unless the
language can resort to the special strategies, presented in (2.61).
31
Moreover, Landau (2010: 30-31), similarly to Grimshaw (1990: 15), analyses
some other peculiarities of OE verbs, namely their inability to form synthetic
compounds, involving a deverbal head and its object (a god-fearing man, a fun-
loving teenager, *a man-frightening god, *a parent-appalling exploit). Another
characteristic of OE verbs is their resistance to Heavy NP Shift (HNPS), analogous
to the inner object in the double object construction, as in (i)-(ii). While overtly
prepositional Experiencers, as in (iii) below, are perfectly moveable.
(i) * These things bothered yesterday the man who visited Sally.
(ii) * We told these things (yesterday) the man who visited Sally.
(iii) These things appealed yesterday to the man who visited Sally.
120 Chapter Two
(2.61) Strategies for Passivization of Quirky Objects

a. P-stranding:
The preposition that governs the object is stranded and reanalysed
with the verb – only available in languages where [V+P] reanalysis
can feed A-movement, e.g. English and Dutch.
Pseudopassive:
[TP [DP Exp]1 [T’ Aux [VP [V VPASS + Øȥ ][DP t1 ] ]]]
b. Pied-Piping: The preposition that governs the object is carried along

to the subject position – only available in languages licensing
quirky subjects, e.g. Finnish.
Quirky passive:
[TP [PP Øȥ [DP Exp]]1 [T’ Aux [VP VPASS [PP t1 ] ]]]
(Landau 2010: 48)
Only eventive (non-stative) psych-verbs can form verbal passives. In

English psych-verbs can occur felicitously with pseudopassive, as
exemplified in (2.62)
(2.62) Pseudopassives in English, e.g.

a. This bed can be slept in.
b. Mary can be relied on.
(Landau 2010: 48 (92))
Stative class II verbs, instead, are unaccusative and do not passivize,
due to the fact that they lack an external argument. However, they can
form adjectival passives. Therefore, Landau (2010) reduces the
generalization in (2.63a) to the one in (2.63b).32
(2.63) Landau’s (2010) generalization about passivization of psych-

verbs:
a. Universally, stative class II verbs do not passivize.
b. Universally, stative class II verbs are unaccusative.
(Landau 2010: 49)
32
Cf. Landau’s (2010) arguments for psych-verbs, which can passivize in Finnish.
There the aspectual distinctions are morphologically marked, so “the relevant
judgments need not appeal to subtle semantic intuitions” (Landau 2010: 54).
Consequently, making an attempt to distinguish stative from non-

stative class II verbs, the syntactic tests need to be applied. In section
2.4.2.3, some space has been devoted to discussing the most crucial
syntactic tests for stativity / non-stativity of OE psych-verbs, but an
interested reader is referred to Landau (2010) for a more detailed analysis.
In a nutshell, Landau’s generalization that the passive in class II is only
found with eventive verbs, while stative verbs are unaccusative, has been
shown to be true. Likewise, Landau’s (2010: 51) book “the more eventive
the verb, the more felicitous verbal passives are” is adequate. Indeed, the
picture concerning the passivization of psych-verbs is thus fairly intricate,
with unaccusativity and obliqueness of the Experiencer being the two
major factors governing the cross-linguistic and within a single language
variations. Finally, Landau (2005, 2010) provides an extensive overview
of syntactic properties of OE verbs in different languages of the world and
reconsiders their status from different angles. On this basis, he claims that
Experiencers behave like locatives. Consequently, Experiencers should
share many of the properties of true locatives, which is manifested through
the following syntactic properties: first, all object Experiencers are oblique
(or dative); second, Experiencers are LF–subjects (Landau 2005: 5). In
Landau’s theory, the special psych properties are linked to the presence of
a (possibly null) locative preposition with a [loc] feature, which licences
the OE. However, raising Experiencer objects in class II verbs to the
subject position is a case of locative inversion. For Landau (2010), all
Experiencers become LF-subjects, namely they end up in [Spec, TP].
2.5.3 Fábregas and Marín’s (2015) layer theory

Fábregas and Marín (2015: 231), in their recent paper, argue that all
formal psych-verbs comprise a core. The core refers to a mental state
which relates the Experiencer with the entity towards which this state is
targeted. While the core matches up the structure of SE psych-verbs,
shown in (2.64a), OE psych-verbs are built over this core. The OE
structure results from adding another layer codifying causation, but
without any process (dynamic part) contained within the event structure,
as illustrated in (2.64b).33
33
In fact, it was Pesetsky’s (1995) original claim that OE psych-verbs subsume SE
psych-verbs.
122 Chapter Two
(2.64)
(Fábregas and Marín 2015: 231)34
Consequently, Fábregas and Marín (2015: 227) maintain that SE psych-

verbs denote individual level (IL) states, i.e. states without boundaries,
whereas OE psych-verbs should be classified as states of the stage level (SL)
class, since they denote states with an onset.35 Similarly, Fábregas and Marín
(2015: 234, 265) propose that Experiencers in SE psych-verbs can also be
viewed in terms of a PP, as in (2.65), following Landau’s (2010)
representation, or in their revised version, as in (2.66) below.
(2.65)
(Fábregas and Marín 2015: 234, 265)
34
I would like to thank Prof. Henryk Kardela for suggesting the term “cognitive-
semantic-syntactic trees” for the syntactic structures of SE and OE psych-verbs
offered by Fábregas and Marín (2015). In their structure the notion of state
belongs to event structure. This type of event structure is not mentioned either in
Landau’s (2010) or Harwood et al.’s (2017) syntactic structures. In addition, the
trees by Fábregas and Marín (2015) refer both to Dowty (1979) and his semantic
concepts of State, or Causer, and to Jackendoff (1990) and his cognitive-semantic
decomposition analysis. In turn, the trees offered by Koopman (1991, 2010) for
verb particle constructions (cf. (4.69)) are clear syntactic structures.
35
Class I (SE) verbs comprise individual-level and stage-level predicates (love vs.
worry), and the latter are often related to inchoative or reflexive morphology,
which is likely to have an agentive interpretation (cf. Pesetsky 1995; Reinhart
2002; and Pylkkänen 2000).
(2.66) The structure of SE psych-verbs like temer “fear”:
Juan teme a María.

Juan fears ACC María.
Juan fears María.
(Fábregas and Marín 2015: 259)
In a nutshell, in (2.66), the tree diagram depicting the structure of SE

psych-verbs like temer “fear” is presented. The structure is a core of a
formal psych structure, viz. a state denoting an emotion which relates an
Experiencer with the target of that emotion (cf. Ramchand 2008: 55-56).
Nonetheless, both Landau’s (2010) and Fábregas and Marín’s (2015)
assumptions about the locative nature of SEs have not be confirmed by
sufficient evidence so far.
2.5.4 Grafmiller’s (2013) account of psych-verbs

In Grafmiller’s (2013) approach, the explanation for the untypical
behaviour of some OE psych-verbs is mainly semantic in nature, and is
said to stem from the ways in which humans conceptualize psychological
events and processes.36 Assuming initially that the special behaviour of
36
Grafmiller (2013) ascertains that providing further insight into the relationship
between language and emotion concepts lies at the heart of his research. To
understand this relationship, he explores how the conceptual properties of
emotions are encoded in the words and constructions used to describe them.
Assuming that human beings build mental concepts of various kinds which reflect
their experience, Grafmiller (2013: 1-2) notes that many of these concepts are
encoded in the meanings of individual words, often conceived of as “entries” in a
mental lexicon (e.g. Jackendoff 1989; and Pustejovsky 1995). Lexical entries
comprise varying degrees of semantic information, conceived of as sets of
privative features, thematic role lists, and/or event structures, and that words are
124 Chapter Two
psych-verbs is obtained only in their stative and/or more controversially

non-agentive readings; Grafmiller (2013) proves, on the ground of the
corpus data, that eventive and stative uses are available to all OE
predicates in both the active and passive.
The purpose of his study was to investigate the acceptability of the two
classes of OE verbs in agentive constructions, namely (i) Agentive-OE
verbs, which allow agentive interpretations; and (ii) Non-agentive-OE
verbs, which are stated to disallow agentive interpretations. Following
Verhoeven (2010), Grafmiller (2013) added, beside the two classes of OE
verbs, two more control groups to the study, i.e. physical transitive verbs
and transitive SE psych-verbs, which are unambiguously dynamic and
compatible with agentive interpretations. SE predicates, in turn, are
typically identified as stative and inherently non-volitional (Grafmiller
2013: 243-244). The verbs used in the research are given in (2.67).
(2.67) Verbs used in Grafmiller’s (2013) research:

a. Agentive-OE verbs:
amuse, anger, annoy, bother, disturb, frighten, irritate, scare,
surprise, upset
b. Non-agentive-OE verbs:
amaze, astonish, bore, captivate, concern, depress, fascinate,
horrify, offend, please
c. SE verbs:
admire, adore, despise, detest, enjoy, fear, hate, like, love,
loathe
d. Transitive verbs:
help, hug, kick, pinch, shove
(Grafmiller 2013: 244)
In his research, Grafmiller (2013) attempts to address the issues of

agentivity. Based on speaker judgments, he draws the conclusion that the
twenty OE verbs he examined do not form clearly agentive and non-
agentive sub-classes. Considering the results from the corpus data and the
judgment studies, there seems to be, according to Grafmiller (2013), a
serious doubt cast on proposals of making a lexicalized distinction
individuated conceptually in terms of the information that they denote. With regard
to verbs, Grafmiller (2013: 2) argues that in a wide variety of theoretical
approaches a verb’s semantic representations determine the range of syntactic
realizations of its arguments (e.g. Dowty 1991; Jackendoff 1990; Langacker 1987;
Pinker 1989; Rappaport Hovav and Levin 1998; van Valin 1990; and van Valin
and LaPolla 1997; among others).
between OE psych-verbs that are definitely non-agentive (e.g. amaze,

concern, depress) and those that are more stretchable to be used with
either agentive or non-agentive subjects (e.g. amuse, frighten, surprise)
(Grafmiller 2013: 258). In the results of the acceptability judgment
surveys, depress and concern are the two verbs revealing a strong
unacceptability in agentive contexts; whereas, amuse clearly has a
tendency to be regarded as a deliberately caused emotion. The other
remaining predicates under scrutiny show no strong bias one way or the
other, with some verbs (fascinate, frighten and startle) showing
considerable variability in subjects’ assessment. In fact, most verbs do
readily appear in at least some agentive contexts (e.g. used as imperatives,
or modified by deliberately, intentionally, etc.) (Grafmiller 2013: 258).
As suggested by Grafmiller (2013), the agentive OE psych-verbs
cannot be easily separated from non-agentive ones. Instead, all OE psych-
predicates fall along a continuum in accordance with the relative potential
of being used with an agent. Regrettably, Grafmiller’s (2013) study lacks a
sufficient explanation for both how this continuous distribution compares
to that of non-psych verbs and the indication of the verbs’ potential for
eventiveness. Nonetheless, the Corpus and judgment studies carried out by
Grafmiller (2013) challenge the claims regarding sub-classes of OE verbs,
made so far by linguists. The “rare” examples of the stative verb depress
used in the progressive, and the agentive adverbial purposely used with the
stative verb bore, which are included in Corpora, are exemplified in
(2.68):
(2.68) a. The human race is constantly depressing me...

b. I’m going to purposely bore you with this tip, but it TOTALLY
WORKS.
(Google, Grafmiller 2013: 114)
Moreover, in his analysis of psych characteristics of OE verbs,

Grafmiller (2013) finds some evidence opposed to Landau (2010: 30-31)
and Grimshaw (1990: 15), who point out the inability of OE verbs to form
synthetic compounds, and their resistance to Heavy NP Shift (HNPS). In
the web corpora Grafmiller (2013) finds some acceptable examples of
shifted Experiencer objects no worse than shifted examples of other kinds
of affected objects. However, he leaves the case open, concluding that
whatever differences in the judgments of these sentences there may be,
they are far too subtle to draw strong conclusions about (potentially
covert) aspects of their syntactic structure (Grafmiller 2013: 69-71).
126 Chapter Two
All in all, the above-mentioned facts argue against analysing

differences in agentivity among psych-verbs at the level of lexical
semantic structure. Instead, Grafmiller (2013) proposes treating agentivity
as an inference arising from the total integration of semantic, syntactic,
and contextual information in the clause (ibid.: v). Importantly, for
Grafmiller (2013), stative OE verbs are not different from regular
transitive (causative) verbs.

The aims of Chapter Two were: (i) to define psych-verbs; (ii) to provide
their typology and characteristics; (iii) to introduce the Lexicon-Syntax
Interface; and (iv) to outline the most prominent approaches to psych-
verbs, and syntactic properties of psych-predicates. To recapitulate, it
should be pointed out, after Klein and Kutscher (2005: 1-2), that psych
predicates, representing a wide range of construction types, challenge the
assumption that verbs with the same theta-grid and event structure would
select the same case patterns. However, it is difficult to find general
linking patterns of psych-verbs. On the other hand, it is characteristic of
psych-verbs to have non-psych-readings as well, which, indeed, may be
responsible for determining case selection.
Unquestionably, beside Belletti and Rizzi’s (1988) theory, there have
been far more approaches to the syntax of psychological predicates. In
general, the thematic roles of psych-verbs are usually assigned to an
Experiencer and stimulus / Theme. Arad (1998, 1999) has argued against
Belletti and Rizzi’s (1988) idea of unaccusativity of OE psych-verbs,
highlighting the ambiguity these verbs give rise to between: stative,
causative eventive (non-agentive), and agentive readings. Landau (2010)
has offered a localist approach to Experiencers. In Fábregas and Marín’s
(2015) layer theory, individual level and stage level psych-verbs have been
distinguished. Finally, Grafmiller (2013) has discussed the issue of
agentivity of OE psych-verbs, and recognised stative OE predicates as
similar to regular transitive (causative) verbs.
Nevertheless, what all of the views have in common is causality that
appears as a relevant concept. The prominence of causation is
unquestionable, although its status differs, since some linguists treat it as a
thematic notion, and others as an aspectual one. The most problematic
class, investigated so far in the literature, are OE verbs. The reason for that
is their special structure and some grammatical rules that they appear to
violate (the most notable of which is the UTAH).
From this perspective and in agreement with Arad (1998), Alexiadou

and Iordachioaia’s (2014) standpoint is adopted, for the purpose of this
book. It is assumed that what makes psych verbs special is their aspectual
ambiguity, rather than their Experiencer argument. Indeed, it is the
diversity of aspectual readings in which SE and OE psych predicates can
occur that makes these predicates distinctive.
CHAPTER THREE
IN SEARCH OF PSYCHOLOGICAL IDIOMS:

A CORPUS-BASED ACCOUNT
3.1 Introduction
The aims of Chapter Three are twofold. On the one hand, we strive to
select psych-verbs, which are a representative set of this type of
predicates. On the other, we intend to determine idioms which are to be
analysed in Chapter Four, and which correspond to the psych predicates
listed earlier. Besides, the methodology adopted in data selection is also
explained in Chapter Three. Two stages of the corpus study need to be
conducted to meet the objectives just stipulated. The first stage involves
eliciting the most frequent class I and III psych-verbs (cf. Belletti and
Rizzi 1988), while class II top psych-verbs are adopted after Grafmiller
(2013). The second stage is based on searching for idiomatic phrases
corresponding to the psychological verbs selected.
The guidelines adopted in data selection in the first stage of the study
are as follows: (i) the psych-verbs to be elicited correspond to the
psychological predicates in Belletti and Rizzi’s (1988) tripartite typology
of psych-verbs, described in detail in Chapter Two of the book; (ii) only
the most frequent psych-verbs, with the top occurrence in The Corpus of
Contemporary American English (COCA), are taken into account; and (iii)
the psych-verbs to be selected belong to various emotion domains.
Once the set of psych-verbs is determined, the second stage of the
study is embarked upon. It aims at investigating any possible
phraseological units, collected from English dictionaries and thesauri,
which correspond to simple psych-verbs. Not only can these idiomatic
phrases be used as substitutes for simple psych-verbs, enriching our
language inventory, but also these idioms will be compared with their
basic psych-verb counterparts in Chapter Four of the book, as regards their
syntactic and aspectual characteristics. As a result, the research is expected
to provide some insight into the complex nature of psychological
predicates.
130 Chapter Three
The chapter is organised into four sections. In section 3.2 the three
above-mentioned criteria of selecting the basic psych-verbs are discussed
in detail. Here, the focus is laid on determining the most frequent class I
(SE) and class III (OE) psych-verbs, based on the corpus analysis. In
section 3.3, the purpose of the research into psych-idioms, the data
collection method, and the methodology adopted in the corpus study are
outlined. Section 3.4 reviews the data analysed and offers some discussion
related to the results of the research. The chapter ends with a brief
summary and provides a scope of further investigation, carried out in
Chapter Four of the book.
3.2 Towards the specification of the selection criteria

As has just been mentioned, the choice of psych-verbs in the first stage of
the study is based on the following criteria: (i) the predicates belong to
Belletti and Rizzi’s (1988) classes of psych-verbs; (ii) they are the top
frequent psych-verbs in the COCA Corpus; and (iii) these psych-verbs
represent different emotion domains. In sections 3.2.1-3.2.3, each of these
criteria is to be elaborated on separately.
3.2.1 Belletti and Rizzi’s (1988) tripartite classification

of psych-verbs
Even though psych-verbs have already been defined and discussed in the
preceding chapter, some crucial points concerning those predicates are
repeated here, for the sake of convenience. On the basis of the well-known
studies represented by Belletti and Rizzi (1988), Pesetsky (1995), and
Landau (2010), among others, it can be specified that psych-verbs express
(a change in) mental or/and emotional state and a relation between the two
arguments: an Experiencer and the Cause / Theme of such a psychological
condition. An Experiencer can be realized as either a subject or as an
object, following Belletti and Rizzi’s (1988) tripartite classification, as
introduced at the very beginning of this book, and illustrated in (2.3) in
Chapter Two, but repeated in (3.1) below for the sake of convenience.
(3.1) Belletti and Rizzi’s (1988) tripartite classification of psych-verbs:
Class I: Mark loves bats. (SE psych-verbs)

Class II: The bats frightened Mark. (OE psych-verbs)
Class III: This film appeals to Joanne. (OE psych-verbs)
In Search of Psychological Idioms: A Corpus-based Account 131
Generally speaking, as can be seen in (3.1), class I involves regular

transitive (agentive) constructions with an Experiencer subject and a
Theme object. Class II is causative in nature with a Causer in the subject
position and an Experiencer in the object position. Class III, in turn,
comprises unaccusative constructions. The basic distinction between class
II and III, provided by Belletti and Rizzi (1988) for Italian and some
languages, refers to the accusative (class II) and dative (class III) case of
an Experiencer object. For English, the difference between class II and III
lies only in their aspectual properties. While class III verbs are stative,
and, consequently, they can never be used agentively, most class II verbs
are ambiguous between eventive (agentive vs. non-agentive) and stative.
Notably, the peculiar behaviour of OE verbs obtains only in the stative
uses (cf. section 2.4.2 in Chapter Two; and cf. Arad 1998, 1999; among
others).
To recall briefly, as argued by Arad (1998), the difference between the
eventive and the stative interpretations is that whereas eventive OE verbs
involve a change of state in the Experiencer, there is no change of state on
the stative reading. The stative reading corresponds to the condition
experienced by the Experiencer that causes him to be in a particular mental
state. Therefore, the verb frighten in the sentence “The bats frightened
Mark” can be interpreted as an eventive predicate if the bats did something
that caused Mark to suddenly be frightened. If Mark, instead, gets
frightened with the idea or the presence of bats in general, then the
sentence has a stative reading.
In short, this three-class typology is to serve as a basis for the selection
of psych-verbs, for which idiomatic expressions are to be provided in the
further part of this chapter.
3.2.2 Top frequent psych-verbs in the COCA Corpus
Due to the fact that certain psychological states are potentially more
common than others, it is expected that the frequency of usage of these
psych-verbs is far higher than the frequency of other emotion / mental
verbs in given corpora. For instance, everyone can recognise the priority
of the state of “love “ over “charm “ or “awe “ with the help of some
basic intuition. Moreover, corpus linguists suggest that “the frequency
distribution of tokens and types of linguistic phenomena in corpora have
(..) some kind of significance. Essentially, more frequently occurring
structures are believed to hold a more prominent place, not only in actual
discourse but also in the linguistic system, than those occurring less often”
(Schmid 2010: 101). Finally, assuming that the essential requirement for
132 Chapter Three
any quantitative analysis is to have the sufficient amount of data to be

meaningful, only those verbs on the higher end of the frequency
distribution of tokens are primarily chosen for any further analysis.1
Therefore, all things considered, it seems justified to treat frequency as
a leading factor in the data search conducted here. Following Grafmiller
(2013: 167), I assume that especially those psych-verbs which are most
commonly encountered in everyday inventory are worth being
investigated more thoroughly.
3.2.2.1 The most frequent class II OE psych-verbs in the COCA

Corpus (Grafmiller 2013)
Recent work by Grafmiller (2013) attempts to address the issue of the

most frequent psych-verbs, with a focus put on OE predicates, belonging
to class II of Belletti and Rizzi’s (1988) typology. There is no need to
repeat the same kind of study; therefore, I adopt the methodology and
results obtained in the first stage of his research (concerning corpus
frequency), and take them as the starting point for my analysis.
Grafmiller (2013) sampled the data for his study from the written and
spoken sections of the COCA, updated up to autumn 2012. The dataset
included 400 tokens, randomly extracted from the COCA. The tokens
covered the 16 OE verbs listed in (3.2).
(3.2) amaze, amuse, anger, annoy, astonish, captivate, concern, depress,

fascinate, frighten, horrify, please, scare, startle, surprise, upset
Grafmiller (2013: 167) justifies his decision concerning the selection

of those specific predicates with: (i) their popularity in the literature; (ii)
their easier and more common reference to these specific senses than to
others; and (iii) their high frequency distribution in the COCA Corpus.
1
Interestingly, some specialists in modern English lexicology (Ginzburg et al.
2004; among others) argue that “frequency value of the word is as a rule the most
reliable and objective factor indicating the relative value of the word in the
language in general and conditioning the grammatical and lexical valency of the
word” (Ginzburg et al. 2004: 180). Besides, the frequency value of the word can
be in many cases “sufficient to judge of its structural, stylistic, semantic and
etymological peculiarities, i e. if the word has a high frequency of occurrence one
may suppose that it is monomorphic, simple, polysemantic and stylistically neutral.
Etymologically it is likely to be native or to belong to early borrowings” (ibid.:
180).
Thanks to his originally constructed project based on a Python script,

Grafmiller (2013) was able to filter class II OE predicates, regardless of
the extremely high number in which they occur in the Corpus. Having
elicited the proper predicates automatically, Grafmiller (2013) managed,
using a manual filter, to remove the tokens with either non-psychological
readings (e.g. Before he depressed the button) or non-verbal uses. Among
the psych-verbs left aside due to their polysemous meanings, there were
bore, bother, and worry. He excluded also those instances in which both
the Stimulus and the Experiencer arguments were not overtly realized in
the sentence. This resulted in eliminating from the dataset agentless
passive sentences (my mother found the scrapbook, and she was just
horrified), active sentences with null objects (More than 20 years and
some restoration later, the necklace still astonishes with the bold
assurance of its design), middle constructions (But the Padres are Alfred
Hitchcock. They don’t scare easily), and psych-verbs with particles (e.g.
frighten away / off) (Grafmiller 2013: 168). In this way, he elicited 16
most frequent class II OE psych-verbs, which are listed in (3.2) above.
For the purpose of my study, dedicated, first and foremost, to idiomatic
phrases which have the meaning equivalent to a given psych predicate,
only nine most frequent psych-verbs are taken from Grafmiller’s (2013)
ranking. They are presented in (3.3), along with the number of tokens
which were found in the corpus after the filtering described above.
(3.3) Most frequent class II OE psych-verbs

(with the number of tokens after the filtering):
a. surprise 389
b. annoy 366
c. fascinate 285
d. amuse 283
e. scare 272
f. depress 210
g. anger 207
h. horrify 159
i. concern 137
Grafmiller (2013) lists seven more top psych-verbs, i.e. captivate (313)
and amaze (268), which I have left aside since they are synonymous with
the verb fascinate, listed in (3.3). The verb frighten (202) is synonymous
with scare; thus, it is omitted as well. Verbs, astonish (169), startle (133),
please (130), and upset (121) are not analysed either, because of their
134 Chapter Three
lower frequency in the Corpus, and because of space limitations of the

book.
The verbs in (3.3) are adopted as the initial dataset for my research on
idiomatic phrases, carried out in section 3.3 of this chapter. They represent
various emotion domains, i.e. happiness, sadness, anger, fear, and disgust;
however, these are only representatives of class II, which makes the list
incomplete from the point of view of Belletti and Rizzi’s (1988) tripartite
classification. As a result, the list must be extended to cover class I, as
well as, class III psych predicates.
3.2.2.2 The most frequent class I (SE) and class III (OE) psych-verbs
in the COCA Corpus
In order to elicit the representatives from all Belletti and Rizzi’s (1988)
classes of psych-verbs, Grafmiller’s (2013) results obtained for the most
common class II OE psych-verbs must be extended to cover both class I
(SE) and class III (OE) psych-verbs. Therefore, class I and III psych
predicates have become the material of my study.
With the aim to select the most frequent members of class I and class
III psych-verbs which appear in the COCA Corpus, I have searched this
online corpus with respect to the occurrence of the predicates under
scrutiny. For the purpose of this research, the list of psych-verbs offered by
Levin (1993) has been adopted as the dataset. Class II psych-verbs,
represented in Levin’s typology by Amuse Verbs (220 members, e.g. amuse,
bother, concern, frighten, scare), have been excluded in order not to repeat
the study, carried out by Grafmiller (2013) (cf. section 3.2.2.1 above).
In comparison with the variety of verb taxonomies offered in the
literature, it is Levin’s (1993) classification which is considered to be far
more sophisticated and thoroughly researched.2 Indeed, in her seminal
work, Levin (1993) categorised over 3,000 English verbs, along with their
shared meaning and behaviour, which makes her classification one of the
most extensive ones. All of the over 3,000 English verbs are put into 57
semantically coherent classes and around 200 subclasses.3 In her
2
There have been numerous attempts to classify English verbs with various
differentiating factors, e.g. argument structure and semantic characteristics, event
structure or thematic role structure (cf. Vendler 1967; Croft 1991; Fillmore and
Baker 2001; and Kipper-Schuler 2005).
3
Those include, e.g. Verbs of Putting; Verbs of Removing; Verbs of Sending and
Carrying; Verbs of Exerting Force: Push / Pull Verbs; Verbs of Change of
Possession; Learn Verbs; Hold and Keep Verbs; Verbs of Concealment; Verbs of
Throwing; Verbs of Contact by Impact; Verbs of Contact: Touch Verbs; Verbs of
taxonomy, verbs which display the same or a similar set of diathesis

alternations in the realization of their argument structure are expected to
share certain meaning components. Levin (1993) identifies verb classes
according to the alternations they can appear in,4 i.e. the causative /
inchoative alternation, the middle alternation; the instrument subject
alternation; the with / against alternation; the conative alternation; the
body-part possessor ascension alternation; the unintentional interpretation
available; and the resultative phrase.5 Accordingly, all the members of the
Psych-Verb class, listed by Levin (1993: 188-195) within classes (31.1)-
(31.4) and (32),6 have become the material for my corpus study to elicit
Cutting; Verbs of Combining and Attaching; Verbs of Separating and

Disassembling; Image Creation Verbs; Verbs of Creation and Transformation;
Engender Verbs; Calve Verbs; Verbs with Predicative Complements, Verbs of
Perception; Psych-Verbs (Verbs of Psychological State); Verbs of Desire; and
Judgment Verbs.
4
For instance, Levin’s (1993) class of Break Verbs (class 45.1), which covers
verbs that refer to actions that bring about a change in the material integrity of
some entity, is categorized by its participation (1-3) or non-participation (4-6) in
the following alternations and other constructions (7-8):
1. Causative/inchoative alternation: Tony broke the window. The window broke.
2. Middle alternation: Tony broke the window. The window broke easily.
3. Instrument subject alternation: Tony broke the window with the hammer. The
hammer broke the window.
4. *With/against alternation: Tony broke the cup against the wall. *Tony broke the
wall with the cup.
5. *Conative alternation: Tony broke the window. *Tony broke at the window.
6. *Body-Part possessor ascension alternation: *Tony broke herself on the arm.
Tony broke her arm.
7. Unintentional interpretation available (some verbs): Reflexive object: *Tony
broke himself.
Body-part object: Tony broke his finger.
8. Resultative phrase: Tony broke the piggy bank open. Tony broke the glass to
pieces.
5
Additional properties connected with subcategorization, morphology and
extended meanings of verbs are also taken into consideration in the
aforementioned taxonomy.
6
Levin’s (1993) classification of 374 English psych-verbs is as follows:
(i) Amuse Verbs, e.g. amuse, bother, concern, frighten, scare (220
members),
(ii) Admire Verbs, e.g. admire, enjoy, love, envy, regret (45 members),
(iii) Marvel Verbs, e.g. mind about, worry about, marvel at, fear for (79
members),
(iv) Appeal Verbs, e.g. niggle, grate, jar, appeal, matter (5 members),
(v) Verbs of Desire, e.g. desire, fancy, need, lust, thirst (25 members).
136 Chapter Three
those most frequent psych-verbs which represent class I (SE) and class III
(OE) psych-verbs in Belletti and Rizzi’s (1988) syntactically-oriented
typology.
When Levin’s (1993) group of psych-verbs is organized in accordance
with Belletti and Rizzi’s (1988) typology, the classification of psych-verbs
looks as in (3.4).
(3.4) Levin’s (1993) set of 374 psych-verbs within Belletti and Rizzi’s
(1988) tripartite syntactically-oriented typology:
a. class I (SE) psych-verbs: 149 members

Admire Verbs, e.g. admire, enjoy, love, envy, regret
(45 members),
Marvel Verbs, e.g. mind about, worry about, marvel at
(79 members),
Verbs of Desire, e.g. desire, fancy, need, lust, thirst
(25 members).
b. class II (OE) psych-verbs: 220 members

Amuse Verbs, e.g. amuse, bother, concern, frighten, scare
(220 members),
c. class III (OE) psych-verbs: 5 members

Appeal Verbs, e.g. niggle, grate, jar, appeal, matter
(5 members).
As shown in (3.4), class I (SE) psych-verbs are represented by Admire

Verbs (45 members), Marvel Verbs (79 members), and Verbs of Desire
(25 members); class II (OE) consists of the most numerous group of
Amuse Verbs (220 members); while class III comprises 5 members of
Appeal verbs. Having excluded class II members, a sample of 154 psych-
verbs (149 members from class I and 5 verbs from class III), out of the
total number of 374 psych-verbs, has been examined in my study, in terms
of their top frequency usage in the COCA Corpus.
The COCA Corpus7 has been chosen as a search tool used for the
analysis since it is generally considered one of the most recent corpora
with its data evenly divided between the five genres of spoken, fiction,
popular magazines, newspapers, and academic journals. Besides, the
7
http://corpus.byu.edu/coca/ and http://www.wordandphrase.info/frequency
List.asp
COCA Corpus is composed of more than 520 million words in 220,225

texts, including 20 million words each year from 1990 to the version
updated in May 2016.
To make my investigation of class I and class III psych-verbs
comparable to Grafmiller’s (2013) study of class II verbs, an analogous
methodology has been used in my research. Thus, similarly to class II, the
selection of class I and III predicates is conditioned by (i) their popularity
in the literature; (ii) their easier and more common reference to these
specific senses than to others; and (iii) their high frequency distribution in
the COCA Corpus. The only difference, which must inevitably appear
between Grafmiller’s (2013) and my corpus research, concerns the
technique to elicit the most frequent psych-verbs. Grafmiller used an
original program, created by himself and written in a Python script, in
order to select the specific sentences with psych predicates. His study
aimed at eliciting the actual language data based on argument structures of
the tokens. My research, on the other hand, only focuses on selecting top
frequent psych-verbs as the starting point for my further research
concerning the idiomatic equivalents of these psych-verbs. Therefore, for
the purpose of my study, I have taken advantage of both the current
interface of the COCA Corpus (updated in May 2016, while Grafmiller’s
data covered the period up to 2012), and the frequency list ready-made for
the first 5,000 most popular lemmas / words in the COCA Corpus.
Accordingly, first, the COCA has been checked via the search tool to
measure roughly how often the predicates analysed have occurred in oral
and written sentences recently. Next, the set of verbs just examined in
relation to their unfiltered frequency of occurrence has been compared
against the Frequency List composed for the whole corpus, with some
overlaps expected. Meantime, following Grafmiller’s (2013) methodology,
each of the 154 predicates under scrutiny (class I and III psych-verbs) has
been filtered to choose their verbal uses only. Simultaneously, the psych-
verbs under scrutiny have been manually sifted to remove those instances
involving non-psychological readings. To be precise, if a verb lacks a
psychological meaning (e.g. The cards were impressed with a halfpenny
stamp), the instance has been disregarded. Moreover, on account of
polysemous meanings of certain psych-verbs, some of them have been left
aside. Among these verbs, there were, e.g. miss, engage, and care, which
have connotations with some physical activity more common than (or
equal to) their psychological readings. The verb miss, according to the
Online International Dictionary,8 means “to fail to hit, reach, or come into
8
Online International Dictionary available at http://idict.org, retrieved 5/12/2016.
138 Chapter Three
contact with (something aimed at): a laser-guided bomb had missed its
target; he shot twice at the cashier, but missed both times.” The verb
engage is defined as “to arrange to employ or hire (someone): he was
engaged as a trainee copywriter;” whereas the verb care means “look after
and provide for the needs of someone.” Undoubtedly, these verbs can refer
to a psychological state, i.e. miss “fail to notice someone,” engage
“attract someone’s interest,” and care “feel concern or interest” but these
are not their sole and main readings, in contradistinction to the other verbs
from the top group under scrutiny. As a result, it would hardly be possible
to discern manually a psychological reading from the non-psychological
one, and present accurate frequency details afterwards. In short, any psych
predicate with an ambiguous reading from the list of 154 items tested has
been excluded from my further research.
Indeed, the search has yielded quite a big number of psych-verbs with
a high frequency of occurrence in the COCA Corpus. It has been found
out that in the dataset of 154 psych-verbs, there are 13 verbs with over
10,000-token-occurrence, excluding those with ambiguous meanings. Two
of them represent class III psych predicates, i.e. matter and appeal;
whereas eleven belong to class I verbs in Belletti and Rizzi’s (1988)
typology, as illustrated in (3.5) below.
(3.5) 13 psych-verbs of class I and III with the highest frequency in the
COCA:
a.class I (SE) psych-verbs: want, need, like, love, enjoy, worry

(about), hate, fear, appreciate, trust, respect
b.class III (OE) psych-verbs: appeal to, matter to
The top 13 psych predicates, listed in (3.5), occur in the COCA Corpus
with the frequency over 10,000. They are also, as expected, included in the
COCA “5,000 frequency word list,” viz. the first 5,000 most frequent
words / lemmas in the whole COCA Corpus.9 Interestingly, Nation (2006),
9
Schmitt and Schmitt (2014) state that for English, “high-frequency vocabulary”
has traditionally been understood as around the first 2,000 most frequent word
families. On the other hand, low-frequency vocabulary has been characterized in
various ways, ranging from anything beyond 2,000 word families all the way up to
all of the word families beyond the 10,000 frequency level. Any frequency list is
accurate provided the corpus (collection of texts) that it is based on is a reliable
source as well. The 450 million word COCA Corpus, which has been chosen as the
source corpus for this study, also contains its own frequency data of English
and Schmitt and Schmitt (2014), among others, take the number 2,000 as
the borderline for high-frequency vocabulary. Thus, psych-verbs ranked
from 1 up to 2,000 can be treated as representatives of “high-frequency
vocabulary” in the language, while those below the 2,000th position are
supposed to be representatives of mid-frequency or low-frequency
vocabulary (Schmitt and Schmitt 2014: 501). Table 3-1 below presents the
ranking position and the exact frequency rates on the COCA frequency
word list for the 13 psych-verbs from class I and III.
Table 3-1. The ranking position of the top 13 psych-verbs on the COCA 5,000
frequency word list10 (above the number of 10,000, ordered from highest to lowest)
No Rank Verb Frequency

1 83 want 514,972
2 132 need 276,744
3 208 like 182,341
4 391 love 103,681
5 884 enjoy 44,020
6 973 worry 40,210
7 1535 hate 24,921
8 1670 fear 21,333
9 1751 appreciate 20,806
10 1763 matter 20,534
11 1855 trust 19,482
12 2836 respect 11,083
13 2927 appeal 10,745
Unquestionably, the 13 psychological predicates, listed in (3.5) and

presented in Table 3-1, belong to the first most frequent 3,000 words of all
words and lemmas in the whole language. The highest ranking position is
associated with the verb want, which is placed in the 83th position of the
common words used in English and collected in the COCA Corpus. The
next verb in the list is need, occupying the 132th position, and like is
(available at http://www.wordfrequency.info/intro.asp) in a number of different

formats (e.g. 100,000 and 60,000 word lists, and a comparison of the two lists).
The list used for the purpose of this study is the one commonly available online at
http://www.wordfrequency.info/top5000. asp and called the “5,000 frequency
word list.”
10
Available at http://www.wordfrequency.info/top5000.asp, last retrieved
5/01/2017.
140 Chapter Three
ranked third in the set of 13 top psych-verbs. Then come the verbs love,
enjoy and worry, which occupy the ranking positions up to the first top
1,000 words. Next, the verbs such as hate, fear, appreciate, matter, and
trust belong to the period between 1,000 and 2,000 of the most frequent
words. Finally, the bottom of the list is covered by the verbs respect and
appeal, which are in the 2836th and 2927th position, respectively. The
frequency of all of the 13 psych-verbs ranges from 514,972 highest to
10,745 lowest in the whole COCA Corpus. Indeed, being so frequently
used in the language inventory, these are the predicates which deserve to
be selected as the starting point for the further stage of the research,
devoted to idiomatic phrases and carried out in section 3.3 of this chapter.
Obviously, since the first top 13 psych verbs have been extracted from the
initial set of 154 class I and III psych-verbs, the remaining 141 verbs have
been classified as less frequent, thus they are outside of our interest and
are left aside here.
In addition, 9 verbs from class II, adopted from Grafmiller’s (2013)
study, have been checked against the current COCA 5,000 frequency word
list. The results obtained are presented in Table 3-2. However, it should be
borne in mind that the number of occurrence of the tokens given by
Grafmiller (2013) was obtained by means of his original Python
programming and thicker filters he had applied to choose the best
examples of psych-verbs and their argument structures.
On the basis of the results presented in Table 3-2, it can be observed
that some of the psych-verbs examined by Grafmiller (2013) are not
included in the list of the first 5,000 most frequent words, updated in 2016.
Only the verbs surprise, scare, and concern can be found between 3,000
and 4,000 in the ranking position. Nonetheless, without shadow of a doubt,
exactly these predicates are commonly referred to in the linguistic
literature, therefore they are highly eligible for further analysis.
Table 3-2. Top class II OE psych-verbs from Grafmiller’s (2013) study and their
ranking position checked against the COCA 5,000 frequency word list11 (with the
number of tokens after Grafmiller’s filtering)
Class II psych-verbs
No Rank with the number of tokens Frequency
(Grafmiller 2013)
1. 3086 surprise 389 10023
2. <5000 annoy 366 >10000
3. <5000 fascinate 285 >10000
4. <5000 amuse 283 >10000
5. 3844 scare 272 7367
6. <5000 depress 210 >10000
7. <5000 anger 207 >10000
8. <5000 horrify 159 >10000
9. 3439 concern 137 8702
Finally, the list of verbs most popular in the literature and in the COCA
Corpus relating to emotional and psychological conditions, is expected to
comprise the complete section of all emotional states, both negative and
positive ones. Thus, the set of 9 class II psych-verbs from Grafmiller’s
(2013) work, and 11 class I and 2 class III psych-verbs from my study (cf.
Table 3-1), elicited so far, should be filtered for the final selection, to
comply with the 5 basic emotion domains referred to in the literature
cross-linguistically.
3.2.3 The top psych-verbs vs. basic emotion domains
Any attempt to show “precise equivalence of the emotion concepts in the

different cultures” is a real challenge (Russell 1991: 435). However, there
are, universally accepted, five basic emotion domains from which
translation-equivalent psych-verbs can emerge (cf. Johnson-Laird and
Oatley 1989; Ekman 1992, 1999; Turner 1999, 2007; Wierzbicka 1992,
1999, 2009; Verhoeven 2010; and Rott and Verhoeven 2016; among
others). According to Johnson-Laird and Oatley (1989), the set of basic
emotion modes they distinguish are encoded in English with the words
happiness, sadness, fear, anger, and disgust, and indeed, they should be
universally accepted as discriminable categories of direct experience.
11
Available at http://www.wordfrequency.info/top5000.asp, last retrieved
5/01/2017.
142 Chapter Three
“Basic emotion signals have no internal structure that is parsed and

interpreted within the system” (Johnson-Laird and Oatley 1989: 90).12
Furthermore, for Johnson-Laird and Oatley (1989), there are some
emotion verbs which exist in connection with the universally
acknowledged five basic emotions. These verbs occur only as emotional
relations and causatives, which is consistent with the fact that psych-verbs
are recognised in the linguistic literature as dyadic relational and causative
verbs. According to theories of argument structure, the former, i.e. the
verbs in the emotional relation class, occur in the syntactic structures in
which the Experiencer is the subject (SE verbs); the latter, i.e. causative
verbs, occur in structures in which the Experiencer is the grammatical
object (OE verbs, including class III OE psych-verbs) (Levin and
Rappaport 2005; Croft 2012; Rozwadowska and Willim 2016).
Moreover, treating emotions as discrete, measurable, and
physiologically distinct (cf. Handel 2011; Shaver et al. 1992), Ekman
(1992, 1999) follows Johnson-Laird and Oatley’s (1989) typology, and
names six emotions as basic ones, i.e. anger, disgust, fear, happiness,
sadness, and surprise. Thus, it is surprise which is added to the list of
basic emotions offered by Johnson-Laird and Oatley (1989). Ekman’s
biologically driven perspective is supported by Plutchik (1980, 2001), who
additionally develops the “wheel of emotions,” reproduced in Figure 3-1
below.
12
Another vital model of viewing emotions is offered by Scherer (2005), who calls
it the Component Processing Model of Emotion, comprising five crucial elements,
i.e. the cognitive component, the peripheral efference component, the motivational
component, the motor expression component, and the subjective feeling
component. From Scherer’s (2005) perspective, emotion experience makes all of
these items coordinated and synchronized for a short period of time, driven by
appraisal processes. Besides, emotions are described as discrete and consistent
responses to internal or external events which have a particular importance for the
organism (cf. Scherer 2005; Frijda 1986, 2007). Scherer (2005) classifies emotions
under the group of short-lived affective phenomena, in contradistinction to moods,
attitudes and personality traits, which are long-lasting. Rozwadowska and Willim
(2016) draw a conclusion that if Johnson-Laird and Oatley’s (1989) approach were
combined with Scherer’s (2005) model and Hartshorne et al.’s (2010) proposal of
distinguishing brief psychological states (fright, anger, delight) from stable
tendencies, viz. dispositions (love, liking, hatred), then “from the psychological
perspective dispositions / attitudes / emotional relations tend to be long-lasting,
whereas causatives and responses to external or internal stimuli are short-lived”
(Rozwadowska and Willim 2016: 19; cf. Myers 2004).
Figure 3-1. Plutchik’s wheel of emotions: two-dimentional circumplex model

(Plutchik [1980] 2001: 349)
The wheel comprises eight primary emotions grouped into positive and
negative ones, i.e. joy versus sadness; anger versus fear; trust versus
disgust; and surprise versus anticipation. In addition, some basic emotions
can blend to form complex emotions. The latter could arise from blending
cultural conditioning or connotation with the basic emotions, e.g. a
mixture of interpersonal anger and disgust may lead to contempt (Plutchik
2001: 350). What is more, Plutchik (2001) argues that advanced emotions,
such as love, are a combination of two or more generic emotions; thus,
love, experienced as a strong feeling of attachment and attraction, is a
combination of joy and trust. While remorse, also regarded as an advanced
emotion, is characterised by feelings of regret and sometimes shame,
whereas regret is a combination of surprise and disgust. In short, Plutchik
(2001) suggests that besides the eight basic emotions, the two, i.e. love and
144 Chapter Three
remorse, form a set of the top ten emotions which each human being
mostly faces in his/her life.13
Additionally, certain emotion words characterised as culture-specific
phenomena can be, nevertheless, comparable to their translational
counterparts. As maintained by Johnson-Laird and Oatley (1989, 1992),
Ekman (1999), Turner (2007), Rott and Verhoeven (2016), some basic
emotional modes are assumed to be stable across cultures. These are
happiness, sadness, fear, anger, and disgust. Plutchnik’s (1980) much
more complex wheel of emotions (see Figure 3-1) comprises these five
basic modes, as well. And these precisely are the most common subject
fields, called emotion domains,14 which are mostly alluded to in the
literature while discussing emotions.
13
What is more, having examined a wide range of emotion theories across
different fields of research, Turner (2007, 2009) identifies four emotions that all
researchers relate to human neurology, viz. assertive-anger, aversion-fear,
satisfaction-happiness, and disappointment-sadness. These four categories are
termed primary emotions, which, when combined, may produce more elaborate
emotional experiences, called first-order elaborations, including sentiments such as
pride, triumph, and awe. In Turner’s theory (2007), emotions can be experienced at
different levels of intensity; thus, feelings of concern are a low-intensity variation
of the primary emotion aversion-fear, whereas depression is a higher intensity
variant. Two of the main eliciting factors for the arousal of emotions within this
theory are expectation states and sanctions.
14
In linguistics the term “domain” denotes “a subject field which has a particular
set of vocabulary associated with it” (Macmillan Dictionary retrieved 30/11/2016
from http://www.macmillan dictionary.com/dictionary /british/domain). This
general definition of “domain” is meant in this context, without making reference
to a far broader concept of “domain” brought by cognitive linguistics.
Cognitive linguistics, in turn, introduces the terms “conceptual domain,”
”conceptual metaphor,” or ”cognitive metaphor,” referring to the understanding of
one idea, or conceptual domain, in terms of another, “mapping the structure of one
domain onto the structure of another” (Lakoff 1987). A conceptual domain deals
with any coherent organization of human experience, and it can be exemplified by
the understanding of quantity in terms of directionality (e.g. “the price of peace is
rising”) (Lakoff 1987). Besides, within the framework of Conceptual Metaphor
Theory (Lakoff and Johnson 1980), metaphors link two conceptual domains, the
“source” domain and the “target” domain. While the “source” domain comprises a
set of literal entities, attributes, processes and relationships, linked semantically
and apparently stored together in the mind, the “target” domain tends to be
abstract, and takes its structure from the source domain, through the metaphorical
link, or “conceptual metaphor.” Target domains are supposed to have relationships
between entities, attributes and processes which mirror those found in the source
domain. At the level of language, “source” domains can be expressed through
Consequently, the basic emotion domains discussed above, proposed

by Johnson-Laird and Oatley (1989, 1992) (cf. Levinson et al. 2007), do
comply with most of the 22 top psych-verbs found both in Grafmiller’s
(2013) study and elicited in my research, listed in (3.6) below. If there are
more pertinent lexicalizations, the verbs are selected on the basis of
frequency (in an established corpus), and the intuition of being common in
use.
(3.6) Correlation of the basic EMOTION DOMAINS with the selected

psych- verbs:
a. HAPPINESS ĺ love, enjoy, fascinate, amuse,
surprise,* appeal to,* matter to*
b. SADNESS ĺ worry, depress
c. ANGER ĺ annoy, anger
d. FEAR ĺ fear, scare
e. DISGUST ĺ hate, horrify
As can be seen in (3.6), 15 of the initial list of the 22 top psych-verbs seem
to match the basic emotion domains intuitively well. The positive emotion
of HAPPINESS comprises 7 psych-verbs which are generally associated
with the state of being delighted or positively influenced. The remaining
four domains, i.e. SADNESS, ANGER, FEAR, and DISGUST, are
negative ones, covering altogether 8 psych-verbs. Thus, the balance is
maintained between positive and negative emotions. However, using the
common intuition, some verbs may be cross-domain specific, e.g. worry15
refers both to SADNESS and FEAR; and hate is linked to DISGUST,
ANGER and SADNESS. In these cases, the definitions of the verbs
provided by Johnson-Laird and Oatley (1989) are decisive, i.e. worry
denotes ANXIETY / SADNESS, while hate refers to DISGUST (Johnson-
Laird and Oatley 1989: 114, 121). Likewise, the verbs surprise, appeal to,
and matter to have been marked with an asterisk to indicate some
difficulty to classify them. In fact, they can be defined as “causing a
sudden unexpected onset of an emotion (generally positive ones),” and the
related words and expressions, which can be understood as organized in groups

resembling “lexical sets” or “lexical fields.” In the target domain entities, attributes
and processes are lexicalized using words and expressions from the source domain.
These words and expressions are occasionally named “linguistic metaphors” or
“metaphorical expressions” to distinguish them from conceptual metaphors
(Lakoff 1987).
15
The verb has been reduced to its single form “to worry” without any
preposition, e.g. “about.”
146 Chapter Three
key to group them under the “happiness” label lies in the positive feeling
they refer to. Besides, the verbs appeal to and matter to are the only
representatives of class III (OE) psych-verbs; that is why, their presence in
the set of psych-verbs, selected for further research, is more than required.
Moreover, seven of the initial list of the 22 top psych-verbs (cf. Table
3-1 and Table 3-2) have not been categorised into the basic emotion
domains at all, and as a result have been left aside. They are systematically
ambiguous due to their multiple meanings, even polar ones. For instance,
the verb concern may refer to anxiety or sympathy for someone else, and
thus, it may belong either to the HAPPINESS (positive) or to SADNESS
(negative) domain. Similarly, want / need are defined as “to have a goal
which if attained causes happiness (or makes good deficiency);” hence,
they do not refer to one specific emotion. Another kind of ambiguity and
difficulty to assign certain psych-verbs to the basic emotional domains, as
Clore et al. (1987) point out, arises in the use of words that do not, strictly
speaking, imply emotions but convey an emotional / mental state, e.g.
“feeling inspired” or “feeling in need of something.” This refers to the
verbs trust “to believe and rely on,” and respect “to judge that someone
deserves to be admired” from Table 3-1. Therefore, these items are also
excluded from further research. Finally, the verb like (cf. Table 3-1) has
been crossed out from the final list of psych-verbs since it has a substitute,
i.e. the verb enjoy. Finally, the verb appreciate from Table 3-1 has not
been chosen either, on account of the fact that, representing the positive
emotional domain, it may range in scale of emotion, from extreme
exhilaration to a neutral state of accepting something.
3.2.4 The final selection of the psych-verbs

All things considered, the final list of the psych-verbs elicited for the
further research concerning psychological idioms is categorised in
accordance with Belletti and Rizzi’s (1988) typology. Besides, the choice
of psych-verbs has been based on the criterion to take into account only
the most frequent psych-verbs, with the top occurrence in the COCA
Corpus. Finally, the verbs under scrutiny represent all the five basic
emotion domains. The results of the final selection of the psych-verbs are
presented in (3.7).
(3.7) The 15 representatives of psych-verb classes (final selection):
class I (SE): love, enjoy, hate, fear, worry–5 verbs
class II (OE): annoy, fascinate, amuse, scare, depress,
anger, horrify, surprise–8 verbs
class III (OE): appeal to, matter to–2 verbs.
As can be seen in (3.7), the number of class members is unbalanced. Class

II (OE) group comprises eight verbs, which are diversified in terms of
emotions they refer to. It is the most numerous class, including both
stative and eventive psych predicates. There are more verbs in this group
purposely because the so-called psych effects arise just in the case of these
predicates. The least numerous is class III, being represented by the only
two top frequent members, which occur significantly in the corpus. Both
of them (appeal to and matter to) denote positive emotions. Class I (SE),
in turn, includes five psych-verbs of multi-emotional reference, from
extremely positive (love) to highly negative ones (hate).
In section 3.3, an attempt is made to elicit psychological idioms which
belong to the same basic emotion domains as the top single psych-verbs.
3.3 Psychological idiomatic phrases:

A Corpus-based study
Since both the set of the most representative class I-III psych-verbs and
the basic emotion domains they comply with have been specified in
section 3.2 of this chapter; the ground for the next stage of the research has
been prepared. The top single psych-verbs are expected to serve as the
basis for selecting idiomatic units from the same emotional field. Those
psychological phraseological units are to be the substitutes for the psych-
verbs in terms of emotions and the psychological condition an Experiencer
undergoes.
Section 3.3 is organised as follows: first, the aim of the study of psych-
idioms is specified, followed by the definition of idiomatic units, repeated
here from Chapter One for the sake of convenience. Then, the data
collection method adopted in the corpus study carried out here is outlined.
3.3.1 The aim of the research
The aim of this stage of the study is to collect as many psychological

phraseological units as possible. Idiomatic phrases qualify for our
purposes given that they fulfil the following two criteria: (i) these are VP
idiomatic expressions that can be used instead of common psychological
predicates to express one’s mental or/and emotional condition; and (ii)
they derive from the same basic emotion domains the simple top psych-
verbs are set in.
Under these primary aims to achieve, it is worth turning now to
reestablishing the working definition of phraseological units and analysing
the relevant data.
148 Chapter Three
3.3.2 The working definition of idiomatic units reconsidered
For the sake of this research, the working definition of idiomatic

expressions is the one adopted after Liu (2008), and O’Dell and McCarthy
(2010). The latter identify idioms as “fixed combinations of words whose
meaning is often difficult to guess from the meaning of each individual
word” (O’Dell and McCarthy 2010: 6). Besides, Harwood et al. (2016)
add that “an idiom is an expression with a non-compositional
interpretation. That is, its meaning as a whole is not derivable from the
literal meanings of its parts” (Harwood et al. 2016: 3). To be precise, in
the idiom show the white feather “scare,” none of the constituents that
compose it conveys this meaning independently. Regardless of this fact,
the figurative reading of the idiom is determined by these specific lexical
elements. Therefore, by replacing any of the component items, the
idiomatic meaning is lost, with only the literal meaning left available, as
illustrated in (3.8).
(3.8) a. #He showed the white plumes.16

b. # He showed the bleached feather.
c. # He exhibited the white feather.
d. # He showed a white feather.
As can be seen in (3.8), the figurative interpretation of the idiom show the
white feather is based upon the elements show, the, white, and feather.
In summary, the working definition provided here complies with the
definition of idioms, called also phraseological units or phrases,
introduced in Chapter One of the book. It should be remembered that in
most cases idiom constituents do not contribute to the overall meaning of
the idiomatic phrase, then the idiomatic unit should be recognised as a
metaphor and a cohesive entity treated as a whole.
3.3.3 Data collection and methodology applied
The research reported here is based on English data collected from English
dictionaries and thesauri listed in (3.9), in which the idiomatic synonyms
of the psych-verbs under scrutiny have been searched for.
16
In this book, # is used to indicate loss of the idiomatic meaning.
(3.9) a. Collins Thesaurus of the English Language. 2002 (Complete

and Unabridged 2nd Ed.). HarperCollins Publishers.
Retrieved from http://www.freethesaurus. com
b. Google Dictionary. Retrieved from https://docs.google.com

/document
c. O’Dell, Felicity; and McCarthy, Michael. 2010. English

Idioms in Use Advanced. Cambridge: Cambridge University
Press.
d. Power Thesaurus. 1969. Retrieved from http://www.

powerthesaurus.org
e. Seidl, Jennifer; and McMordie, W. 1978. English Idioms

and How to Use Them (Fourth Ed.). Oxford: Oxford
University Press.
f. Wielki multimedialny sáownik angielsko-polski i polsko-

angielski [Great Multimedia English-Polish and Polish-
English Dictionary]. 2005. Oxford: PWN.
g. WordNet 3.0, Farlex Clipart Collection. 2012. Princeton:

Princeton University, Farlex Inc. Retrieved from
http://www.freethesaurus.com
Every single verb has been first checked and associated with a specific
definition, for which an idiomatic equivalent has been searched.
The study focuses on psychological verbal idioms, i.e. instances in
which a psych meaning occurs within idiomatic expressions that begin
with a verb. A number of criteria defining a verbal idiom have been
identified in the literature (Marantz 1984; Kiparsky 1987; Nunberg et al.
1994; Croft and Cruse 2004; Svenonius 2005; Evans and Green 2006; and
Harwood et al. 2016; among others). These are summarised in (3.10)
below.
(3.10) Criteria for defining a phraseological unit as a verbal idiom:

a. It must contain a lexical verb.
b. It must have a non-literal interpretation.
c. It must be able to interact with productive syntax.
d. It must be comprised of lexical items that are found outside
of the context of the idiom.
150 Chapter Three
e. It must be formed in a manner that obeys the regular

syntactic rules of the language.
(Harwood et al. 2016: 4)
Accordingly, any expressions which form a literal meaning and lack any
figurative / idiomatic meaning, have been omitted. The precise instances
excluded from the study by the criteria mentioned in (3.10) are listed in
(3.11).
(3.11) Phraseological verbal units which have been excluded from the
study by the criteria listed in (3.10):
a. (i) nominal idioms, e.g. butterflies in one’s stomach ”fear”
(ii) causative structures, such as make / get someone do

something, e.g. make somebody roll in the aisles “amuse”
vs. “make” used figuratively not in causative structures,
e.g. make the grade “amuse;”
b. units which have predictable, literal meanings, viz. phrases

with get / become / be / have + adjective / past participle
NP (a cognate of the psych-verb), e.g. be crazy / mad
about, get worried, have / get pleasure in “enjoy.”
But if a phrase obtains a figurative meaning, e.g. have a

bee in one’s bonnet “worry,” it is included in the list of
idioms to be examined;
c. idioms in the form of fixed clauses, working as sayings,

similes, and proverbs, e.g. Curiosity killed the cat, since
these are frozen expressions which do not interact with
productive syntax. In other words, once these phrases are
embedded in subordinate clauses (#I was wondering if
curiosity killed the cat), or undergo question-type shifting
(#Did curiosity kill the cat), their figurative meaning is
lost;
d. idioms containing irregular lexical items, e.g. give

someone the heebie-jeebies: “Walking across the park
after dark gives me the heebie-jeebies.” (Google), since the
item the heebie-jeebies is never used in any other present-
day English collocations, except for the idiom itself;
e. phrases with irregular syntax such as be that as it may,

because verbal idioms are usually well-formed in terms of
grammar, obeying the structure building mechanisms of
the language.
Besides, throughout my research, I have only concentrated on the most

canonical type of verbal idioms, i.e. comprising a lexical verb and the
direct object of the idiom. Thus, all instances lacking a direct object are
left aside. These are (i) all phrasal verbs comprising only a verb and a
particle (go in for “enjoy”); (ii) idioms composed of intransitive verbs (go
off the deep end “fall in love”); and (iii) copular verbs in combination with
some predicate (be down in the dumps “worry,” be nuts/wild/potty about
“love”).
In brief, following the definition of idioms, adopted from O’Dell and
McCarthy (2010: 6) at the very beginning of the book, and repeated in this
section, only those psychological idioms have been chosen whose
constituents mostly do not contribute to the overall meaning of the idiom.
Moreover, such phraseological units, similarly to psych predicates,
comprise a participant which experiences some emotional or mental state,
i.e. an Experiencer, and a Stimulus / Causer / Cause / Target, which has
contributed to this specific state or become a target of it. Following the
convenient terminology and typology suggested in Belletti and Rizzi
(1988) (cf. Dowty 1991; Pesetsky 1995; and Landau 2010), the data are
divided into three classes, in the same way as their psych-verbal
counterparts in (3.7). The search results are commented upon after each set
of idioms.
3.4 Data analysis and preliminary discussion of the results

The search has yielded a relatively big number of psychological idiomatic
expressions, which are equivalents of the top selected psych-verbs. Both
the psych-verbs and their idiomatic counterparts are set in the same
emotion domains, i.e. HAPPINESS, SADNESS, ANGER, FEAR, and
DISGUST. Nonetheless, the idiomatic counterparts express different
shades of the same emotion, which has not been noted scrupulously for
each idiom under scrutiny. For instance, while the idiom have a soft spot
for X means “to feel a lot of affection for someone, without knowing
why,” another idiom from the same emotion domain (LOVE), lose one’s
heart to X, means “to fall in love,” which makes the intensity of the
emotion far greater in this case, in comparison with the one mentioned
above.
152 Chapter Three
In total, the database thus created contains 161 English idiomatic units,
which are grouped according to the surface syntactic patterns they occur in
for each of Belletti and Rizzi’s (1988) classes.
Taking into consideration the pattern the core of an idiom forms and
the position an argument is placed in, the canonical idioms I have
scrutinised fall into 12 distinct types, schematized in (3.12).17
(3.12) Types of syntactic patterns of the psychological idioms examined:

Type (A): V + NP (no open position)
paint the town (red)
Type (A’): V + possessor + N

float Y’s boats
Type (A’’): V + possessor + N + particle

eat Y’s heart out
Type (B): V + NP + preposition + NP (complement of a P)

carry a torch for X
Type (B’): V + NP + preposition + possessor +N (complement of a

P)
have a yellow streak / belly down Y’s back
Type (C): V + NP + preposition + NP (complement of a P)

hold X in abomination / contempt
Type (C’): V + possessor +N + preposition + NP (complement of a

P)
put Y’s nose out of joint
17
Interestingly, Bruening (2010), cited in Kim (2014), argues that in English there
are three existent classes of idioms with Double Object CConstructions (DOC) and
prepositional dative constructions / prepositional datives (PD). He suggests the
following overall distribution of idiomatic patterns with DOC and PD
constructions (idiomatic parts are underlined, X in brackets is a variant):
a. Class I: Verb NP NP (give X the creeps)
b. Class II: Verb NP to NP (give rise to X)
c. Class III: Verb NP to NP (send X to the showers)
d. Class IV: V NP NP (nonexistent)
(Kim 2014: 216)
Type (D): V + possessor + N + preposition + NP (complement of a

P)
lose Y’s heart to X
Type (D’): V + NP + preposition + possessor + N (complement of a

P)
sweep Y off Y’s feet
Type (D’’): V + possessor + N +preposition + possessor + N

(complement of a P)
have Y’s heart in X’s mouth
Type (E): V + NP + NP (Double Object Construction)

give Y the blues
Type (E’): V + a complement small clause

drive Y batty / nuts/ bananas / bonkers
In the patterns listed above, the idiomatic part in each expression is

underlined, and it is called “the core of the idiom.” Besides a verb itself,
an idiomatic core can be composed of a fixed NP without any open
position, as in the rarely occurring Type A. In Type A’, a verb occurs with
a fixed noun modified by a possessor realising an open position, while in
Type A’’, both a verb with a noun with an open position of a possessor,
and a fixed particle comprise the idiom.
Furthermore, in Type B, the idiom formed by a verb and its object is
extended to a fixed preposition and an NP (a complement of a P) realising
an open position. Type B’ differs from Type B in that the PP consists of a
fixed preposition and the complement NP, in which the noun is fixed, but
it is modified by a possessor realising an open position.
Both Type C and Type C’ have fixed PPs, but in the former the noun
object of the idiom realises an open position, while the object of the latter
(Type C’) comprises a fixed N modified by a possessor.
Type D is similar to Type C’ in that its fixed object is modified by a
possessor. However, in Type D, the PP comprises a fixed P and an NP
with an open position. In Type D’, the open position occurs in an object
NP and in a possessor modifying an NP, which is the complement of a
fixed preposition. In Type D’’, there are two possessors realising an open
position, one modifying a fixed object, and the other modifying a
complement of P.
154 Chapter Three
In Type E, a Double Object Construction appears, but the first NP

realises an open position. Finally, in the last of the twelve patterns of idioms,
in Type E’, a verb is followed by a small clause, which is formed by an NP
realising an open position and a fixed complement of the small clause.
Moreover, in the above-mentioned patterns of idioms, X is an NP
argument which functions as a Target or a Subject Matter or a Causer,
while Y as an argument corresponding to an Experiencer, either in a
subject or an object position. The arguments also represent open positions,
i.e. certain positions in an idiomatic phrase which are not fixed, but
occupied by X or Y (cf. WitkoĞ and Dziemianko 2006).
3.4.1 Idiomatic units with an Experiencer in the subject position
The data in (3.13)-(3.17) below illustrate all the types of idioms possible
for five SE psych-verb which have been on the research list, i.e. love,
enjoy, hate, fear, worry. All the elicited idiomatic phrases are arranged
according to a syntactic pattern they exhibit, while the exemplary
sentences for those idioms, taken from the COCA and/or obtained via the
Google Search, are listed in APPENDIX 1.
(3.13) love - feel deep affection (13 idioms)

carry a torch for X
fall head over heels in love with X
set store by X
think the world of X
take (great) delight/ interest / joy/ satisfaction in X
take a fancy / a liking / a shine to X
have a soft spot for X
have a thing about X
have a weakness for X
have eyes for X
go a bundle on X

P)
lose Y’s18 heart to X
set Y’s heart on X
18
The genitive possessor in the case of this idiom has to agree with the subject,
represented by an Experiencer (Y).
(3.14) enjoy - receive pleasure or satisfaction from something

(11 idioms)
Type (A): V + NP
paint the town (red19)
raise the roof
have a ball
blow / let off20 (some) steam

kick (up) Y’s heels

make the most of X
derive/gain/get pleasure from X
take/find pleasure in X
get a buzz out of X
get a kick / a charge / a bang out of X
reap the benefits of X
(3.15) hate - feel strong dislike for or hostility toward (10 idioms)
pour scorn on X
bear ill will toward X
bear / owe a grudge against X
bear / feel aversion / malice / hostility / repugnance
toward (to) X
have no stomach / taste / use for X
show contempt for X

hold X in abomination / contempt21
19
“Red” in this idiom is an optional element, though if used, it is a part of a
resultative phrase.
20
“The particle “off” is the constituent of the verb “let/blow” itself, not a
preposition requiring a complement.
21
The idioms hold X in contempt and look down Y’s nose at X refer to one’s
CONTEMPT and mean “to spurn, to regard oneself as superior to others and thus
act in a haughty or snobbish manner”. Nonetheless, these idioms have been
grouped into the basic emotion domain HATE, which consists of disgust, anger,
and/or sadness (cf. Johnson-Laird and Oatley 1989: 121). Truly, as suggested by
Prof BoĪena Cetnarowska, the aforementioned idioms might have been grouped as
156 Chapter Three

P)
turn Y’s back on X
turn Y’s nose up at X
look down (Y’s) nose at X22
(3.16) fear - feel apprehensive, afraid or frightened of something /
someone (9 idioms)
Type (A): V + NP
give / raise an alarm
show the white feather
get cold feet
have goose bumps
have / get pins and needles
turn tail (and run)

take dim view of X23

P)
Type (D’’): V + possessor + N +preposition + possessor +N

(complement of a P)
have Y’s heart in Y’s mouth
(3.17) worry – to feel uneasy or concerned about something; to fret or be

anxious about the welfare of someone or something (SE)
(6 idioms)
the representatives of the complex emotion CONTEMPT, instead of the basic

emotion HATE. Even though both of these domains comprise the same primary
emotions, i.e. disgust and anger; the complex domain CONTEMPT would be a
more adequate choice for these two idioms.
22
In the idiom look down (Y”s) nose at X, the particle “down” is a part of the
phrasal verb “look down.”
23
I would like to thank Prof BoĪena Cetnarowska for her suggestion to mention
here that in the idiom take (a) dim view of X, the uncountable noun view can be
altered into a countable one, still providing the same meaning “to worry; to view
sceptically, pessimistically; to regard with disapproval,” e.g.
They take (a) dim view of those they consider outsiders. The public, however,
seems willing. (COCA)
Type (A): V + NP
have the blues

eat Y’s heart out

lose sleep over X

P)
have a bee in Y’s bonnet
have ants in Y’s pants
have butterflies in Y’s stomach24
On the basis of the idioms listed in (3.13)-(3.17), and exemplified by

the sentences, reproduced from the COCA Corpus / via the Google Search
(cf. APPENDIX 1), it can be said that the equivalents of the SE psych-
verbs are quite numerous (49 units). The most frequent type of the eleven
syntactic patterns (51%) for SE psychological idiomatic phrases is Type
(B), comprising a verb, a fixed object, a fixed preposition, and an NP
complement within the PP, realising an open position (V+NP+P +NP).
Nearly one quarter of all the types is covered by idioms of Type (A) viz.
without any open position, i.e. comprising a verb and a fixed object
(V+NP).
Type B’ (V + NP + preposition + possessor +N), including idioms
which are constructed of a verb, a fixed object, a fixed preposition, and an
NP complement of the P, in which the possessor realises an open position
while the N is fixed (V+ NP + P + possessor+ N), covers 8%. A similar
percentage (10%) corresponds to Type D (V + possessor + N + preposition
+ NP), in which there are two open positions, i.e. the one represented by a
possessor of an NP object, and the other realised by an NP that is a
complement of a P.
Extremely rare (2%) are the following four types: Type A’ (V +
possessor + N) with a possessor placed within an NP object; Type A’’ (V
+ possessor + N + particle), covering, beside a verb and a N, a fixed
24
Cf. (3.11) in Chapter Three, where the examples of phraseological verbal units,
excluded from the study by the criteria listed in (3.10), are listed. The phrase have
butterflies in the stomach comprises the verb have and a NP, and it obtains a
figurative meaning “to worry.” Therefore, it is not excluded from the list of idioms
to be examined.
158 Chapter Three
particle; Type C (V + NP + preposition + NP), in which the verb is

followed by an object realised as an open position, and by the fixed
Prepositional Phrase (V+NP+ PP); and Type (D’’) with two possessors
within the NPs, in the object position and within a PP (V + possessor + N
+preposition + possessor +N).
Four types do not occur with SE psychological idioms at all. They are
as follows: Type C’ (V + possessor +N + preposition + NP); Type D’ (V +
NP + preposition + possessor + N), in which an open position within a PP
is realised by a possessor while the N is fixed; Type E (V + NP + NP) with
Double Object Constructions; and Type E’ (V + a complement small
clause), covering a small clause, i.e. an NP realising an open position,
followed by a fixed complement of the small clause. Table 3-3 and Figure
3-2 summarise the results for the five SE psych-idioms.
Table 3-3. Types of syntactic patterns for SE psychological idioms
Idiom types and their syntactic patterns
A A’ A’’ B B’ C D D’’
Verb/
Tokens V+ V+ V+ V+ NP V+ NP V+NP V + V+
NP possessor possessor +P +P+ +P +NP possessor possessor
+N + N+ +NP possessor +N +N
particle +N +P+NP +P+
possessor
+N
love 13 --- --- --- 11 --- --- 2 ---
enjoy 11 4 1 --- 6 --- --- --- ---
hate 10 --- --- --- 6 --- 1 3 ---
fear 9 6 --- --- 1 1 --- --- 1
worry 6 1 --- 1 1 3 --- --- ---
TOTAL /
11 1 1 25 4 1 5 1
49
In Search of Psycchological Idiom
ms: A Corpus-bbased Account 159
Syn
ntactic patteerns for SE psychologicall idioms A
(in percenta
age)
A'
D''
D
C 2% A A''
%
10%
2% 23%
B
B' A'
8% 2% B'
A''
2% C
B D
51%
D''
Figure 3-2. S
Syntactic patterrns for SE psych
hological idiom
ms (in percentag
ge)
Furthermmore, on the basis

b of the results obtainedd, we can state that the
number of ppsychological idioms found d for the sampple of 5 top SE
S psych-
verbs is quiite equally distributed, and d all of them constitute a fairly
f big
number of 449 units of thee idiomatic in nventory. Reggardless of thee fact that
SE psych-vverbs are recoognised as sttative, their iidiomatic cou unterparts
comprise m mostly agentivve verbs, e.g. fall, set, takke, make, turn n, paint),
carrying a specific sem mantic interpretation, whiich for somee idioms
contributes tto the overall idiomatic meaaning (carry a torch for X “love”,
“ or
hold X in abbomination “hhate”). On the other hand, tthere are somee frozen /
opaque idiooms which doo not distributte their idiom matic meaning g between
their single constituents (ppaint the town n “enjoy”, or show the whitte feather
“fear”) (cf. tthe discussionn of idiom commpositionalityy in Chapter Foour).
3.4.22 Idiomatic phrases

p for OE (class III) psych-verrbs
The data wiith the psychoological idiom

matic equivalennts of the eight psych-
verbs whichh represent OE O (class II)) predicates, i.e. annoy, fascinate,
f
amuse, scarre, depress, anger,
a horrifyy, surprise, arre presented in
i (3.18)-
(3.25) beloww. These iddioms are org ganised in 112 types (A)-(E”), as
discussed inn (3.12), andd exemplified with sentencces obtained from the
COCA Corppus and via thee Google Searrch in APPEN NDIX 2.
160 Chapter Three
(3.18) annoy - make (someone) a little angry; irritate, upset, irritate,

aggravate (26 idioms)
Type (A): V + NP25

play (a game of) cat and mouse
get the hump

get Y’s goat
raise Y’s hackles
rattle Y’s cage / chain
try Y’s patience / try the patience of Y
ruffle Y’s feathers

get Y’s dander /hacklers/ Irish up
put/get Y’s back up

get a rise out of Y
take it out of Y
stir up a hornet’s nest (of something) amongst/ in Y
upset the apple cart against Y

P)
bring a hornet’s nest round Y’s ears
25
In Type A, all the idiom components are fixed, and an Experiencer (Y) is
realised covertly, but it can be found in the context of a particular sentence, e.g. in
(i) below, American fighters are Experiencers, while in (ii) Tony is an Experiencer
(placed in the position of a subject), who got annoyed by us and the fact that we
hadn”t invited him to the party.
(i) play (a game of) cat and mouse [against /towards / with Y] “to annoy Y”
Enemy warplanes have been playing a deadly game of cat and mouse [against
American fighters], trying to bring American fighter planes into rang of their
missiles. (Google)
(ii) get the hump “to get annoyed” Tony got the hump because he thought we
hadn”t invited him to the party. (Google)

cut Y to the quick / bone
drive Y up the wall
put Y in a hole /a bind
put Y on the spot
put Y out of countenance
put Y through wringer
throw/send Y into a tizzy / tizz
Type (C’): V + possessor +N + prep. + NP (compl. of a P)

put Y’s nose out of joint

P)
thumb X’s nose at Y

give Y the pip
rub Y (up) the wrong way

drive Y batty / nuts/ bananas / bonkers / crazy
(3.19) fascinate – attract the strong attention and interest of (someone)

(11 idioms)
catch Y’s eye / catch the eye of Y
win Y’s heart / win the heart of Y
turn Y’s head
tickle Y’s fancy
stir the /Y’s blood

cast a spell on Y
make a hit with Y
pay court to Y

put Y in a trance
set Y on fire
162 Chapter Three

P)
(3.20) amuse – cause (someone) to find something funny; provide

interesting and enjoyable occupation for (someone); entertain
(4 idioms)
gladden Y’s hearts / the heart of Y

make the grade for Y
play the fool for Y

tickle Y to death / to pieces / pink26
(3.21) scare – cause great fear or nervousness in; frighten
(11 idioms)
curdle / chill Y’s blood

frighten / scare the life / the hell out of Y
put the fear of God into Y
put/get the wind up Y
strike terror into Y

chill Y to the bone/marrow
send Y into a cold sweat
throw /send Y into a panic

P)
shake X’s / a fist at Y
26
If the component “to death /to pieces” in the idiom “tickle Y to death / to pieces
/ pink” is changed into “pink,” the idioms forms Type (E”): V + a complement
small clause.
give Y goose pimples/ bumps

give Y (quite) a (bit of) turn / a fright
(3.22) depress – make (someone) feel utterly dispirited or dejected

(20 idioms)
Type (A): V + NP
upset the applecart27

break Y’s heart / spirit
damp Y’s hopes
dash / wither Y’s hopes

do a number on Y
knock the stuffing out of Y
put a damper on Y

P)
bring tears to Y’s eyes

cut Y down to size
take/ knock Y down a peg / notch (or two)
put / send / throw Y into a (blue) funk
leave Y in the / a lurch
bring Y into disrepute
knock Y for six
27
This idiom belongs to Type A, in which all the idiom components are fixed,
whereas an Experiencer (Y) is realised covertly, but recognisable from the context
of the sentence, e.g. in (i) below, you is the Experiencer, who experiences the state
of being depressed.
(i) upset the applecart [in Y] “to depress Y”
I don”t want to upset the applecart now by asking you to change the date
for the meeting. (Google)
164 Chapter Three


bring Y low
knock Y sideways
(3.23) anger – fill (someone) with anger; provoke anger in

(10 idioms)
Type (A): V + NP
fan the fire / flame28 (of something)

kindle Y’s wrath

get Y’s back / dander up29
put / set Y’s back (up)

blow a gasket on Y
breathe fire over Y

drive Y out of mind
drive Y to distraction
drive Y round the bend / twist

P)
vent X’s spleen at Y
28
This idiom represents Type A, with all components fixed. Therefore an
Experiencer (Y) is realised covertly, but still it can be noticed from the context of
the sentence, e.g. in (i) below, she is the Experiencer, who experiences the state of
being angered.
(i) fan the fire / flame (of something) [in Y] “to anger Y”
She already found him attractive, but his letters really fanned the flames
[in her]. (Google)
29
The particle “up” is a constituent of the verb “get.”
(3.24) surprise – cause (someone) to feel mild astonishment or shock.

(15 idioms)

take Y’s breath away30
knock Y’s socks off

knock Y down / over31 with a feather
blow Y out of the water
catch Y off balance / up short32 / napping33
catch Y on the hop
throw Y off balance
strike Y with awe
strike Y with wonder

P)
catch Y off (Y’s) guards

throw Y a curve (ball)

catch Y unawares34
leave Y open-mouthed
leave Y at a loss for words
strike Y dumb
30
The particles “away,” and “off” work as particles of the phrasal verbs.
31
The particles “down / over” work as particles of the phrasal verb “knock down”
or “knock over.”
32
In the idiom “catch Y up short,” “short” is an adjective, which works as a
complement of the preposition “up,” and may be treated as a resultative phrase
(Type E”).
33
If the component “off balance” in the idiom “catch Y off balance / up short /
napping” is changed into “napping,” the idiom forms Type (E”): V + a
complement small clause.
34
All these idioms include a resultative phrase, i.e. “open-mouthed / at a loss for
words / dumb / with awe / with wonder,” which are treated as complements of the
small clauses.
166 Chapter Three
(3.25) horrify – fill with horror; shock greatly (8 idioms)

freeze Y’s (the) blood
put the screws on Y
scare the shit / the wits out of Y
frighten / scare the pants off Y
scare the bejesus out of Y
frighten / scare Y to death
frighten / scare Y out of their wits
give Y a turn
As can be seen in (3.18)-(3.25), the psychological equivalents of the OE

psych-verbs (class II) yield a big number of 105 units. Some verbs, such as
amuse or horrify have only a few psychological idioms; while others, i.e.
annoy, depress, and surprise, provide a set of 15 up to 26 idiomatic units.
Table 3-4 and Figure 3-3 summarise the results obtained for the eight OE
(class II) psych-verbs.
Table 3-4. Types of syntactic patterns for OE (class II) psychological idioms
Idiom types and their syntactic patterns

A A’ A’’ B B’ C C’ D D’ E E’
Verb/ Tokens
V+ V+ V+ V+ V+ V+N V+ V + V V V+a
NP posse- posse- NP + NP + P posse- posse- +NP +NP+ compl.
ssor+ ssor+ P P+ +P ssor+ ssor+ +P+ NP small
clause
N N+par +NP possess +NP N +P N posse-
-ticle
or+ N +NP +P+N ssor+
P N
annoy
2 5 2 4 1 7 1 1 --- 2 1
/ 26
fasci-
--- 5 --- 3 --- 2 --- --- 1 --- ---
nate/ 11
amuse
--- 1 --- 2 --- 1 --- --- --- --- ---
/ 4
scare
--- 1 --- 4 --- 3 --- 1 --- 2 ---
/ 11
depress
/ 20
1 4 --- 4 2 7 --- ---- --- 1 1
anger
/ 10
1 1 2 2 --- 3 --- 1 --- --- ---
surprise
/ 15
---- --- 2 --- --- 7 --- 1 --- 1 4
horrify
/ 8
---- 1 --- 4 --- 2 --- ---- --- 1 ---
TOTAL
/105
4 18 6 23 3 32 1 4 1 7 8
Syntactic patterns foor OE (class I I) psychologgical

idiom
ms (in percentagee)
A
D'
D
1% E E' A A'
4%
% 6% 4% A'
C' 5% A''
17%
1% B
A''
6% B'
C C
31% B C'
B' 22%
3% D
D'
E
E'
Figure 3-3. Syntactic patternns for OE (classs II) psychologiical idioms

(inn percentage)
The most ppopular typess of syntacticc patterns foor these psycchological

idioms are TTypes (C)-(C’) comprising 32%,
3 which iss a distinctive result, in
comparison with SE idiom ms. In these patterns
p an oppen position is realised
either as an object NP or as a possessor within the N NP object, whiile the PP
following thhe object is fixxed (V+NP+ PP / V+ poss essor +N+ PP P)). Then,
Types (A), (A’) and (A’’’), viz. pattern ns with or witthout an open n position
realised in aan NP (V+NP P, or V+ posssessor + N, orr V+ possessor + N +
particle), arre the secondd largest of all the types, comprising altogether
a
27%. Typess (B)-(B’) aree number thrree as far as the popularitty of the
syntactic paatterns is conncerned, coverring 25%. Thhese types co omprise a
168 Chapter Three
verb and a fixed object, followed by a Prepositional Phrase realising an

open position in a complement of the P, i.e. either an NP (V+NP+P +NP),
or in a possessor modifying the fixed noun (V+NP+P +possessor +N).
Moreover, both Type (E) with a Double Object Construction (V + NP
+ NP) and Type (E’) with a small clause complement, containing an NP
subject realising an open position, cover only 11% of all the types, but
they were absent for SE psych-idioms entirely. Types (D)-(D’) with two
open positions, i.e. in the object position and within a PP, hardly ever
occur in this idiomatic dataset (5%). All these types are distributed in a
different manner in comparison with the psychological idioms for SE
psych-verbs.
In addition, although regular OE psych-verbs (class II) consist of
agentive, stative and eventive verbs, their idiomatic equivalents comprise
mostly agentive verbs, e.g. warm, stir, cast, feed, pay, and turn. An
Experiencer Y is generally explicit, located within a PP or realised as a
direct object (an NP or a possessor modifying a fixed N) of the idiomatic
unit. In Type A (4%) psych-idioms, in which all idiom components are
fixed, an Experiencer is covert, but noticeable from the context of the
sentence.
3.4.3 Idiomatic phrases for OE (class III) psych-verbs
Class III psych-verbs is represented by two significant predicates, i.e.

appeal to, and matter to, which have yielded a small group of idioms in
the number of 7, arranged into syntactic patterns, as illustrated in (3.26)
and (3.27). The complete list supplemented with sentences extracted from
the COCA or obtained via the Google Search is available in APPENDIX
3.
(3.26). appeal – be attractive or interesting (4 idioms)

float Y’s boat
tickle Y’s fancy
whet Y’s appetite

P)
set / put Y on Y’s ear
(3.27) matter to – be important

i or significant for someone
(3 idioms)
Type (B): V + NP + preposition n + NP (compllement of a P))
carry soome / a lot of weight
w with Y
cut (no)) ice with Y
make a difference to Y
On the bbasis of the aforementione

a ed data, illustrrated in Figure 3-4, it
can be stateed that this cllass of psych--verbs lacks iidioms which occur in
n (3.12), viz. T
any of the 9 syntactic pattterns listed in Type (C): V++NP+ PP;
Type (C’): V+ possessoor +N + PP; Type (E): w with a Doublle Object
Constructionn (V + NP + NP);N Type (E E’): V + a com mplement smaall clause;
Type (D): V + possessor +N + preeposition + N NP; Type (D D’’): V +
possessor +N N + prepositiion + possesssor +N; Type (A) with a fixed f NP;
Type (A’’): V+ possessor +N +parrticle; or Typ ype (B’): V + NP +
preposition + possessor +N.
+
Howeverr, the idioms obtained in the t search forr Class III psy ych-verbs
appear in thhree syntactic patterns. The least commoon (14%) is Type (D’):
V + NP + preposition + possessor + N, while it is Type (A A’): V +
possessor + N, and Type (B): V + NP N + preposiition + NP, which
w are
mostly repreesented by thiis class of psych idioms, eacch covering 43%
4 of all
the patterns.. The agentive verbs, e.g. tickle,
t whet, oor set, can be found in
those idiomaatic structuress.
Syntacticc patterns for OE (class IIII) psychological

idioms (in percentagee)
D'
14% A'
A'
43%
B
B
43%
D'
Figure 3-4. S
Syntactic patterrns for OE (classs III) psychologgical idioms
(iin percentage)
170 Chapter Three

The aim of Chapter Three has been (i) to elicit the set of psych-verbs,
which both conforms with Belletti and Rizzi’s tripartite taxonomy, and
belongs to various but cross-linguistically adopted emotion domains; and
(ii) to find as many psychological idiomatic phrases as possible, equivalent
to the elicited top psych-verbs. The results obtained in the study have
yielded a great deal of relevant data as far as psychological predicates are
concerned.
Accordingly, the results have proved that the language inventory to
express one’s mental, emotional, and/or psychological condition comprises
quite a long list for each emotion domain, for any single psych-verb,
regardless of either what class it belongs to, or whether an Experiencer is
in the Subject or Object position. Significantly, all the idiomatic phrases
have been grouped into 12 distinct syntactic verbal patterns. Thus, the
idioms under scrutiny comprise nearly always agentive verbs, which occur
necessarily, with an object, followed or not by a Prepositional Phrase. The
open position is located either within the object complement or/and within
a PP. The open position can be realised as either an NP or a possessive
modifier. Some idiomatic phrases are opaque and have all their
constituents (NP, PP, or a small clause) fixed, whereas others comprise
elements which greatly contribute to the overall meaning of the idiom, and
make its meaning more predictable.
To conclude, the heterogeneity of these idiomatic expressions indicates
the necessity of learning at least some constraints and irregularities related
to idioms. This task is to be taken in the subsequent chapter. Besides, the
next part of the book is to address the question concerning the syntactic
and aspectual characteristics of these psych-idiomatic expressions.
CHAPTER FOUR
PSYCHOLOGICAL IDIOMS:
SYNTACTIC CONSTRAINTS
AND ASPECTUAL PROPERTIES
4.1 Introduction
Idioms represent “a multidimensional lexical space, characterized by a
number of distinct properties: semantic, syntactic, poetical, discursive, and
rhetorical” (Nunberg et al. 1994: 492); therefore, any attempt to categorise
idioms along single-criterion definitions or rules is futile. Taking into
account various dimensions of idiomaticity, two main opposing
approaches to idioms are usually distinguished in the literature, which
have been outlined in Chapter One of the book, and are now briefly
returned to, for the sake of convenience. These are: (i) the more standard
non-compositional approach, in which idioms are recognised as arbitrary
configurations of words with nonliteral meanings, syntactically and
semantically behaving as lexical entries; and (ii) the non-traditional
compositional approach, which refers to the non-arbitrary internal
semantic and syntactic structure of idioms.
In the traditional approach, typically adopted in the generative
grammar, it is believed that the meanings of idioms cannot be derived
compositionally by the morpho-syntactic rules of a language (e.g. the
meaning of the canonical idiom kick the bucket (“to die”) cannot be
derived from the meanings of its constituents, i.e. “kick,” “the,” and
“bucket”); thus, idioms are thought to be arbitrarily stipulated in memory
(cf. Chomsky 1980; Cruse 1986; Fraser 1970; and Katz 1973; among
others). In the generative literature, idiom chunks are not associated with
meanings, and it is said that “there is no relation between the meanings of
the parts and the meaning of the whole from the viewpoint of
“synchronic” structure” (Kiparsky 1976:79). Chomsky (1980: 149) names
idioms “non-compositional” structures; Machonis (1985: 306) defines
them as “frozen” expressions, “not predictable from the composition;”
while van der Linden (1992) follows Katz and Postal (1963: 275) in
172 Chapter Four
claiming that the meaning of an idiom “is not a function of the meanings
of its parts and the way these are syntactically combined; that is, an idiom
is a noncompositional expression” (van der Linden 1992: 223). Consequently,
since there is no relation between the linguistic forms of such idioms and
their meanings, and due to their lack of internal syntactic or semantic
structure, hardly any modification is possible in the case of this type of
“frozen” idioms (called idiomatic phrases, after Nunberg et al. 1994).
However, this standard view of idioms as non-compositional strings or
long words has been objected to in the current research. It has been
demonstrated that not all idioms are “frozen,” and as opaque or
unanalysable as kick the bucket. For most idioms (called idiomatically
combining expressions, after Nunberg et al. 1994), some relations between
their meaning and form can be established. In fact, the meanings of
particular components of idioms do play a role in the way idioms are used
and understood (cf. Cacciari 1993; Cacciari and Glucksberg 1991; Gibbs
1990; Keysar and Bly 1995, 1999). What is more, this type of idioms can
be syntactically transformed in various ways and their parts can be
modified, while some elements within an idiomatic expression can be
semantically productive (cf. Makkai 1972; Nunberg et al. 1994; O’Grady
1998; and Ifill 2002; among others). Thus, a question arises of what sort of
syntactic modifications can compositional idioms undergo.
Following this second research trend, attention is paid here to the
linguistic side of the distinction between so-called “decomposable vs. non-
decomposable idioms” (Nunberg 1978). Taking this division into
consideration, the purpose of Chapter Four is to discuss and analyse
idioms, with a special focus laid on the psychological idioms elicited in
Chapter Three, with the aim of gaining better theoretical and empirical
insights into syntactic rules they are governed by. More precisely, issues
that are to be addressed here include the following: (i) semantic properties
of idioms, with special reference made to psych-idioms, and a distinction
between idiomatically combining expressions and idiomatic phrases
(section 4.2); (ii) syntactic and semantic flexibility of psychological
idiomatically combining expressions (section 4.3); (iii) syntactic
constraints on psychological idiomatic phrases and the structure of idioms
analysed in previous accounts (section 4.4) and within the Phase Theory
(section 4.5); and (iv) the position of an Experiencer in psychological
idioms (section 4.6). The chapter ends with an overview of the aspectual
properties of psychological idioms (section 4.7), followed by the
concluding remarks, outlined in section 4.8.
Psychological Idioms: Syntactic Constraints and Aspectual Properties 173
4.2 Classifying and characterising idioms

Idioms used to be commonly treated as fixed phrases both lexically and
syntactically frozen, with no internal structure open to grammatical or
lexical operations (cf. Chomsky 1980; Cruse 1986; Fraser 1970; and Katz
1973; among others), until extensive corpus studies of selected idioms
have revealed that some idioms can undergo substantial lexical and
morphosyntactic modifications (cf. Moon 1998; and Fellbaum 2011;
among others). However, drawing a clear-cut line between non-
compositional, frozen idioms, on the one hand, and partly compositional,
modifiable idioms, on the other, does not reflect the way speakers actually
use idioms (cf. Zhu and Fellbaum 2015). Diversity seems to be the
fundamental characteristics of all idioms.
4.2.1 Semantic dimensions of idiomaticity

While making an attempt to classify idioms as either non-compositional or
compositional ones, it should first be stated that idioms may differ along
several orthogonal semantic dimensions, such as: conventionality,
compositionality, analysability and transparency, figuration, proverbiality,
informality, and the so-called “affect” (Nunberg et al. 1994: 492-493,
498). Since these dimensions have been discussed in detail in Chapter One
of the book, they are now briefly mentioned in connection with
psychological idioms.
4.2.1.1 Conventionality
The first semantic dimension is conventionality, defined as the relation

between some string of words and a particular semantic representation. An
element of arbitrariness is involved here in that a certain linguistic label is
used to express a certain conceptual representation. Some arbitrariness
may also be found in psychological idioms, listed in (4.1) below. To
recall, for the sake of convenience, in the instances of psychological
idioms in (4.1a-l), Y refers to an Experiencer, who experiences some
emotion / psychological condition, whereas X is a Stimulus / Causer /
Cause / Target, which has contributed to this specific state or become a
target of it.
174 Chapter Four
(4.1) Conventionality of psychological idioms:
a. have a yellow streak / belly down Y’s back “to fear”1

b. show the white feather “to behave in a cowardly manner” ĺ “to
fear”
c. get Y’s Irish up “to become angry or outraged” ĺ “to annoy”
d. drive Y bananas (batty / nuts/ bonkers / crazy) “to cause someone
to feel out of his/her mind” ĺ “to annoy”
e. upset the applecart “to cause upset by ruining plans or
arrangements; spoiling something” ĺ “to depress”
f. give Y the blues ĺ “to depress”
g. knock Y for six “to upset or overwhelm someone completely” ĺ
“to depress”
h. carry a torch for X “to be in love with, especially to suffer from
unreciprocated love for” ĺ “to love”
i. go a bundle on X “ to be very keen on” ĺ “to love”
j. paint the town (red) “to go out and celebrate” ĺ ”to enjoy”
k. raise the roof “to show great enthusiasm” ĺ ”to enjoy”
l. float Y’s boat “to appeal to, make someone excited” ĺ “to
appeal.”
Analysing the psychological idioms in (4.1a-l), the dimension of

conventionality may be recognised in the specific contextual selection of
the particular idiomatic components which bear some conventionally
significant connotation. Thus, to express “fear” we say have a yellow
streak / belly down Y’s back, as in (4.1a), but never have a buttery smudge
/ stomach down Y’s backbone, because “a yellow streak” is associated
with a trait of cowardice in a person’s character, which dates back to 1910-
1915.2 Likewise, we can say to show the white feather to mean “to behave
in a cowardly manner,” as in (4.1b), for the reason that “the white feather”
refers to one’s cowardice, which dates back to 1775-1785.
Moreover, to show that someone has become annoyed, one may use
the idiom get Y’s Irish up, as in (4.1c), but not get Y’s Canadian up, since
“Irish” works here to reveal prejudice towards Irish people, which
originated in 1834. The idiom drive Y bananas (batty / nuts/ bonkers /
1
The meanings provided for the idioms refer to psych-verbs corresponding to the
idiomatic phrases. In fact, the exact meanings may differ in the degree from these
basic psychological domains they belong to.
2
The etymological background and the information about the cultural /
conventional origins of idioms have been retrieved from http://www.phrases.
org.uk/meanings/118400.html, in March, 2017.
crazy), as in (4.1d), may be interpreted as “to cause someone to feel out of

his / her mind,” and it dates back to the 1970s.
Besides, the idiom upset the applecart, as in (4.1e), means “to cause
upset by ruining plans or arrangements,” but if changed into distress the
fruit wagon, its idiomatic meaning is lost. The OED traces this expression
back to 1780-1790, and, indeed, this allusive phrase is first recorded by
Jeremy Belknap in The History of New Hampshire (1788) “Adams had
almost overset the apple-cart by intruding an amendment of his own
fabrication on the morning of the day of ratification” [of the Constitution].
Then, the verb “overset” was changed into “upset.”
The state of getting depressed may also be expressed either by saying
give Y the blues, as in (4.1f), since “the blues” is meant as depressed
spirits, despondency, or melancholy, which dates back to 1800-1810; or by
the idiom knock Y for six, as in (4.1g). The latter expression means “to
upset or overwhelm someone completely,” and “six” in this idiom derives
from the highest scoring action in the sport of cricket: where the ball exits
the circle without touching the ground, viz. hitting the ball over the
boundary scores of six runs, the maximum for one shot.
Furthermore, a few conventionally fixed idioms express the state of
loving someone, i.e. carry a torch for X, as in (4.1h), which is traced back
to 1927, and means “to be in love with, especially to suffer from
unreciprocated love for.” Next, the idiom go a bundle on X , as in (4.1i),
with the meaning “to be very keen on,” dates back to Middle English,
when “a bundle” implied a close binding or grouping together. To express
someone’s excitement or enjoyment, the idioms paint the town (red) “to
go out and celebrate,” as in (4.1j), or raise the roof “to show great
enthusiasm,” as in (4.1k), may be used. Both of them originated in the
1880s-90s. Finally, to express that someone or something appeals to a
human Experiencer, the idiom float Y’s boat, as in (4.1l), originating from
Middle English (1100), may be used.
In other words, conventionality is viewed as the difference between the
idiomatic meaning and the meaning of a collocation analysed along the
rules that apply to constituents when they are in isolation from one
another. Conventions are recognised as conditions to use each of the
component parts of the expression, not the phrase as a whole (Nunberg et
al. 1994: 496, 498). What is more, referring to meaning, Nunberg et al.
(1994: 492, 499) propose that conventionality, instead of non-
compositionality, is the defining characteristics of idioms. Once the
meaning is established conventionally, we can predict what each part of an
idiom means. Thus, in contradistinction to many linguists who treat
conventionality and non-compositionality interchangeably while talking
176 Chapter Four
about idioms, Nunberg et al. (ibid.) propose to attach conventionality to

the use of the idiom constituents, rather than to the collocation as a whole,
which results in making the particular component parts, not the entire
idiom, fixed and non-compositional.
4.2.1.2 Inflexibility
The second dimension which characterizes idiomaticity of an idiomatic

expression is its inflexibility, which is recognized by Nunberg et al. (1994)
as “a limited number of syntactic frames or constructions, unlike freely
composed expressions” (Nunberg et al. (1994: 492). More specifically,
concerning syntactic inflexibility, idioms are subject to syntactic (and pos-
sibly also morphological and phonological) constraints. Usually, idioms
which represent grammatically well-formed patterns of phrasal or sentential
structure, are assumed to show both semantic and syntactic coherence of
their components. As a result, there are also fewer possibilities with
respect to recursion, movement, transformations, modification, and other
syntactic processes which normally could be applied to these constituent
items. To be precise, the constituents of many idioms cannot take enclitics,
be replaced by pronouns or be independently modified, nor can they be
raised (i.e. they cannot change position, as required by another syntactic
construction). Thus, some syntactic modifications make an idiom sound
ungrammatical, as exemplified in (4.2a-b). Besides, the presence of
particular morphemes or a particular grammatical category (tense, number,
person etc.) may be required in some idioms, as shown in (4.2c-d).
(4.2) Inflexibility of certain psychological idioms:
a. upset the applecart “to depress”

*the applecart was upset3 (passivization)
b. paint the town (red) “to enjoy”

# paint the old town (internal modification)
c. scare the pants off Y “to horrify”

*scare a pant off Y (number of the DP)
3
The sign * [asterisk], as explained earlier, will be used to mark the sentence /
phrase as ungrammatical, while # [hash] indicates the loss of idiomatical
interpretation.
d. scare the bejesus out of Y “to horrify”

*is scaring the bejesus out of Y (aspect of the V)
Nevertheless, the instances listed in (4.2a-d) are only a few examples to

prove the inflexibility of certain psychological idioms. More space is
devoted to the discussion concerning the syntactic analysis of
psychological idioms in section 4.3 of this chapter.
On the other hand, it is worth mentioning that some researchers (e.g.
Nunberg et al. 1994) argue that a number of (lexical) idioms may be
syntactically manipulated by speakers, making an idiom acceptable in a
given form.
4.2.1.3 Analysability and transparency
Analysability and transparency also characterise idiomaticity. Analysability

refers to the degree to which the component concepts encoded by an
idiomatic phrase may be used to access suppositions in memory
contributing to the derivation of the proposed reading. More specifically,
semantic analysability defines the extent to which a speaker can
comprehend the meaning of the idiom as a whole, on the ground of the
information conveyed by the constituents of the idiom. The range to which
idiom elements contribute to their overall reading has been termed
semantic decomposition (cf. Nunberg 1978) or their analysability (cf.
Cacciari and Tabossi 1988; Gibbs 1994; and Glucksberg 1991; among
others).
Langacker (1987) argues that the concept of analysability is subtle.
According to him, analysability does not refer to “the intrinsic complexity
of a structure, but rather to a person’s awareness of certain aspects of this
complexity” (Langacker 1987: 457). By presuming that “word sequences
are stored in memory and accessed as units” (Bybee 2006: 714), Bybee
underlines the significance of frequent repetition of an idiomatic
expression, and the consequence it has for improved analysability. This
fact has been proven by some usage of idioms, where speakers were to
provide just the first part of the expression, assuming that the hearer will
complement it mentally and will access the whole meaning of the idiom
(Bybee 2006; and CieĞlicka 2004, 2006; among others).
In the same vein, Nunberg et al. (1994: 496) state that once the
meaning of an idiom is acknowledged, e.g. by hearing it in a sufficiently
informative context, the meaning is not devolved on the constituents of the
expression. However, this does not entail simple non-predictability or non-
analysability of the idiom on the basis of the knowledge of the meanings
178 Chapter Four
of its parts. Instead, idioms are viewed to be listed as phrasal entries in the
lexicon, associating each entry directly with a single semantic representation.
Furthermore, once an idiom is recognised, we are able to establish
correspondences between the parts of the structured denotation of the
phrase, e.g. the relation of making someone dispirited and depressed in the
idiom give Y the blues in (4.1f), and the components of this idiom (give
and the blues), in such a way that each constituent is to be predictable in
terms of metaphorical reference to an element of the interpretation. Thus,
the idiom is given a compositional, i.e. idiosyncratic, analysis (cf. Nunberg
et al. 1994: 496).
In addition, a lot of experimental research has been done concerning
the variability in the degree of idiomatic compositionality (analysability).
As a result, it has been shown that people appear to have strong intuitions
enabling them to judge an idiom as being decomposable or non-
decomposable (Gibbs and Nayak 1989). It can be easily identified how
each part of the idiom combines in the derivation of the figurative
interpretation for decomposable / compositional idioms (e.g. pull strings,
cast a spell on Y, get cold feet), but not for non-decomposable / non-
compositional idioms (e.g. kick the bucket, raise the roof, paint the town
(red)).
While analysability defines the range to which idiom components
contribute to the overall idiomatic interpretation; transparency refers to
the relative ease with which any assumptions about the meaning of an
idiom are accessed and implications derived. Idiom transparency is
sometimes called “literalness” of an idiom, and is also identified by some
scholars as the degree to which the meaning of the idiom can be derived
from the phrase. Thus, analysability and transparency are so closely
related that they are mostly used interchangeably.
According to CieĞlicka (2004: 98), Mäntylä (2004: 28-29), and
Peacock (2009: 2), among others, there are directly / fully transparent
idioms, for which the literal meaning of their constituents is clearly linked
to the figurative overall meaning, e.g. give the green light “to accept
something.” These idioms display a high degree of analysability and
transparency. However, in the case of semi-/ partially / relatively
transparent idioms, the literal meaning of their components gives only
some hint of the figurative meaning (e.g. quake in your shoes “to fear”)
but the link is not as noticeable as with fully transparent idioms. Finally,
there are (transparently) opaque (semi-opaque) idioms, in which the
motivation behind the figurative meaning is hard to perceive unless the
etymology is already known, e.g. be home and dry “to succeed.”
In (4.3) below, there are examples of compositional psychological

idioms, which display either a high or partial degree of semantic
analysability and transparency.
(4.3) Examples of directly / fully transparent psychological idioms,

with a high degree of analysability and transparency:
a. have a ball ”to have a pleasant time”ĺ “to enjoy”

b. set Y’s heart on X “to put someone’s emotional involvement
into something / someone” ĺ “to love”
c. turn Y’s nose up at X “to turn someone’s attention and
enthusiasm against something / someone; to reject” ĺ “to
hate”
d. bear / owe a grudge against X “to continue feeling an old
resentment for someone / something” ĺ “to hate”
e. catch Y’s eye / catch the eye “to draw someone’s attention”
ĺ “to fascinate.”
Examples of partially analysable and transparent psychological

idioms:
f. give Y the pip “to cause someone to feel depressed” ĺ “to

depress”
g. bring Y low “to cause someone to feel dispirited and
depressed” ĺ “to depress”
h. whet Y’s appetite “to evoke excitement in someone” ĺ “to
appeal”
i. make a difference to Y “to be distinctive while compared to
someone / something else” ĺ “to matter.”
Some psychological idioms. as in (4.3a-e), are fully transparent, which

means that much of their meaning may be comprehended if they are taken
literally; thus, their meaning is predictable on the basis of their constituent
items. Others (4.3f-i) may not be entirely literally interpretable, but require
a slight metaphorical broadening.
Importantly, as argued by Vega-Moreno (2005:395-396), transparency
and opacity are not fixed properties of idioms, but, instead, should be
treated as semantic dimensions, which differ between individual language
users in a particular context. Indeed, whether an individual perceives an
idiom as more or less transparent, would largely depend on the
connotations available to him / her at the time, and their degree of
accessibility. Thus, the more implications derivable from the literal
180 Chapter Four
meaning of the phrase, the more transparent the idiom seems to be, and the
easier the idiomatic interpretation will be obtained. For example, in the
idiom spill the beans “to divulge a secret,” spill directly implies “divulge,”
and the word the beans refers to the noun “a secret,” as a result of the
long-lasting convention concerning this expression. McGlone, Glucksberg
and Cacciari (1994) propose that the idiom spill the beans is more
transparent than an alternative expression spill the mud, which may stem
from the fact that the beans make a great contribution to the idiom
meaning in such a way that beans, like “secrets,” are many and countable.
Finally, it is decomposable / compositional idioms which are more
flexible (syntactically, semantically, and lexically) than non-decomposable
ones. As argued by Gibbs and Nayak (1989), Gibbs et al. (1989a,b),
Glucksberg (1993, 2001), among others, the more analysable and
transparent an idiom is, the more syntactic flexibility it is expected to
reveal, which is to be discussed in section 4.3 below.
4.2.1.4 Figuration, proverbiality, informality and “affect”
Besides compositionality, conventionality, analysability and transparency,

Nunberg et al. (1994) discuss several other properties of idioms, namely
figuration, proverbiality, informality, and affect (Nunberg et al. 1994: 492-
493, 498).
Figuration refers to the fact that most idioms pertain to so-called
“figures of speech,” like metaphors, metonymies and hyperboles (Nunberg
et al. 1994: 492). According to The Merriam-Webster Collegiate
Dictionary,4 in a metaphor, a word or phrase literally denoting an object or
idea is used, instead of another, to express likeness or analogy between
them. One of the most frequently cited examples of a metaphor in English
literature is “All the world’s a stage” from William Shakespeare’s As You
Like It. In (4.4) below, psychological idioms function as metaphors since
the objects which are the constituent elements of the idioms draw an
analogy with the emotional / psychological state, e.g. “steam” is analogous
to “joy” in blow / let off (some) steam “to enjoy,” or “one’s temper and
patience,” which might be lost when someone (an Experiencer) is enraged
or irritated, is compared to “fire /flame,” which is fanned in the idiom fan
the fire / flame to mean “to anger someone.”
4
The online version of The Merriam-Webster Collegiate Dictionary, available at
https://www.mer riam-webster.com/ was retrieved on March 11, 2017.
(4.4) Psychological idioms as metaphors:

blow / let off (some) steam “to enjoy”
fall head over heels in love with X “to love”
have butterflies in Y’s stomach “to worry”
knock Y’s socks off “to surprise”
fan the fire / flame “to anger”
bring a hornet’s nest round Y’s ears “to annoy”
rattle Y’s cage / chain “to annoy”
ruffle Y’s feathers “to annoy”
take the wind out of Y’s sails “to depress.”
A metonymy, as defined by The Merriam-Webster Collegiate Dictionary,

is a figure of speech consisting of a thing or concept which is referred to
by the name of something closely associated with that thing or concept,
e.g. “crown” in “lands belonging to the crown.” The psychological idioms
in (4.5) below work as metonymies because the idiomatic components
evoke some connotation with some different concept, such as “the white
feather” in show the white feather “to fear” or “yellow streak” in have a
yellow streak / belly down Y’s back “to fear” refer to cowardice; “Irish” in
get Y’s Irish up “to annoy” evokes human prejudice and annoyance
towards the Irish; and “the blues” in give Y the blues ”to depress” or “six”
in knock Y for six “to depress” are associated with melancholy, failure and
depression.
(4.5) Psychological idioms as metonymies:

have a yellow streak / belly down Y’s back “to fear”
get Y’s Irish up “to annoy”
show the white feather “to fear”
give Y the blues “to depress”
knock Y for six “to depress.”
Accordingly, both metonymy and metaphor involve the substitution of one

term for another; but in metaphor, the substitution is based on some
specific analogy between two things, while in metonymy the substitution
concerns some understood association.
Finally, a hyperbole as a figure of speech is defined, according to the
Online International Dictionary,5 as an exaggerated statement or claim,
not meant to be taken literally. Hyperboles are often used in casual speech
5
Online International Dictionary available at http://idict.org, retrieved on March
11, 2017.
182 Chapter Four
as intensifiers. More specifically, in the example “the bag weighed a ton,”

the hyperbole stresses the fact that a speaker found the bag to be extremely
heavy, although it could not be as heavy as a literal ton. Besides,
hyperboles are said to relay emotions, and serve as, e.g. a form of humour,
excitement, or distress. The examples of psychological idioms which
realise the role of hyperboles are illustrated in (4.6) below.
(4.6) Psychological idioms as hyperboles:

eat Y’s heart out “to worry”
cut Y to the quick / bone “to annoy”
curdle / chill Y’s blood “to scare”
put the screws on Y “to horrify”
Indeed, as seen in (4.6), to eat Y’s heart out is an exaggerated way to mean
that someone is worried. In the same vein, trying to express somebody’s
annoyance the idiom cut Y to the quick / bone reveals someone’s psychical
and emotional condition clearly enough. To mean that someone or
something scares or horrifies a person, the idioms curdle / chill Y’s blood
or put the screws on Y may be used, and the emotional state is almost
visibly exposed.
Another feature defining idiomaticity is proverbiality. It concerns the

typical use of idioms for the description of a recurring situation of
particular social interest (becoming restless, talking informally, divulging
a secret) on account of its resemblance to a familiar scenario (Nunberg el
al. 1994: 493). Defined as short concise, frequent and widespread sayings
which contain wisdom, truth, moral or practical guidelines, proverbs have
a fixed and memorisable form, and are usually handed down from
generation to generation.6 Interestingly, Norrick (1985: 73) lists some
distinctive features, according to which prototypical proverbs differ from
idioms, clichés, etc., even though the latter may bear some characteristics
of proverbiality.
There is a relatively big number of psychological idioms which may
work as proverbs, making reference to some social issues, e.g. hatred, fear,
scaring or horrifying others, depressing others, or filling others with anger.
To be precise, in some idioms the psychological states may refer strictly to
the personal condition experienced by an Experiencer himself or herself.
6
Retrieved from American Heritage Dictionary of the English Language,
available at https://www. ahdictionary.com/word/search.html?q=online on March
11, 2017.
However, emotions may also (if not always) occur in an Experiencer in

relation with others (other humans or entities), regardless of the fact
whether the psychological state or emotion is evoked in an Experiencer by
some Stimulus / Cause / Causer (in OE psych-idioms), or if an
Experiencer in the role of a subject directs his / her emotion towards
Theme / Goal (in SE psych-idioms). In such a relationship, rarely does it
happen that the emotion is held as a matter of privacy, locked within one’s
private inner sphere. More often, the emotion is distinguishable by others,
may become noticeable by the surroundings, and even lead to a social
problem, especially when the emotion is extremely negative, expressed in
a problematic manner or directed towards a significant group of recipients
/ citizens. Accordingly, in (4.7) some instances of psychological idioms
connected with hatred, fear, scaring or horrifying others, depressing
others, or filling others with anger are provided.
(4.7) Proverbiality of psychological idioms:

a. the social problem of hatred expressed in the following idioms:
pour scorn on X
bear ill will toward X
bear / owe a grudge against X
bear / feel aversion / malice / hostility / repugnance toward
(to) X
have no stomach / taste / use for X
show contempt for X
hold X in abomination / contempt
turn Y’s nose up at X
look down (Y’s) nose at X
b. the social problem of fear expressed in the following idioms:

take dim view of X
have Y’s heart in Y’s mouth
c. the social problem of worry expressed in the following idioms:

eat Y’s heart out
lose sleep over X
have a bee in Y’s bonnet
184 Chapter Four
d. the social problem of scaring or horrifying others expressed in the

following idioms:
curdle / chill Y’s blood
frighten / scare the life / the hell out of Y
put the fear of God into Y
chill Y to the bone/marrow
send Y into a cold sweat
throw /send Y into a panic
shake X’s / a fist at Y
give Y goose pimples/ bumps
give Y (quite) a (bit of) turn / a fright
freeze Y’s (the) blood
put the screws on Y
scare the shit / the wits out of Y
frighten / scare the pants off Y
scare the bejesus out of Y
frighten / scare Y to death
frighten / scare Y out of their wits
give Y a turn
e. the social problem of depressing others expressed in the following
idioms:
break Y’s heart / spirit
damp Y’s hopes
dash / wither Y’s hopes
do a number on Y
knock the stuffing out of Y
put a damper on Y
bring tears to Y’s eyes
cut Y down to size
take/ knock Y down a peg / notch (or two)
put / send / throw Y into a (blue) funk
leave Y in the / a lurch
bring Y into disrepute
knock Y for six
bring Y low
knock Y sideways
f. the social problem of anger, viz. filling someone with anger; expressed
in the following idioms, e.g.:
kindle Y’s wrath
get Y’s back / dander up
put / set Y’s back (up)
blow a gasket on Y
breathe fire over Y
drive Y out of mind
drive Y to distraction
drive Y round the bend/ twist
vent X’s spleen at Y
Another property of idioms, discussed by Nunberg et al. (1994: 493),

is informality, which refers to the tendency of idioms to appear in collo-
quial registers, in popular speech and oral culture. As defined by The
Merriam-Webster Collegiate Dictionary,7 “colloquial” is related to
registers characteristic of familiar and informal conversation. Thus, in
colloquial English, “kind of” is often used for “somewhat” or “rather;” a
greeting “what’s up?” between friends sounds more natural and real and
appropriate than the formal “How are you?” or “How do you do?;” while
idioms sound better in popular and oral discourse than in informal speech
or writing.8
In addition, a colloquial register is the variety of language that speakers
usually use when they are relaxed and not particularly self-conscious;
therefore, some colloquial speech may contain a great deal of slang,
contractions or even profanity (cf. Trask 1999). Nonetheless, colloquial
expressions, in a piece of literature, may provide deep insights into the
writer’s society, and the real language they use. For that reason, colloquial
phrases bring a sense of realism to a piece of literature, which, in turn,
draws readers’ and listeners’ attention since colloquialisms are identified
7
https://www.merriam-webster.com/ retrieved on April 12, 2017.
8
Importantly, as commented by Prof BoĪena Cetnarowska, some idioms are
currently used less frequently, e.g. in the Oxford English Dictionary (online
version https://en.oxforddictionaries.com), the idiomatic phrase give someone a pip
is marked as “dated,” while give someone a turn and pay court to someone are
noted by the Longman Dictionary (http://www.ldoceon line.com/dictionary) as
“old fashioned.”
186 Chapter Four
with their real life. Besides, colloquial expressions add variety to

characters, which makes them more remarkable and unforgettable.9
In order to check whether psychological idioms occur relatively more
frequently in colloquial / spoken English than in formal language, the
Corpus of Contemporary American English has been searched for a few
psych-idioms listed in (4.8) below. Here, some psychological idioms have
been compared in terms of their occurrence in different types of register,
namely those offered by the COCA, viz. spoken discourse, fiction, popular
magazines, newspapers, and academic texts. With its data containing more
than 520 million words of text (20 million words each year 1990-2015),
the COCA Corpus appears to work as a reliable source for researchers to
compare data across registers, and across different periods of time.
(4.8) Informality of psychological idioms:

a. blow off steam “to enjoy”
SECTION
FREQ SIZE (M) PER MIL
Type of register
spoken 17 109.4 0.16
fiction 24 104.9 0.23
magazine 38 110.1 0.35
newspaper 27 106.0 0.25
academic 7 103.4 0.07
b. ruffle Y’s feathers “to annoy”

SECTION
Type of register
spoken 11 109.4 0.10
fiction 9 104.9 0.09
magazine 16 110.1 0.15
newspaper 31 106.0 0.29
academic 4 103.4 0.04
c. carry a torch for X “to love”

SECTION
Type of register
spoken 5 109.4 0.05
fiction 20 104.9 0.19
magazine 7 110.1 0.06
newspaper 6 106.0 0.06
academic 1 103.4 0.01
9
Literary Devices Editors. 2013. Colloquialism. Retrieved April 12, 2017, from
https://literarydevices. net/ colloquialism/
d. paint the town “to enjoy”

SECTION
Type of register
spoken 10 109.4 0.09
fiction 8 104.9 0.08
magazine 12 110.1 0.11
newspaper 10 106.0 0.09
academic 4 103.4 0.04
As can be seen in (4.8), the most common registers that idioms are likely
to occur in are newspapers, magazines, fiction, and spoken discourse.
Academic sources, in turn, record the lowest scores of the idioms under
scrutiny. In fact, newspapers, magazines, and fiction are not taken into
consideration, since there is no clear-cut distinction between the sources
written in colloquial language and the ones which contain more scientific
and formal language. In what follows, having compared the spoken
discourse to the academic text the psychological idioms occur in, the
superiority of the former over the latter type of register may be evidently
noticed. That is why, the assumption that informality is one of the features
of idiomaticity, including psychological idioms, has been confirmed.
The last dimension of idiomaticity, mentioned by Nunberg et al. (1994:

493), is called “affect.” It stems from the fact that idioms, in the majority
of cases, imply a certain evaluation or affective attitude toward the things
they denote (ibid.). According to The Merriam-Webster Collegiate
Dictionary, affect is “a set of observable manifestations of a subjectively
experienced emotion.”10 The noun affect derives from late 14th-century
Middle English affect, which means “mental state,” and from Latin
affectus, adfectus, denoting a disposition, mood, state of mind or body
produced by some external influence, especially sympathy or love.11
While taking psychological idioms into account, there seems to be no
doubt that they are expected to exhibit this idiomatic dimension, since
affection and emotional and / or psychical attitude is the core of their
nature. Indeed, in most cases the relation between an Experiencer (Y) and
a Stimulus / Cause / Causer or a Theme / Goal (X) of the emotional state is
clearly present, as illustrated in the instances in (4.9a) below.
10
https://www.mer riam-webster.com/ retrieved on April 12, 2017.
11
Online Etymology Dictionary, available at http://www.etymonline.com retrieved
on April 12, 2017.
188 Chapter Four
(4.9) Affect” of psychological idioms:

a. [Y] have eyes for X “to love”
[Y] show contempt for X “to hate”
[Y] take dim view of X “to fear”
[Y] lose sleep over X “to worry”
[X] get a rise out of Y “to annoy”
b. [Y] paint the town (red) “to enjoy”

[Y] raise the roof “to enjoy”
[Y] have goose bumps “to fear”
[Y] turn tail (and run) “to fear”
[Y] have the blues ”to worry.”
However, in some psychological idioms, as exemplified in (4.9b) above,

the Causer / Cause / Stimulus/ Theme / Goal (X) is covert. In short, the
“affect,” i.e. any affective relationship in which an Experiencer interacts
with the Causer / Theme, is clearly stated for the former group of
psychological idioms, in (4.9a), but concealed for the latter one, in (4.9b),
even though this kind of emotional affect presumably exists.
In a nutshell, as argued by Nunberg et al. (1994), apart from the
dimension of conventionality, none of these properties have to apply
compulsorily to all idioms. As far as psychological idioms are concerned,
all of them are associated with certain conventional connotations.
However, there are some psych idioms which do not involve figuration,
but reveal a high degree of analysability and transparency (e.g. set Y’s
heart on X “to put someone’s emotional involvement into something /
someone” ĺ “to love;” or turn Y’s nose up at X “to turn someone’s
attention and enthusiasm against something / someone; to reject” ĺ “to
hate”). Others are highly metaphorical (e.g. ruffle Y’s feathers “to annoy;”
take the wind out of Y’s sails “to depress;” or have a yellow streak / belly
down Y’s back “to fear”), have some proverbial allusion (e.g. drive Y out
of mind “to anger;” or leave Y in the / a lurch “to depress”), or overtly
refer to some affective scenario (e.g. lose sleep over X “to worry;” or [X]
get a rise out of Y “to annoy”). Finally, most idioms represent an informal
register and a colloquial type of discourse (e.g. paint the town “to enjoy;”
or carry a torch for X “to love”).
Importantly, it is the triple semantic distinction based on compositionality,
conventionality, and transparency, which is most commonly referred to in
the literature while searching for the indicators of idiomaticity. As briefly
summarised by Titone and Connine (1999: 1663-1664), compositionality
is inferred from the degree to which the phrasal meaning, once recognised,
can be analysed in terms of the contributions of the constituents of the

idiom; conventionality refers to the degree to which idiomatic meanings
are not predictable from the word components left in isolation, and
knowledge of the conventions of a particular language environment;
finally, transparency concerns the degree to which the original motivation
of these phrases is immediately accessible.
4.2.2 Idiomatically combining expressions vs. idiomatic phrases
The Principle of Semantic Compositionality (sometimes called “Frege’s

Principle” (1884)), according to which the meaning of a (syntactically
complex) whole is only a function of the meanings of its (syntactic) parts
placed together and the rules used to combine them, is contradicted by
idioms (cf. Fraser 1970; Katz 1973; Chomsky 1980; Machonis 1985;
Schenk 1994; and Grégoire 2009; among others). Indeed, in every
language, there are such idiomatic expressions, as illustrated in (4.10a-c),
conveying the meaning that does not comprise (the combination of) the
meanings of the individual lexical items of that expression.
(4.10) Examples of idiomatic phrases (IdPs):
a. kick the bucket “to die”

Didn’t you hear? He kicked the bucket–had a heart attack, I
think.
(Google)
b. raise the roof “to show great enthusiasm” ĺ “to enjoy”
The whole college is ready to raise the roof at next weekend’s
homecoming celebrations.
(Google)
c. paint the town (red) “to go out and celebrate” ĺ
“to enjoy”
After the show, we went out to paint the town red. We’d been
sitting at an outdoor cafe, drinking rum.
(COCA)
In (4.10a), there is a canonical idiomatic phrase, usually cited in the

literature, the meaning of which is completely not derivable from its
components, i.e. neither “kicking” nor “buckets” contribute to the overall
meaning of the phrase, which is “to die.” Likewise, the particular lexical
items comprising the psychological idiomatic expressions in (4.10b-c), viz.
“raise” and “roof,” or “paint” and “town,” do not account for the whole
190 Chapter Four
meaning of these idioms, which corresponds to the psychological verb “to

enjoy.” Nunberg, Sag and Wasow (1994) call these expressions idiomatic
phrases (IdPs), which do not distribute their meanings among their
components. Instead, the whole VPs of kick the bucket, raise the roof, and
paint the town (red) are related to the overall interpretation of “to die,” “to
show great enthusiasm,” and “to go out and celebrate,” respectively.
Furthermore, Nunberg et al. (1994) distinguish idiomatic phrases
(IdPs) from the other class of idioms, i.e. idiomatically combining
expressions (ICEs). This group, exemplified in (4.11a-c), is usually
referred to as compositional idioms, and it is far more extensive. The
meaning of these idioms is predictable on the basis of their constituent
elements, but it is often not a sum of the meanings of all their lexical
items.
(4.11) Examples of idiomatically combining expressions (ICEs):
a. pull strings “to use connections”

You yourself pulled strings so that they’d transfer me to your
department. You spoke with the board.
(COCA)
b. cast a spell on Y “to intrigue and delight someone”
ĺ “to fascinate”
She is a real beauty. She cast a spell on every man she met.
(Google)
c. give Y the pip “to severely annoy or dispirit someone” ĺ “to
annoy”
That sort of talk gave Jimmy the pip.
(Google)
The typical example of idiomatically combining expressions is pull strings

(“to use connections”), cited in (4.11a), the overall interpretation of which
is distributed among its parts, even though these are associated with
conventional meanings, viz. pull ĺ use, and strings ĺ connections.
Similarly, the psychological idiomatically combining expressions, listed in
(4.11b-c), comprise lexical items that contribute to the general meaning of
these idioms. The meaning of cast a spell on Y “to intrigue and delight
someone” (in 4.11b) is totally predictable from the meaning of its
particular constituents; whereas the idiom give Y the pip “to severely
annoy or dispirit someone” in (4.11c) refers to a conventional connotation,
i.e. “pip” meaning a bad temper or depression.
Moreover, it is generally acknowledged that idioms do not form a

homogeneous class but rather a highly heterogeneous one which lies on “a
continuum of compositionality” (Vega-Moreno 2003). Idioms differ as
regards the extent to which the meanings of their individual components
contribute to the overall figurative interpretation. At one end of the scale,
there are extremely flexible idiomatically combining expressions which
are derived fully compositionally from the meanings of their constituents.
At the other end, there are highly fixed idiomatic phrases, recognised as
non-compositional idiom strings, whose individual elements are in an
entirely arbitrary relation to the overall idiom meaning. However, the
majority of idioms take the position of somewhere between these
extremes, exhibiting a variety in the degree and ways in which the internal
semantics of their components is derived.
Even though different typologies of idiomatic expressions have been
proposed (Cacciari and Glucksberg 1991; and Nunberg et al. 1994; among
others), for the sake of this book, I follow the proposal, offered by
Nunberg et al. (1994), and Harwood et al. (2016), that there are two types
of psychological idioms, i.e. (i) idiomatically combining expressions, and
(ii) idiomatic phrases, as illustrated in (4.11b-c) and (4.10b-c),
respectively. Idiomatically combining expressions (ICEs) have meanings,
even conventional ones, which are distributed among their components,
and the particular elements of these literal (compositional) expressions can
be mapped onto the elements of the figurative meaning. Besides, DPs of
the idioms have a referent, e.g. strings mean connections, and pip means
sickness; and the nouns are used metaphorically, making a collocation
with a particular verb. On the other hand, idiomatic phrases (IdPs) do not
distribute their meanings onto their constituents, and they form a whole
unit which is mapped onto the figurative meaning.12
12
Interestingly, it should be repeated here, for the sake of convenience that, taking
into account the relation between the form of an idiom and its meaning, Gibbs and
Nayak (1989), and Nunberg (1978), among others, distinguish three types of
idioms. First, when a one-to-one relation occurs, i.e. a relation in which each word
contributes independently to the figurative interpretation (e.g. the semantic relation
between “pop” and “utter” and “question” and “marriage proposal” in pop the
question), the idioms are typically known as “normally decomposable” idioms.
Second, they mention “abnormally decomposable” idioms, in which there is an all-
to-one relation with the (literal) meaning of the whole phrase being semantically
connected with the figurative interpretation (e.g. bury the hatchet, and push the
panic button). Finally, in case of “non- decomposable” idioms, the relation may be
none-to-one in that the component words neither individually nor as a whole are in
a semantic relation to the idiomatic meaning (e.g. chew the fat, and break a leg).
Moreover, the relation may differ in terms of transparency between the constituent
192 Chapter Four
In brief, idioms whose overall idiomatic interpretation is derivable

(normally or abnormally, literally or figuratively) from their component
parts are generally known as decomposable / compositional or analysable.
These are named here idiomatically combining expressions. While idioms
with the overall idiomatic interpretation not derived from the constituent
parts are non-decomposable / non-compositional, frozen, opaque, and
unanalysable, and these are typically referred to as idiomatic phrases.
Furthermore, with the aforementioned triple semantic distinction of
idiomaticity in mind, viz. the one based on compositionality,
conventionality, and transparency, idiomatic phrases (IdPs) are assumed
to differ from idiomatically combining expressions (ICEs) in having
essentially a lower compositionality, a higher conventionality, and a lower
transparency (cf. Espinal and Jaume 2010: 1399). Besides, idiomatically
combining expressions (ICEs) are not obligatorily required to be
transparent, i.e. providing a speaker the reasoning for the figural
interpretation they involve (though ICEs are mostly transparent and
analysable). What is only essential is a correspondence between the
expression and the relevant element of the idiomatic denotation that can be
established. In turn, the idiomatic interpretations of idiomatic phrases
(IdPs), cannot be distributed over their parts; thus, they must be entered in
the lexicon as complete phrases (cf. Nunberg et al. 1994). Nevertheless,
the class of idioms defined by the criterion of predictability comprises a
far bigger group than the class of idiomatic phrases.
Finally, it must be admitted that the border line between those two
types of idioms, i.e. idiomatic phrases (IdPs) and idiomatically combining
expressions (ICEs), is not clear cut. Zhu and Fellbaum (2015: 339) argue
that Corpus data have revealed that fully non-compositional idioms are not
frozen, and semantic compositionality and variation are in fact independent of
each other. Dobrovol’skij and Piirainen (2005: 14), in turn, treat idioms as
a radial category, which comprises fully non-compositional, frozen idioms
as prototypes, and partly compositional idioms which “radiate out” i.e.
deviate from the prototype, when they have lexically substituted and
morphosyntactically operated components.
Because a clear division between idiomatic phrases (IdPs) and
idiomatically combining expressions (ICEs), based on the aforementioned
semantic characteristics of idiomaticity, is hardly possible to make, some
valid syntactic diagnostics should be established to make the distinction
words contributing to idiom meaning, viz. an idiom may be literally transparent

(e.g. “miss” in miss the boat), metaphorical (e.g. “blow” in blow one’s stack),
hyperbolical (e.g. eat one’s heart out), or not deriving the meaning of an idiom at
all (e.g. kick the bucket, chew the fat, and shoot the breeze).
explicit. For Wasow et al. (1984), Fillmore et al. (1988), Nunberg et al.
(1994), Everaert et al. (1995), Harwood et al. (2016), and Corver et al.
(2017), the extent to which idioms can be syntactically and lexically
modified accounts for the difference between IdPs and ICEs. Especially
the possibility for an idiom to undergo passivization, topicalization, and
adjectival modification, while leaving the idiomatic interpretation intact,
are the most commonly cited diagnostics to distinguish IdPs from ICEs, as
these tests produce the most categorical results. The tests on the level of
conventionality, compositionality or opacity of an idiom are, in turn, far
more ambiguous and a matter of one’s individual interpretation (cf. Corver
et al. 2017: 12). Consequently, IdPs are expected to remain completely
resistant to any syntactic or lexical modifications, in contradistinction to
ICEs, which are to show a considerable degree of syntactic and lexical
flexibility (cf. Nunberg et al. 1994; and Gibbs et al. 1995; among others).
The issue of syntactic and lexical variability of idioms is to be discussed
more thoroughly in section 4.3 of the chapter.
4.3 Syntactic and semantic variability of idiomatically

combining expressions (ICEs) with psychological meaning
As regards compositional idioms, viz. idiomatically combining expressions
(ICEs), the question arises what sort of syntactic, semantic, and lexical
modifications this type of idioms can undergo. It is worth noting that parts
of idiomatic phrases (IdPs) can also be occasionally modified but are
subject to certain restrictions. Having studied the case of Italian idioms,
Vietri (2014: 89) makes a general observation that modification of idiom
parts is relevant either for the (supposed) decomposable or non-decomposable
idioms.
With this in mind, the purpose of section 4.3 is to make an attempt to
find out what types of syntactic and semantic modifications psychological
idiomatically combining expressions may undergo. The instances of the
possible modifications have been taken either from the COCA Corpus, or
obtained via the Google Search. In addition, some value judgements have
been pronounced by native speakers,13 which is always indicated in
parentheses below the cited examples. If a sentence is judged by native
13
There have been 20 native speakers who have been asked for their judgements.
Two of them (10%) were aged 20-25, the other two (10%) were in their 30s-40s,
whereas the remaining sixteen (80%) were aged 60-70. All of them either teach
English presently or used to do so before they retired.
194 Chapter Four
speakers as weird or rarely used, it is marked with a question mark /

question marks [?].
4.3.1 Alternations affecting the idiomatic

object NP in psychological idioms
The first type of variability attested in idiomatically combining
expressions refers to the alternations affecting the idiomatic object NP,
including pronominalization of the object, called anaphoric reference
(4.12), quantification of the object (4.13), changing the number of the
object (4.14), and modification of the object by means of an intensifier
(4.15), or an adjective (4.16)-(4.18).
To begin with, the existence of co-reference relations between
pronouns and parts of idiomatic expressions, commonly known as
anaphora, anaphoric reference, or pronominalization of the object, has
been noted by Chomsky (1981), Bresnan (1982), Langacker (1987), and
Nunberg et al. (1994), among others. While Bresnan (1982: 49), cited in
Nunberg et al. (1994: 502), argues that genuine idiom chunks may not
serve as antecedents for pronouns; Nunberg et al. (1994), find at least
some idiom chunks, which are possible antecedents for pronouns.14
Based on native speakers’ judgement and the data found via the
Google Search, a few psychological idioms with anaphoric reference have
been found. In the case of these phrases, a pronoun occurring later in the
sentence may refer back to a noun which belongs to the core of an
idiomatic expression. Therefore, in (4.12a) below, we have the pronoun
them referring back to the idiomatic NP butterflies in the idiom have
butterflies in Y’s stomach “to worry and fear.” The context of the passage
below gives no other reference possibilities except for the idiomatic object
NP “butterflies.” Likewise, in (4.12b), the NP object curves of the idiom
throw Y a curve “to surprise” is referred to by the pronoun they in the
subsequent clause. In (4.12c), the idiomatic NP object goose bumps in the
idiom have / get goose bumps “to worry” is replaced with the pronoun
they.
14
For further discussion concerning anaphora see Nunberg et al. (1994: 501-503).
(4.12) Pronominalization of an idiomatic object NP in psychological

idioms:
a. have butterflies in Y’s stomach ĺ “to worry and fear”
I did have butterflies in my stomach, for some reason, and ... I
had them in both of my thighs and both of my calves.
(Google)
b. throw Y a curve “to surprise”
Life may throw us many curves, but sometimes they’re just a test
of our faith.
(Google)
c. have / get goose bumps “to worry”15
Why do we get goose bumps? (…) Sometimes goosebumps come
with a fright, sometimes simply when we’re chilly. They don’t
seem to perform any purpose, so why are they there?
(Google)
Moreover, a comparatively common variation concerns quantification

of the object NP. Thus, in (4.13a), the idiomatic NP is quantified by many,
while in (4.13b-e), a more extended range of quantifiers is provided, such
as a bit of, a lot of, no, and quite so much.
(4.13) Quantification of the object NP in psychological idioms:
a. throw Y a curve “to surprise”

Life may throw us many curves, but sometimes they're just a test
of our faith.
(Google)
b. give Y a turn / a fright “to scare”
Recently I had a high blood pressure reading from my doctor. It
gave me a bit of a fright. (Google)
c. carry weight with Y “to matter to”
Your argument does not carry a lot of weight with me.
(Google)
d. cut ice with Y “to matter to”
As before, his deference cuts no ice with Carol, who is even more
merciless than before
(COCA)
15
The expression may be also interpreted literally “to get cold,” referring to a
bodily sensation, but in our analysis of the phrase only its figurative interpretation,
viz. “to worry,” is taken into consideration.
196 Chapter Four
e. hold X in abomination / contempt “to hate”

For some reason he didn’t hold her in quite so much contempt as
he used to.
(Google)
Furthermore, when the number of the object is changed, viz. mainly

when the canonically singular object NP appears as plural, some quantifier
must be altered at the same time to license this variation, e.g. genitive
possessor / pronoun (somebody’s) into zero article [ø], as illustrated in
(4.14a-b), a into many, as in (4.14c).
(4.14) Changing the number of the object NP in psychological idioms:

a. whet Y’s appetite “to appeal”
According to Communist, sportsmen were controlled by
capitalists to whet imperialistic appetites. (COCA)
b. break Y’s heart / spirit “to depress”

That conflict left deep scars. It broke spirits, destroyed cities,
scattered families, and sent thousands of Lebanese into exile.
(COCA)
c. throw Y a curve “to surprise”
Life may throw us many curves, but sometimes they’re just a test
of our faith.
(Google)
Another type of modification is the one with an intensifier inserted

next to an NP object. An intensifier is a modifier that makes no
contribution to the propositional meaning of a clause but aims to enhance
and give additional emotional emphasis to the word it modifies. This is
illustrated in (4.15a-b) for psychological idioms, in which the intensifier
quite is used. Other intensifiers, e.g. fairly, pretty, very, absolutely, or
really, are mostly placed directly in front of adjectives or adverbs to
strengthen their meaning. In (4.15c), the idiomatic adjective soft is
enhanced by the intensifier very in the idiom have a soft spot for X “to
love;” while in (4.15d), the adjective batty, which is a constituent of the
idiom drive Y batty “to annoy,” is modified by the intensifier absolutely.
(4.15) Intensifier modification in psychological idioms:

a. give Y a turn / a fright “to horrify, scare”
You gave her quite a turn, suddenly appearing like that!
(Google)
b. get a rise out of Y “to annoy”

The seat has gotten quite a rise out of some of the unsuspecting
men.
(Google)
c. have a soft spot for X “to love”
She has a very soft spot for young Victoria.
(Google)
d. drive Y batty “to annoy”
You and I both know that James drives me absolutely batty.
(Google)
Moreover, adjectival modification of the object is the most complex of

these variations (cf. Ernst 1981; and Stone 2008, 2016). The most frequent
kinds of adjectival modification, which occur with psychological idioms,
include: (i) external adjectival modification; (ii) internal adjectival
modification; and (iii) conjunctive adjectival modification, as exemplified
in (4.16), (4.17), and (4.18), respectively.
(4.16) External adjectival modification of the object NP in psychological

idioms:
a. raise Y’s hackles “to annoy”

Children’s health insurance bill in the current Congress has also
raised conservative hackles.
(COCA)
b. scare the shit / the wits out of Y “to horrify”
When all of a sudden my cell phone rings and scares the total shit
out of me.
(COCA)
c. take a fancy / a liking / a shine to X “to love”
She took a sudden liking to him.
(Google)
d. cast a gloom / a shadow over Y “to depress”
The TB-ridden slums cast a horrifying shadow on the comfortable
neighbour-hoods around them.
(COCA)
External modification, illustrated in (4.16a-d), is said to be more common

in the case of idioms compared with literal expressions. In this type of
adjectival modification, the adjective modifies the entire expression rather
than just the NP object. In (4.16a), conservative does not modify hackles
198 Chapter Four
but, instead, it denotes a figurative reading of the idiom, which means here
“annoying the conservative part of the society.” In (4.16b), the adjective
total does not describe the shit, but, in turn, it builds the overall idiomatic
interpretation of “being scared totally.” Similarly in (4.16c-d), a sudden
liking and a horrifying shadow refer to the context in which the
expressions should be interpreted; viz. “Suddenly, she took a liking,” and
“Horrifyingly, the TB-ridden slums cast a shadow.” In brief, external
modification exists outside of the domain of idioms. Thus, examples like
(4.16a-d), allow an interpretation where the NP-internal modifier does not
work as restricting the reference of the nominal constituent, but, instead,
as an operator taking the nominal within its scope (cf. Nunberg et al. 1994:
500).
In (4.17a-d), there are some instances of the internal adjectival
modification of the object NP, attested for psychological idioms.
(4.17) Internal adjectival modification of the object NP in psychological

idioms:
a. play (a game of) cat and mouse “to annoy”

Enemy warplanes have been playing a deadly game of cat and
mouse, trying to bring American fighter planes into range of their
missiles. (COCA)
b. bring a hornet’s nest round Y’s ears “to annoy”
However, the chief offenders for the time were flogged and kept in
bounds; but the victorious party had brought a nice hornet's nest
about their ears.
(Google)
c. put / send / throw Y into a funk “to depress”
Having to change her menu threw the whole day off schedule and
put her into a blue funk.
(Google)
d. whet Y’s appetite “to appeal”
According to Communist, sportsmen were controlled by
capitalists to whet imperialistic appetites and divert attention
away from social and political problems.
(COCA)
Internal modification shows a lot of similarities with regular adjectival

modification, with the exception that these adjectives may be interpreted
either literally or figuratively, as illustrated in (4.17a-d) for psychological
idioms. To be more precise, “deadly” in playing a deadly game of cat and
mouse in (4.17a) may refer to a dull and exhausting way of causing

annoyance, when taken figuratively, or if literally retrieved, it may
describe extremely dangerous and death-bringing enemies, as the context
of the whole sentence suggests. In (4.17b), “nice” in the idiom had
brought a nice hornet’s nest about their ears should not be treated literally
as pleasant and enjoyable, since annoying someone is not nice at all.
Instead, the adjective is used ironically to emphasise the state of being
annoyed even more. In (4.17c), “blue” in put her into a blue funk evokes a
figurative interpretation of being “affected by fear or anxiety,” which is
traced back to the 19th century.16 In (4.17d), the adjective “imperialistic” in
whet imperialistic appetites must undergo metaphorization in order to be
interpreted properly. It alludes to the metaphorical reading of the
irresistible and indomitable power of imperialism. In all these cases, the
adjectives contribute to the figurative interpretation of the idiomatic
expressions.
Another type of adjectival modification of the object NP is called
conjunctive modification (cf. Ernst 1981; and Stone 2016). In contrast to
the two types of adjectival variation just discussed, the conjunctive
modification concerns the reading of an expression which is deduced both
on the literal and figurative basis, while the adjective itself is taken to
modify the literal meaning of the object NP, as illustrated in (4.18a-c)
below.
(4.18) Conjunctive adjectival modification of the object NP in

psychological idioms:
a. get a kick / a charge / a bang out of X “to enjoy”

This book is just the kind you like and you’ll get a real kick out of
it.
(Google)
b. cast a gloom / a shadow over Y “to depress”
The Palestinian Authority and its violent takeover of Gaza in June
have cast
a heavy shadow over politics in Jordan.
(COCA)
c. scare the shit / the wits out of Y “to horrify”
Granny used to scare the holy shit out of us kids with her
campfire tales.
(Google)
16
As given by American Heritage Idioms Dictionary (2002). Retrieved from
http://www.dictionary. com/browse/blue-funk--in-a
200 Chapter Four
In (4.18a), real modifies kick, which is understood as a true strike or thrill;

while simultaneously, the idiom is retrieved figuratively and the overall
meaning of the phrase get a real kick means “to enjoy,” with an emphasis
put on the fact that the enjoyment is real. In the same vein, in
(4.18b),heavy describes shadow to underline the extensive range of it. On
the other hand, if the expression cast a heavy shadow over politics in
Jordan is treated figuratively, the adjective heavy adds far more stress to
the idiomatic meaning of the phrase, i.e. “to depress.” This makes the
subject, viz. the Palestinian Authority and its violent takeover of Gaza in
June, even more responsible for the disheartening and hopeless situation in
Jordan. Interestingly, holy, in (4.18c), if treated literally, appears in
conjunction with the contradictory noun shit. This rhetorical figure, by
which apparently opposing terms are conjoined, so as to give emphasis to
the statement or expression, is called an oxymoron. When the idiom scare
the holy shit out of Y is interpreted as a whole, it should imply the reading
“to extremely horrify Y.”
In a nutshell, it has been demonstrated here that all the aforementioned
types of modification, in which the idiomatic object NP may occur, are
valid for psychological idiomatically combining expressions, as
exemplified by the instances taken either from the COCA Corpus or
retrieved via the Google Search. The kinds of alternation, the idiomatic
object NP may undergo, comprise anaphoric reference, quantification of
the object, changing the number of the object, and modification of the
object by means of an intensifier or an adjective. All of these variations
retain the idiomatic interpretation of an idiom, while conjunctive
modification refers both to the literal and figurative meaning of the idiom
at the same time.
4.3.2 Alternations of the syntactic configuration of elements

in psychological idiomatically combining expressions
Further variations refer to changes in the syntactic configuration of

idiomatic elements. Among these, the modifications relevant to psychological
idioms include: passivization (4.19), subject-to-subject raising (4.20),
control (4.21), tough-movement (4.22), relativization (4.23), proxy clause
formation (4.24), clefting (4.25), topicalization (4.26), object incorporation
(4.27), VP-ellipsis (4.28), figurative modification (4.29), and extendibility
(4.30).
To start with, passivization is one of the most universally discussed
variations, analysed by, e.g. Katz and Postal (1964), Fraser (1970),
Chomsky (1970), Katz (1973), Fiengo (1974), Newmeyer (1974), Nunberg
(1978), Nunberg et al. (1994), Folli and Harley (2007), and Stone (2013),
among many others. For decades, many theoretical proposals have been
offered to account for the distinction between idioms which can be
passivized, while still retaining their idiomatic meaning in the passive (e.g.
The beans were spilled “The secret was divulged”), and idioms which are
non-passivizable since their idiomatic interpretation is then lost (e.g. #The
bucket was kicked receives only a literal interpretation).
One of the most widely-recognised accounts for (lack of) passivization
of certain idioms is the one discussed in Nunberg et al. (1994). They argue
that compositionality of idioms is the key factor for an idiom to be
passivized. Due to semantic non-compositionality of kick the bucket “to
die,” which has its idiomatic meaning distributed over the entire phrase,
the idiom cannot passivize and still retain its idiomatic meaning. Whereas
a semantically compositional idiom spill the beans “to divulge a secret” is
passivizable, because the meaning of the idiom is shared among the
constituent elements of the idiom; as spill denotes “to divulge” and the
beans refer to “a secret.”17 In addition, Folli and Harley (2007), and Stone
(2008, 2013) offer a syntax-based account for distinguishing passivizable
and non-passivizable idioms.18
There is a fair number of psychological idiomatically combining
expressions which retain their idiomatic interpretation in the passive, as in
(4.19a-d) below, while some of them, when passivized, sound
grammatically unacceptable, as in (4.19e).
(4.19) Passivization of psychological idioms:

a. drive Y bananas “to annoy”
Adults also can be driven bananas by loud music, loud parties,
and loud cars.
(Google)
b. raise Y’s hackles “to annoy”
After 1976, Western European hackles were raised by the
tendency of President Jimmy Carter to make the application of
detente contingent.
(Google)
c. bear / owe a grudge against X “to hate”
Evil will and a grudge are born against me.
(Google)
17
Cf. McGinnis’s (2002) counterarguments against Nunberg et al.’s (1994)
account of passivization. For McGinnis (2002), aspect plays a significant role
since it is compositional even in non-passivizable idioms.
18
Cf. Stone (2013: 4-5) for a more detailed analysis.
202 Chapter Four
d. take/find / gain pleasure in / from X “to enjoy”

There is much pleasure to be gained from useless knowledge.
(Google)
e. catch Y’s eye “to fascinate”
*His eyes were caught by her look.
(native speakers’ judgement)
Subject-to-subject raising is another common modification, discussed

by Perlmutter (1970), Postal (1974), Abeillé (1995), Lasnik and Saito
(1999), and Osborne et al. (2012), among others. Abeillé (1995) argues
that both an idiomatic subject and an idiomatic passivized object can
undergo raising from a subordinate clause, to become the subject of a
higher sentence, usually the main clause. This type of variation is
illustrated for psychological idioms in (4.20a-c) below. In (4.20a) a night
fright is a passivized object of the idiom give Y a fright, and here the
object a fright belongs to the core of the idiomatic phrase. In (4.20b), an
object the good old monarch’s heart of the idiom gladden Y’s hearts “to
amuse,” realises an open position in the possessive structure. The object
originates in the complement position of gladdened and is first passivized,
and hence lands in the subject position of a non-finite clause, from which
it raises to the matrix clause subject position. In (4.20c), an object
politicians realises an open position of the idiom hold X in contempt. In
this case, there is passivization in the lower clause of an object politicians,
raised into the lower subject position, and then raising from the lower
clause subject position to the matrix clause.19
(4.20) Subject-to-subject raising with psychological idioms:
a. give Y (quite) a (bit of) turn / a fright “to scare”

A night fright seems to have been given to Brownie.
(Google)
19
Cf. Petersen’s (2016) comment on raising out of some idioms, which, if
modified this way, lose their idiomatic meaning, as in (i).
(i) O kombos fenete oti exi ftasi sto xteni (Greek)
the knot seem-3SG that have-3SG reached to-the comb
“The knot seems to have reached the comb.”
Idiomatic reading: #”Things seem to have come to an end.”
(Petersen 2016: 248)
Petersen (2016: 247-248) claims that (parts of) an idiom may not carry discourse
effects, and their raising, as an instance of A-movement, renders an idiomatic
interpretation infelicitous.
b. gladden Y’s hearts Y “to amuse”

The good old monarch’s heart appears to have been gladdened.
c. hold X in contempt “to hate”
Politicians seem to be generally held in contempt by ordinary
people.
(Google)
On the other hand, when idioms are broken up by means of some

control verbs, e.g. want, or fail, the sentences can only get a literal
interpretation, as illustrated in (4.21a-b) for psychological idioms.
(4.21) Control with psychological idioms:

a. play the fool for Y “to amuse”
# The fool wants to be played for Elisa.
(literal interpretation ĺ “A silly person has a desire to be
used, to please / gain Elisa’s favour”)
b. carry a torch for X “to love”
# A torch fails to be carried for Monica.
(literal interpretation ĺ “A flashlight isn’t delivered to
Monica”)
Van der Linden (1991: 27) claims that the control-construction is hard to
test for idioms.20
20
Interestingly, control has been thoroughly studied in the Government and
Binding framework in the 1980s, while in the days of Transformational Grammar,
it has been referred to in terms of Equi-NP deletion. Since control is said to be only
applicable to meaningful expressions; thus, an expression (the controller) is related
to an abstract pronominal (the controllee), and both the controller and the
controllee have to be meaningful expressions (cf. Schenk 1995: 260-261). Idiom
chunks cannot become controllers, as illustrated in the following examples:
(i) spill the beans ĺ “to reveal a secret”
a. Pete instructs John to spill the beans.
b. John tries to spill the beans.
c. *John instructs the beans to be spilled.
d. *The beans try to be spilled.
(van der Linden 1991: 27)
(ii) paint the town ĺ “to enjoy”
a. Mary instructs James to paint the town.
b. James tries to paint the town.
204 Chapter Four
Another type of variation, viz. tough-movement, refers to sentences in

which the syntactic subject of the main verb works logically as the object
of an embedded non-finite verb. In contradistinction to a freely applicable
subject-to-subject raising, tough-movement cannot occur freely in the case
of idiomatic expressions. Some idioms retain their idiomatic interpretation
in the tough-movement construction (cf. A can of worms is often easier to
open than one expects ĺ open a can of worms “to examine or attempt to
solve some problem, only to inadvertently complicate it and create even
more trouble,” in Ruwet 1991: 186). However, others cannot be used in
tough-movement structures since then the expressions are interpreted
literally (e.g. #The ice was easy to break ĺ “The frozen surface was easy
to be broken”).
In our analysis of psychological idioms, the idiomatic phrases which
can be affected by tough-movement, are presented in (4.22a-b) below,
while those which lose their figurative reading, if modified this way, are
provided in (4.22c-d).
(4.22) Tough-movement construction with psychological idioms:

a. cast a spell on Y “to fascinate”
Love Spells are easy to cast and can bring new love or bring back
an old love or lost love fast and simple.
(Google)
b. whet Y’s appetite “to appeal”
Appetites of consumers are more and more difficult to whet when
it comes to new products,’ said the advertising and marketing
workers at the conference.
c. freeze Y’s (the) blood “to horrify”
# The blood isn’t tough to freeze, and based on ion concentration,
the freezing is probably -2°C. But the protein (albumin) may
further lower the freezing point.
c. *Mary instructs the town to be painted.

d. *The town tries to be painted.
In (i-a, b), John, as a free argument of spill the beans, can be part of a control
structure. In (i-c, d), the idiom chunks sound bad when they occur in a control
structure. Likewise, the example of the psychological idiom paint the town ĺ “to
enjoy” in the control structure, in (ii), James is a free argument of the idiom paint
the town in (ii-a, b), which makes these examples acceptable. In turn, in (ii-c, d),
the idiomatic NP object the town occurs in the control structure, which results in
the ungrammaticality of these sentences.
d. fan the fire / flame (of something) “to anger”

# The fire is easy to fan, especially when the wind is heavy.
As can be seen in (4.22a-b), some idioms, including psychological idioms,

can retain their idiomatic interpretation when they undergo tough-
movement. Indeed, in (4.22a), love spells are easy to cast can be
understood as “it is easy to fascinate someone,” and in (4.22b), Appetites
of consumers are more and more difficult to whet has the idiomatic
meaning of “appealing” not referring to someone’s culinary preferences.
However, the idiomatic interpretation in (4.22c) for the idiom freeze Y’s
(the) blood “to horrify” is lost, and the literal interpretation is triggered,
referring to the blood and the process of freezing. Likewise, in (4.22d), the
idiom fan the fire / flame (of something) does not mean “to anger
someone,” but the literal action of kindling a flame is possible. It seems
right to conclude then that context in which a specific idiom appears
makes a big difference for an idiomatic or literal meaning (cf. Cacciari
1993: 27; Marschark et al. 1983; Moore 1982; Popiel and McRae 1988;
and Needham 1992; among others). In fact, these meanings (“subsenses”),
contextually dependent, may be combined to form either a literal or
idiomatic meaning (cf. Everaert 2010: 83).
A more detailed discussion of tough-movement taking place in idioms
can be found in Rosenbaum (1967), Berman (1973), Lasnik and Fiengo
(1974), Ruwet (1991), and Hicks (2009), among others.
Likewise, relativization, viz. the process of forming a relative clause,

can apply only to certain idioms. Schachter (1973) and McCawley (1981)
identify two possibilities for relativization, relevant for idioms, i.e. when
the idiomatic verb is in the relative clause (4.23a-d), or in the main clause
(4.23e). Besides, Siemund (2013) mentions three strategies to form
relative clauses, i.e. with the use of (i) relative pronouns, e.g. which, who,
as in (4.23e); (ii) the subordinator (or relativizer) that, as in (4.23d); and
(iii) by means of a null relative marker [ø], which seems to be the most
common one, as can be seen in (4.23a-c), for psychological idioms.
(4.23) Relativization of psychological idioms:
a. have / get goose bumps “to fear”

Yes, [I mean] goose bumps [ø] one can get when you hear
childbirth.
(Google)
206 Chapter Four

It’ll help extinguish that torch [ø] he carries for you when he sees
once and for all you
(Google)
She can finally let go of the grudge [ø] she has borne against
them all the time.
(Google)
d. bear / owe a grudge against X “to hate”
I tell first of the grudge that I bore against him.
(Google)
e. lose Y’s heart( to X) “to love”
My heart, which is bleeding, I’ve lost to him.
f. have ants in Y’s pants “to worry”
# These are the ants which I have in my pants.
g. get cold feet “to fear”
# Mary got cold feet which she had taken earlier out of her high-
heeled shoes.
In (4.23a) the idiomatic object goose bumps is modified by means of the

relative clause, introduced by a null relative marker [ø]. Likewise, in
(4.23b-d), the object idiom chunks, i.e. that torch and the grudge are
modified by relative clauses, which are introduced either by a null relative
marker [ø], or by means of the relativizer that. The sentence in (4.23e) is
grammatical, and the idiomatic object my heart is relativized and appears
in the subject position, while the idiomatic verb lost occurs in the main
clause. The instances in (4.23f-g) confirm the fact that some idioms lose
their idiomatic meaning in the process of forming a relative clause. Thus,
the sentence in (4.23f), These are the ants which I have in my pants,
should be interpreted literally as “having some small insects in my shorts”
but not as “to worry,” which the psychological idiom have ants in Y’s
pants denotes. Similarly, the sentence in (4.23g), Mary got cold feet which
she had taken earlier out of her high-heeled shoes, refers to literally
understood “cold feet which one may get when they are taken out of one’s
shoes.” Obviously, the phrase get cold feet loses here its idiomatic
interpretation “to fear,” and the literal meaning is enforced by the content
of the which-clause. In brief, relativization works only for certain idioms.
Another type of variation, relatively underrepresented in the

literature, refers to proxy clauses, which represent a subclass of relative
clauses. Such a clause works as a proxy mostly for some noun within the
clause itself. In (4.24a), an example of proxy clause, given by Higgins
(1974; 1981), is presented, where the clause what I would regard as
headway is serving as a proxy for the noun headway. The fact that this
modification is only valid for a small set of idioms, has been proven by the
fact that no relevant data has been found for psychological idioms in the
COCA Corpus and by means of The Google Search. It can only be
assumed that some version of a proxy clause is theoretically possible for
psychological idioms, as in (4.24b). However, according to native
speakers’ judgements, the sentence in (4.24b) is grammatical but sounds
artificial and weird, which is indicated with a question mark “?,” since it is
really dubious whether a speaker can use it in a real language.
(4.24) Proxy clause formation:
a. make headway “to develop, make progress”

John certainly isn’t making what I would regard as headway.
(Higgins 1974: 3; 1981, unpublished manuscript)
b. have a yellow streak / belly down X’s back “to fear”

? Tex has what I would regard as a yellow streak down his back.
The next type of variation which is little discussed in the literature, is

called clefting. The purpose of cleft sentences, mostly realised as It-clefts,
Wh-clefts, inverted-clefts, or Pseudo-clefts, is to put a particular (new)
constituent in focus. In clefts, a single message is divided (cleft) into two
clauses, which allows us to focus on the new information (cf. Collins
1991; and Lambrecht 2001). Fraser (1970), Carlson (1977), Gramley and
Pátzold (1992), Reeve (2010), and Anastasiou (2010), among others, find
clefting generally disallowed for English idioms, e.g. *It was the bucket
that Mark kicked (kick the bucket “to die”), or *It were the beans that
Agnes spilled (spill the beans “to divulge a secret”). French idioms, in
turn, can be modified by means of cleft structures easily, as discussed
thoroughly by Ruwet (1991) and Abeillé (1995). This variation is
exemplified in (4.25a) for the French idiom promettremonts et merveilles
“to promise the moon,” and in (4.25b-h) for English psychological idioms.
208 Chapter Four
(4.25) Cleft-structures with psychological idioms:
a. promettremonts et merveilles “to promise the moon’ [French]

C’est des monts et (des) merveilles que nous a promis le
président.
It is–mountains–and–marvels–that–us–promised–President
It is mountains and marvels that the President has promised us.
(Ruwet 1991: 192)
b. cast a gloom / a shadow over Y “to depress”
? It’s a shadow that is cast over the President.
But cf.
b’. It’s a shadow that hangs over the President’s capability of
conducting foreign affairs.
(COCA)
c. dampen/damp Y’s spirits “to depress”
?? It’s her spirits that were dampened by that difficult case.
But cf.
c’. It was her spirit that ravished him.
(COCA)
d. give Y the blues “to depress”
# All you ever give me is the blues.
e. show contempt for X “to hate”
What he often shows is contempt for the local authorities.
f. have / get goose bumps “to fear”
What I got were goose bumps, when I saw him so pale.
g. get cold feet “to fear”
# What she got in this terrifying situation was cold feet.
h. cast a gloom / a shadow over Y “to depress”
# What has been cast over the President is the shadow.
The French idiom in (4.25a) is completely grammatical in the cleft

structure; while the English psychological idioms in (4.25b-c), seem to be
syntactically acceptable, but they sound pretty strange, when judged by
native speakers. Interestingly, when the idiomatic verbs in (4.25b’-c’) are
changed into different ones, viz. cast ĺ hang over, and dampen ĺ ravish,
the sentences get improved, but the psychological idiomatic meaning is
lost, even though some figurativeness of the sentences is still retained.
Moreover, the instance illustrated in (4.25d) is an example of Inverted-
clefts, which is completely licit as regards grammar, but its meaning is
ambiguous between the idiomatic one, viz. give Y the blues “to depress
someone,” and the literal interpretation “to provide somebody with this
specific kind of music, i.e. blues.” Likewise, the idiomatic meaning is lost
for most of the Pseudo-cleft structures, illustrated in (4.25e-h). Only the
cleft examples with the idioms show contempt for X “to hate” (4.25e), and
have / get goose bumps “to fear” (4.25f), seem to be acceptable and retain
their idiomatic interpretation, but native speakers assess these structures as
rarely used in a real life language.
To sum up, the cleft structures, provided in (4.25b-h), seem to capture
the fact that the more transparent and compositional an idiom is, the easier
it forms the cleft structure. Furthermore, Gramley and Pátzold (1992)
propose that some of the reasons why certain idioms do or do not allow
transformations seem to be idiosyncratic; while for others the semantic
reasons can be given. Therefore, most idioms usually resist the isolation of
one formative for emphasis, as in the case of cleft structures (*It was her
throat that he jumped down, jump down someone’s throat “to respond to
what someone has said in a sudden and angrily critical way”), because in
this operation word forms are treated as semantic constituents, which they
are not. Throat in he jumped down her throat has no isolable meaning in
the idiom, and that is why it cannot be modified (cf. Gramley and Pátzold
1992: 57).21
Furthermore, a certain set of English idioms, including psychological

idioms, under some restrictions, can take part in topicalization, to
emphasize a certain part of an idiomatic phrase. Nunberg et al. (1994), in
21
Reeve (2010) provides some remarkable account for possible or disallowed cleft
structures. He argues that VO idioms, such as keep track and make headway, must
“be base-generated as a constituent.” This contributes to the fact that the idiomatic
object is dependent on the verb. However, in some cases, it is possible for the
object to undergo movement and still retain its idiomatic meaning, e.g. What kind
of track was she keeping t of her expenses?, or The careful track that she’s keeping
t of her expenses pleases me (cf. Carlson 1977, cited in Reeve 2010: 65). Whereas,
other cleft variations are found unacceptable, e.g. *What she is keeping of her
expenses is CAREFUL TRACK (den Dikken et al. 2000, cited in Reeve 2010: 65),
or * CAREFUL TRACK is what she is keeping of her expenses (cf. Reeve 2010:
65-66).
210 Chapter Four
their discussion of topicalization, argue that emphasizing parts of idioms

in this way would be pointless unless these parts have distinguishable
meanings in their idiomatic uses (ibid.: 501). An important account for
topicalization is also given by Schenk (1995), who confirms that this type
of syntactic variation is applicable only to meaningful expressions. To be
precise, idiom parts cannot be the focus of topicalization, viz. idiom
components cannot be discontinued, as illustrated in (4.26a-a’) and (4.26b-
b’), because they constitute one complete idiomatic whole, which, if
broken, results in losing the idiomatic interpretation and leads to
ungrammaticality of the sentence.
(4.26) Topicalization with psychological idioms:
a. spill the beans “to divulge a secret”

*The beans, John spilt.
(Schenk 1995: 259)
a’. kick the bucket “to die”
*The bucket, John kicked.
(Schenk 1995: 259)
a’’. read the riot act “to give someone a severe scolding”
To the class, Pete read the riot act.
(Schenk 1995: 259)
b. upset the apple cart against Y “to annoy”
*The apple cart, they upset.
b’. get a kick / a charge / a bang out of X “to enjoy”
* A kick, I got out of this incredible event.
Against them, she has borne a grudge all the time.
d. lose Y’s heart to X “to love”
To this lady, he’s lost his heart recently.
e. make a difference to Y “to matter”
Only to Jason, she makes a difference.
On the other hand, topicalization is allowed for the instance illustrated in

(4.26a’’) above, since the class is a free slot of the idiom, and thus it can
be topicalized (cf. Schenk 1995: 259). Similarly, as can be seen in (4.26c-
e) on the example of psychological idioms, against them, to this lady, and
only to Jason are the free slots of the idioms, which, when shifted into the
topic position at the beginning of the sentence, do not result in breaking
the core meaningful structure of these idioms. Therefore, such
constructions are acceptable, although judged by native speakers as quite
rare. A more detailed analysis of topicalization is provided in Wasow et al.
(1984), Gazdar et al. (1985), and Osborne et al. (2012), among others.
Another type of variation, hardly ever analysed in the literature, is

called object incorporation, viz. synthetic compound formation. It seems
to be applicable to English idioms, when an idiomatic verb forms a
compound with its direct object, still retaining its original syntactic
function. Nonetheless, at least two changes must occur in this variation,
i.e. (i) the bare NP object must take a pre-verbal position without its
determiner; and (ii) it is an adjectival rather than verbal compound, with
participial -ing morphology for the verb (cf. Baker 1988; Rosen 1984,
1989; and Mattissen 2006). The application of object incorporation to
idioms is illustrated on the example of psychological idioms in (4.27)
below.
(4.27) Object incorporation in psychological idioms:
a. give Y a turn “to horrify”

The purpose is to make explicit the simultaneously occurring
markers and cues of the turn-giving intention of the current
speaker based on information coming from different modalities.
(Google)
b. curdle / chill Y’s blood “to scare”

Then she let out a blood chilling scream and ran to the other side
of the table.
(COCA)
The example in (4.27a) shows object incorporation with give Y a turn “to
horrify,” which appears as the adjectival expression turn-giving. Whereas,
in (4.27b) the adjectival compound blood chilling derives from the idiom
chill Y’s blood “to scare,” and it comprises the idiomatic NP object blood
placed pre-verbally, before the idiomatic verb chill. The determiner of the
NP, represented in (4.27b) by a genitive possessor, which reveals the
identity of the Experiencer Y, must be omitted in this type of modification.
212 Chapter Four
Furthermore, certain parts of idioms can sometimes be affected by VP-

ellipsis. The antecedents of the missing elements in elliptical constructions
must correspond to semantic units, i.e. to pieces of interpretation (cf.
Nunberg et al. 1994: 501). Due to the fact that the antecedents are parts of
idioms, they must bear some kind of interpretation of their own (ibid.:
501). The application of VP-ellipsis to idioms has been discussed by
Wasow et al. (1984), Gazdar et al. (1985), Johnson (2001), Goldberg
(2005), and Bos and Spenader (2011), among others. This elliptical
construction is illustrated in (4.28a-b) below for psychological idioms, in
which certain parts of the idioms have a ball, or carry a torch, are omitted.
(4.28) VP-ellipsis with psychological idioms:
a. have a ball “to enjoy”

At school he worked a lot harder when he had the ball than when
he didn’t.
(Google)
“People carry torches when they’re young.” “They do,” he
grinned.
(Google)
What is more, Egan (2008: 19-20) introduces two further types of

idiomatic variations, viz. figurative modification (4.29), and extendibility
(4.30). The figurative modification bears some resemblance to internal
modification as in (4.17) above, except for the fact that the figurative
modification is more flexible, and the modifier is not required to be in a
prenominal position. Besides, as noted by Vietri (2014: 91), the figurative
modification may involve “wordplay.” In addition, Egan (2008) argues, in
his PRETENSE hypothesis, that manipulations may concern the literal
meaning of some parts of an idiom, rather than the figurative interpretation.
Indeed, Egan’s (2008) prediction that figurative modification is relevant to
both idiomatic combinations and phrasal idioms seems to be confirmed.
Our study of psychological idioms provides support for Egan’s (2008)
hypothesis as well, as illustrated in (4.29a-b).
(4.29) Figurative modification for psychological idioms:
a. knock Y’s socks off “to surprise”

Yeah, you said you could give me a kiss that would knock my
socks off. I'm still not wearing any. (Google)
b. ruffle Y’s feathers “to annoy”

The Bishop of Edinburgh doesn’t just ruffle feathers–he tears
them out in handfuls.
(Google)
While any in (4.29a) literally refers to socks in the preceding sentence in

the idiom knock Y’s socks off, it is interpreted figuratively when the
idiomatic meaning is activated. If knock Y’s socks off means roughly “to
surprise,” then “I'm still not wearing any” indicates here that so far there
has been no surprise, or no surprise is expected to happen. In the same
way, in (4.29b), the modifying clause “he tears them out in handfuls”
implies that them is a substitute for feathers, which not only are ruffled, as
the idiom suggests, which means “to annoy;” but literally retrieved the
feathers are also said to be torn out in handfuls. This “wordplay,” related
to the literal reading of particular parts of the idioms, may occur owing to
the activated figurative interpretation of the idioms.
Extendibility is another process discussed by Egan (2008), and it refers

to the idiomatic interpretation which is extended to other parts of the
discourse, be they idiomatic or not, while sticking to the same semantic
concept.22 In (4.30) below, extendibility is illustrated on some examples of
psychological idioms.23
(4.30) Extendibility relevant to psychological idioms:
a. “to annoy:”
drive Y bananas (batty / nuts/ bonkers / crazy) ĺ ruffle Y’s
feathers
Things that drive me crazy might not bother you, while something
that has you climbing the walls might hardly ruffle my feathers.
(Google)
22
While discussing extendibility, Egan (2008) provides the following example:
(i) Speaker A: I hear Mr. Jones kicked the bucket.
Speaker B: Yeah. He almost connected yesterday; today he really put the
boot on it.
(kick the bucket; Egan 2008: 393).
In Egan’s (2008) example above, extendibility means that once Speaker A, in (i).
has invoked the idiom kick the bucket to mean that Mr. Jones has died, Speaker B
relates to the same semantic space of death, using terms connected, and boot.
23
Egan (2008) argues that the line between extendibility (and perhaps even
figurative modification) and wordplay is tenuous, and no dividing line of this kind
should be drawn.
214 Chapter Four
b. “to worry and fear”

have butterflies in Y’s stomach ĺ have a yellow streak / belly
down Y’s back
The psychosomatic component to our suffering has been

recognized as “a pain in the neck,” or, “I’ve got butterflies in my
stomach,” or (…) We say, “He’s got a yellow streak up his
spine,”
(Google)
c. “to worry and fear”
have butterflies in Y’s stomach ĺ shaking
GIBSON: Well, we win the 200-meter butterfly, both the men's

and women's. We talked to Tom Malchow yesterday, and I asked
him the question I've always wanted to ask swimmers in the
butterfly races. And I ask you the same question, did you have
butterflies in the butterfly?
Ms-HYMAN: Well, I definitely had butterflies all day long

yesterday. There were some points where I was actually shaking.
(COCA)
In the example in (4.30a), an Experiencer describes the stimuli which

make him / her annoyed, referring to them as “things that drive me crazy,”
while others “ruffle my feathers.” In (4.30b), an Experiencer has invoked
two idioms to denote the state of being worried and afraid, i.e. have
butterflies in Y’s stomach and have a yellow streak down Y’s back. The
speaker meant to define “the psychosomatic components to our suffering.”
Whereas in the conversation in (4.30c), Ms-HYMAN refers to the state of
being worried and anxious, using the idiom have butterflies in Y’s
stomach, and the non-idiomatic predicate shake. In all these instances, the
Experiencers have extended their idiomatic interpretations, using related
terms, even related idiomatic expressions, to invoke the same semantic
space of either “annoyance” (4.30a), or “worry” (4.30b-c).
4.3.3 Co-occurrence dependencies in psychological ICEs

One of the distinguishing features of idiomatically combining expressions,
which is underlined by Nunberg et al. (1994: 504-505), is the dependency
between the idiomatic verbs and their objects. The authors claim that this
dependency is semantic in nature, which stems from the fact that an idiom
consists of a (literal) meaning which receives a particular conventional and

homomorphic association. This conventional mapping from literal to
idiomatic reading is homomorphic taking into account certain properties of
the interpretations of the idiom’s constituents. Therefore, some
idiomatically combining expressions are expected to have families of
idioms. More specifically, either the same verb can occur with different
NPs / PPs to form distinct, but semantically related, idioms; or the same
idiomatic NP / PP takes different verbs; or the same or semantically
related idiomatic meaning is retained, but all the idiom components are
syntactically or/and lexically different. All of these possibilities are
illustrated on the examples of psychological ICEs, in (4.31a-j), (4.32a-i),
and (4.33a-o), respectively.24
(4.31) Homomorphic families of psychological idioms, with the same

verb, e.g.:
a. love ĺ set store by X; set Y’s heart on X

b. enjoy ĺ get a buzz / get a kick / get a charge / get a bang / out of X
c. hate ĺ bear ill will toward X; bear a grudge against X; bear aversion /
malice / hostility / repugnance toward (to) X
d. annoy ĺ rattle Y’s cage /chain; get Y’s dander /hacklers/ Irish up
e. amuse ĺ tickle Y to death / to pieces / pink
f. scare ĺ frighten the life / the hell out of Y; give Y a turn / a fright
g. depress ĺ break Y’s heart / spirit; cast a gloom / a shadow over Y;
knock Y down a peg / notch (or two); give Y a bad / hard
time/ the blues / the run around; bring Y into disrepute/
bring Y low; knock the stuffing out of Y/ knock Y for six /
knock Y sideways
h. anger ĺ fan the fire / flame (of something); get Y’s back / dander up;
drive Y out of mind /drive Y to distraction /drive Y round
the bend / twist
24
It is worth recalling here, for the sake of convenience, what has been introduced
in Chapter Three of the book. Specifically, all psychological idioms are not
provided with detailed meanings they have, but classified into the general
psychological domains, directly related to psych-verbs. In fact, the interpretation of
every single idiom differs, e.g. both the idiom carry a torch for X “to secretly love
someone who does not love you,” and set store by X “to regard as valuable or
worthwhile, worthy to be loved” are classified into the same semantic domain of
LOVE, and the psych-verb related to these idioms is “to love.”
216 Chapter Four
i. surprise ĺ knock Y’s socks off / knock Y down/over with a feather;

catch Y off balance / up short / napping / catch Y on the hop
/ catch Y off (Y’s) guards / catch Y unawares; leave Y open-
mouthed / leave Y at a loss for words; strike Y dumb / strike
Y with awe / strike Y with wonder; throw Y off balance/
throw Y a curve (ball)
j. horrify ĺ scare the shit / the wits out of Y / scare the bejesus out of Y
(4.32) Homomorphic families of psychological idioms, with the same

NP / PP, e.g.:
a. enjoy ĺ derive/gain/get pleasure from X

b. hate ĺ bear / owe a grudge against X; bear / feel aversion / malice /
hostility / repugnance toward (to) X
c. fear ĺ have a yellow streak / belly down Y’s back
d. annoy ĺ throw/send Y into a tizzy / tizz
e. scare ĺ curdle / chill Y’s blood; throw /send Y into a panic
f. depress ĺ damp/ dash / wither Y’s hopes; put / send / throw Y into a
funk
g. anger ĺ put / set Y’s back (up)
h. horrify ĺ frighten/scare the pants off Y; frighten / scare Y to death;
frighten/ scare Y out of their wits
i. appeal ĺ set / put Y on Y’s ear
(4.33) Homomorphic families of psychological idioms, with

syntactically or / and lexically different components but with the
same or semantically related meaning:
a. love ĺ 13 idioms, 25 e.g.

carry a torch for X; fall head over heels in love with X;
b. enjoy ĺ 11 idioms, e.g.
have a ball; blow / let off (some) steam; kick (up) Y’s
heels
c. hate ĺ 10 idioms, e.g.
show contempt for X; hold X in abomination / contempt;
d. fear ĺ 9 idioms, e.g.
turn tail (and run); take dim view of X; have a yellow
streak down Y’s back
25
These numbers are based on the corpus study, presented in Chapter Three.
e. worry ĺ 6 idioms, e.g.

have the blues; eat Y’s heart out;lose sleep over X
f. annoy ĺ 26 idioms, e.g.
raise the hump; get Y’s goat; raise Y’s hackles
g. fascinate ĺ 11 idioms, e.g.
catch Y’s eye; tickle Y’s fancy; stir the /Y’s blood
h. amuse ĺ 4 idioms, e.g.
gladden Y’s hearts; make the grade for Y; play the fool
for Y
i. scare ĺ 11 idioms, e.g.
curdle / chill Y’s blood; put the fear of God into Y;
j. depress ĺ 20 idioms, e.g.
put a damper on Y; bring tears to Y’s eyes; take the wind
out of Y’s sails
k. anger ĺ 10 idioms, e.g.
fan the fire / flame (of something); kindle Y’s wrath; get
Y’s back up
l. surprise ĺ 15 idioms, e.g.
knock Y’s socks off; knock Y down / over with a
feather; blow Y out of the water
m. horrify ĺ 8 idioms, e.g.
freeze Y’s (the) blood; put the screws on Y; scare the
shit / the wits out of Y
n. appeal ĺ 4 idioms, e.g.
float Y’s boat; tickle Y’s fancy; whet Y’s appetite
o. matter ĺ 3 idioms, e.g.
carry some weight with Y; cut (no) ice with Y; make a
difference to Y
As can be seen in (4.31a-j), (4.32a-i), and (4.33a-o), there are numerous

instances of psychological ICEs, which form homomorphic families of
idioms, preserving the co-occurrence dependency. It is necessarily
required for all these idiom components (idiomatic verb, NP, or PP) to
stay in the same semantically coherent domain. However, as admitted by
Nunberg et al. (1994: 505), this semantic dependency in idiomatically
combining expressions may miss some factors, such as the definiteness of
idiomatic NPs. As a result, this fact may provide some justifiable account
for the marginality of certain psychological idioms with definite articles,
e.g. raise the hump, get the blues, or make the grade for, in comparison
with those which comprise an open slot realised in a possessor, e.g. damp /
218 Chapter Four
dash / wither Y’s hopes; rattle Y’s cage / chain; or get Y’s dander /
hacklers / Irish up.
To sum up, in section 4.3, the issue of syntactic and lexical variability
of ICEs (on the example of psychological idioms) has been discussed.
Quantification, topicalization, ellipsis, and anaphora, among others, are the
numerous examples of variation forms idiomatically combining
expressions can occur in. As exemplified by psychological idioms, ICEs
tend to exhibit also a degree of lexical substitution, due to their more
compositional nature. IdPs, instead, are usually entirely resistant to any
such alterations, as has been discussed in section 4.2.1.2, and exemplified
by inflexibility of psychological IdPs in (4.10). Table 4-1 summarises the
differences between IdPs and ICEs. Compositionality
Conventionality
Flexibility
Variation
Syntactic
Opacity
Lexical
Open
Slots
Idiom
Non-
IdP High Opaque No No No
compositional
Trans- More
ICE Low Yes Yes Yes
parent compositional
Table 4-1. Properties IdPs and ICEs

(Corver et al. 2017: 10; cf. also Harwood et al. 2016)
Besides, taking into account this plentiful variation idiomatically

combining expressions can undergo, illustrated on the aforementioned
examples of psychological idioms, it seems to be hardly possible to sustain
a notion of idioms as purely frozen expressions. However, due to the
scarcity of systematic investigations concerning idiom flexibility, the
difficulty to distinguish idiomatic from literal interpretations, and semantic
from syntactic ill-formedness; it has been problematic to draw any explicit
conclusions concerning permissible idiomatic variability. Nonetheless,
some introspective grammaticality conclusions, made by experienced
linguists who recognise these nuances well, have helped to form the
primary source of data about the nature of idioms.
4.4 Constraints on the syntactic structure of psychological

idioms. Previous accounts
In the previous sections, some crucial issues concerning the characteristics
of idioms have been reviewed, together with the examples of flexibility
which psychological idiomatically combining expressions (ICEs) may
display. However, idioms have puzzled linguists for decades. The puzzle
refers both to the syntactic rules that idiomatic phrases seem to disobey,
and the way in which their overall idiomatic reading is to be deduced.
Various proposals in this field have been made by linguists, on the ground
of both theoretical and empirical observations, but unfortunately, little
agreement concerning the behaviour of idioms has been reached so far.
The aim of this part of the chapter is to deal with syntactic constraints
imposed on idiomaticity by the grammar. Section 4.4, specifically,
presents an outline of the most crucial approaches towards the behaviour
of idioms, reviewing Nunberg et al.’s (1994) semantic alternative to the
Hierarchy Constraint (section 4.4.1), and O’Grady’s (1998) Continuity
Constraint (section 4.4.2). These accounts have been found useful to
analyse the syntactic structure of psychological idioms before the onset of
the recent Idioms As Phases Hypothesis (Svenonius 2005; Kim 2014;
Harwood 2013, 2015, 2016). In the discussion concerning constraints on
idioms, reference is to be made to particular syntactic patterns of
psychological idioms, elicited in Chapter Three.
4.4.1 Nunberg et al.’s (1994) semantic alternative

to the Hierarchy Constraint
To begin with, it has been Nunberg, Sag and Wasow (1994) who, in their
seminal work, distinguish idiomatic phrases (IdPs), e.g. kick the bucket “to
die,” or raise the roof “to enjoy,” from idiomatically combining
expressions (ICEs), e.g. spill the beans “to reveal a secret,” or give Y a
fright “to scare.” This distinction, adopted for the sake of the book, and
analysed thoroughly in section 4.2.2, seems to be useful to understand the
behaviour of the so-called non-compositional and fixed idioms, on the one
hand, and compositional and flexible ones, on the other.
What is more, while analysing idioms, generative linguists have never
dealt satisfactorily with idiomatic compositionality, productivity and
syntactic variance, which has led them to treating idioms as “extra-
grammatical,” placed at the periphery of grammar. Nunberg et al. (1994),
instead, argue that certain components of an idiom can be assigned an
interpretation, and that “modification, quantification, topicalization,
220 Chapter Four
ellipsis, and anaphora provide powerful evidence that the pieces of many
idioms have identifiable meanings which interact semantically with other”
(Nunberg et al. 1994: 503). Contradicting this way a well-established
assumptions made in generative grammar, Nunberg et al. (1994: 503)
conclude that the meanings of idiom chunks are not their literal meanings,
but, instead, idiomatic meanings are largely derived from literal meanings
in the conventionalized, but not entirely arbitrary, manner. As a result,
conventionality should not be identified with non-compositionality, and
differences in syntactic flexibility among idioms can be explained by
means of the compatibility of semantics with the semantics and pragmatics
of various constructions (cf. Nunberg et al.’s (1994: 504-505) postulation
of the existence of homomorphic families of idioms, discussed in section
4.3.3).
Furthermore, Nunberg et al. (1994) comment on Marantz’s (1984) and
Kiparsky’s (1987) generalization concerning the rarity of idiomatic Agents
and Goals in idioms, providing some additional argumentation about why
idioms contain far fewer animate NPs than inanimate ones in general.
Nunberg et al. (1994) argue that, in normal discourse, verbs mostly take
animate objects, while hardly any animate objects are used in idiomatic
expressions (e.g. hit the ceiling “to get very angry,” hit rock bottom “to
reach the lowest point,” kiss the dust “to fall down due to being shot / hit,”
or kiss the cup “to drink”). Indeed, this assumption seems to be true when
considering the case of psychological idioms under scrutiny. Truly, out of
161 English idioms related to a psychological condition, which have been
elicited in Chapter Three, there is only one instance including a literally
animate NP, i.e. get Y’s Irish up “to annoy,” providing the nationality
adjective Irish, used in this idiom, is an NP. Moreover, Nunberg et al.
(1994) assume that the predisposition of metaphorical mappings to shift
from concrete to abstract in idioms does not offer a full explanation of the
extreme rarity of concrete / animate idiomatic meanings in idioms. Such
metaphorical reference to concrete things, actions, or situations does not
commonly occur.
Likewise, according to Nunberg et al. (1994), it is relatively rare for an
idiom to feature a fixed Goal or Possessor argument. This also follows
from their generalization concerning the scarcity of fixed Agent
arguments, since Goals and Possessors tend to be animate. Nunberg et
al.’s (1994) assumption is confirmed by psychological idioms. In the set of
161 psychological idioms, both idioms with a possessor (get Y’s goat;
raise Y’s hackles “to annoy”) and a fixed Goal (tickle Y to death “to
amuse”) comprise fewer than one fourth of all the dataset.
In short, in their proposal accounting for asymmetries in the

grammatical or thematic roles of idiom chunks, Nunberg et al. (1994)
offer an alternative approach to both Marantz’s (1984) and Kiparsky’s
(1987) hierarchies of thematic roles. Nunberg et al. (1994) recognise these
asymmetries as a consequence of broader and multifactorial tendencies in
figurative uses of language, and the way the world is conceived by human
beings (Nunberg et al. 1994: 531).26
4.4.2 O’Grady’s (1998) Continuity Constraint

In this section, the twelve patterns of psychological idiomatic constructions,
established in Chapter Three, are to be tested against O’Grady’s (1998)
Continuity Constraint. Later on, some space is devoted to the Hierarchy
Constraint, which is to be investigated in terms of possible vs. impossible
argument structures of psychological idiomatic verbs.
To start with, in his influential work, O’Grady (1998) puts forward the
claim that idioms are subject to the Continuity Constraint. This
grammatical principle, stated in (4.34), delimits the general architecture of
idioms, in terms of a continuous chain of head-to-head relationships.
(4.34) The Continuity Constraint:

An idiom’s component parts must form a chain.27
(O’Grady 1998: 284)
O’Grady (1998) argues that idiomatic constructions display a relationship

between heads and their dependents. In addition, lexical selection between
lexical heads is always more prominent than the dependency relationship
between a lexical head and a functional head. To be precise, if an idiom is
26
It is significant to add that 3 years later after Nunberg et al.’s (1994) seminal
works, Jackendoff (1997) in his framework of Representational Modularity (RM),
postulates that both syntactic and lexical conceptual structures (LCS, Jackendoff
1997: 49) are involved in the production of meaning. Jackendoff (1997, 2002)
argues that idioms have phonological structure, syntactic structure, and conceptual
structure, but due to the fact that not all of the syntactic constituents of an idiom
correspond to conceptual constituents, the idiom, in fact, may give rise to idiomatic
interpretation. Most idiomatic expressions do not have compositional meaning;
thus, they comprise complex lexical items whose meaning is not syntactically
analysed but rather determined by the syntactic-conceptual structure interface
component. In short, in his representational modularity approach, Jackendoff
(1997) postulates the lexical licensing of units larger than Xº.
27
A chain is identified by O’Grady (1998: 284) in the sense that “iff x [in the
string x…y…z…] licenses y and z, or if x licenses y and y licenses z.”
222 Chapter Four
to be illustrated via a tree structure, the continuity chain will go down the
tree, from heads down to their dependents.
Let us take all the syntactically different patterns of the psychological
idioms, elicited in Chapter Three, to examine how they behave with
respect to the Continuity Constraint.The twelve types, which 161
psychological idioms under scrutiny represent, show the patterns of chains
specified in (4.35) below.
(4.35) The patterns of chains for the psychological idioms (cf. O’Grady
1998):
Thus, as can be seen in (4.35), eight out of the twelve patterns of

psychological idioms show a chain of heads, remaining in a relation with
their dependents viz. verbs select nouns (objects), prepositions, particles,
and a given type of a small clause as a whole; nouns select their specifiers,
i.e. genitive possessors and articles, and possibly adjectives or quantifiers
if the idiom is modified; prepositions within a PP select NPs which play
224 Chapter Four
the role of a complement of the P. In four types of the idioms analysed, the
chain is broken due to the fact that Y (the Experiencer) or X (the Theme)
occurs in between the verb and other dependents and interrupt the chain
relation between them. For the four types of idioms, listed in (4.36) below,
the Continuity Constraint does not hold. Thus, they cannot be captured in
O’Grady’s (1998) model.
(4.36) Illicit idioms with respect to the Continuity Constraint:
a. Type (C)*: V + NP + preposition + NP (complement of a

P)
hold X in abomination
b. Type (D’)*: V + NP +preposition +possessor +N

(complement of a P)
c. Type (E)*: V + NP + NP (double object construction)

give Y the blues
d. Type (E’)*: V + a complement small clause

drive Y bananas
Table 3-4 in Chapter Three shows that the number of OE (class II)
psychological idioms, occurring in the aforementioned illicit idioms is
extremely big for Type (C), quite big as for Types (E) and (E’), while
Type (D’) is marginal. Therefore, based on the chain-like patterns of
psychological idioms, analysed in (4.35), it can be concluded that the
chain cannot replace all syntactic structures ever possible, but, instead, it
works as a bare minimum constraint on the very structure of an idiom.
Nonetheless, the Continuity Constraint seems to be applicable to most
cases.
All in all, in O’Grady’s (1998) Continuity Constraint, the lexical
choices sound explicitly fixed within idiomatic constructions, i.e. the
selection restrictions are based on specific Head-to-Head relationships.
The Continuity Constraint accurately specifies the organization of existing
idioms, including non-constituent idioms, simultaneously predicting that
certain types of patterns are impossible. Additionally, it sheds new light on
the relevance of argument structure for idiom formation, pointing to some
asymmetries in the composition of idioms, making space, this way, for
thematic hierarchy effects. Finally, non-idiomatic open slots are placed by
O’Grady (1998) at the very end of chains, just to preserve the chain from
breaking. Instead, there should be more possibilities for idiomatic patterns
with open positions. Therefore, the existence of psychological idioms with
open slots realised as NP objects, makes the Continuity Constraint flawed
and, thus, worth revising.
What is more, developing Kiparsky’s (1987) and Baker’s (1989)
Thematic Hierarchies, reproduced in (4.37a) and in (4.37b) respectively,
O’Grady (1998) formulates the Hierarchy Constraint, as in (4.38).
(4.37) a. Kiparsky’s (1987) Thematic Hierarchy Constraint:

Agent > Theme > Goal/Location
(Kiparsky 1987: 35-36)
b. Baker’s (1989) Extended Thematic Hierarchy Constraint:
Agent > Instrument > |Experiencer > |Patient / Theme >
Goal/Location
(Baker 1989: 544)
(4.38) O’Grady’s (1998) Hierarchy Constraint:
Any arguments that are part of a verbal idiom must be lower on
the hierarchy than arguments that are not part of the idiom.
(O’Grady 1998: 293)
Following O’Grady’s (1998) Hierarchy Constraint, it can be predicted

which arguments may or may not be part of a verbal idiom. Thus, any
fixed element in a VP idiom is expected to be placed lower in the
Thematic Hierarchy than an argument realizing an open slot. The validity
of O’Grady’s (1998) Hierarchy Constraint is confirmed by eight types of
psychological idioms, viz. idioms with Theme NP-objects: Types (A),
(A’), (A’’), e.g. float Y’s boat; idioms with fixed Goal / Location PPs and
open Theme slots: Types (C), e.g. hold X in abomination; idioms with
fixed Goal / Location PPs and open Experiencer slots: Types (C), (C’),
(D’), e.g. drive Y up the wall; and idioms with fixed Theme NPs and open
Experiencer slots: Types (E), (E’), e.g. give Y the blues. In all these idioms
their fixed components are situated lower in the Thematic Hierarchy than
the non-idiomatic parts.
However, two Types of idioms, i.e. (B) and (D), e.g. carry a torch for
X, lose Y’s heart to X, contradict the Hierarchy Constraint, postulated by
O’Grady (1998), since the fixed idiomatic Theme argument (a torch , Y’s
heart) is placed higher than the Goal / Location PP with an open slot (for
X, to X). Moreover, the issue of two more types of idioms in which the
open position is realized within the possessor of the NP-complement of a
PP (i.e. Types (D’’) and (B’): have Y’s heart to X’s mouth, have a yellow
226 Chapter Four
streak down Y’s back), is unresolved. In fact, the types of idioms

problematic for the Hierarchy Constraint do not coincide with the types of
idioms which contradict O’Grady’s (1998) Continuity Constraint, as
illustrated in (4.36). Furthermore, O’Grady’s (1998) Hierarchy Constraint
does not specify what lexical categories it refers to. Indeed, within
idiomatic constructions, the distinction between what is an argument and
what is an adjunct is not maintained, since for O’Grady (1998) the most
crucial defining property of idioms is the requirement of chain formation
among lexical heads, without specifying further different levels of
projection.
Interestingly, Bruening (2010) formulates a restriction similar to the
Continuity Constraint, specifying the lexical categories. He reformulates
the Hierarchy Constraint, offered by O’Grady (1998), as follows:
(4.39) Constraint on Idiomatic Interpretation:
If X selects a lexical category Y, and X and Y are interpreted

idiomatically, all of the selected arguments of Y must be
interpreted as part of the idiom that includes X and Y.
Lexical categories are V, N, A, Adv.
(Bruening 2010: 532 (25-26))
Bruening’s (2010) constraint accounts successfully for V-NP idioms with

possessors realising an open position (e.g. eat Y’s heart out; float Y’s
boat), and PPs realising an open slot (e.g. carry a torch for X), assuming
that the selected arguments of the NP-idiomatic objects must be
interpreted as part of the idiom as well. Nonetheless, Bruening’s (2010)
constraint cannot account for V-NP-PP idioms (e.g. drive Y up the wall),
in which an open slot is represented by an NP-object, and the fixed PP
belongs to an idiom, even though it is not a lexical category listed in
Bruening’s (2010) constraint.
In section 4.4 of the chapter, some analyses related to syntactic
constraints on idioms have been briefly outlined. A special focus has been
laid on those approaches which have sounded instrumental to our analysis
of psychological idioms. Thus, some space has been devoted to Nunberg
et al.’s (1994) semantic analysis, and to the so-called Continuity
Constraint formulated by O’Grady (1998). It has been shown that
O’Grady’s (1998) analysis can account for most, but not all, of the
syntactic patterns that psychological idioms under scrutiny occur in.
4.5 The structure of psychological idioms

in the Phase Theory
Among many recent approaches referring to the relations between
idiomaticity and syntax, the notion of a phase-bound structure, termed as
the Idioms as Phases Hypothesis, has been found worth analysing. The
Phase Theory, which is still under debate in the literature, has been
developed in various forms, represented by Chomsky (1998, 2000, 2001,
2008), Radford (2004), Matushansky (2005), Boškoviü and Lasnik (2006),
Gallego (2010), and Citko (2014), among others.
The aim of this part of Chapter Four is to make an attempt at looking at
the structure of idioms, with reference made to psychological idioms under
scrutiny, in the light of the Phase Theory. This part is structured as
follows: section 4.5.1 introduces the basic terms relevant for the Phase
Theory. The subsequent sections are devoted to the structure of different
types of idioms in the Phase Theory, i.e. fixed IdP-idioms confined to vP-
phases (section 4.5.2), ICE (V-O) idioms with DP-phases (section 4.5.3),
Double Object ICE-idioms (section 4.5.4), ICE (V-O-PP) idioms with PP-
phases (section 4.5.5), idioms with particles (section 4.5.6), and ICE-
idioms with small clauses comprising Predication-phases (section 4.5.7).
The purpose of the discussion undertaken in this section is to test whether
the predictions of the Phase Theory work well for the idiomatic
expressions analysed here.
4.5.1 Defining phases
In the Minimalist Program, Chomsky (1995, 1998, 2000, 2001, 2008)

presents a noteworthy contribution to the generative tradition in
linguistics, making an attempt to situate linguistic theory within the
broader range of cognitive sciences. The minimalist framework is built on
the theory of Principles and Parameters and, particularly, on principles of
economy of derivation and representation.
Within Chomsky’s theoretical framework, Universal Grammar works
as a unique computational system, derivations are driven by morphological
properties, and linguistic expressions are generated by optimally efficient
derivations which are required to satisfy the conditions that hold on
interface levels, the only levels of linguistic representation. The interface
levels, viz. Phonetic Form (PF) and Logical Form (LF) interfaces, provide
instructions to two types of performance systems, namely the articulatory-
perceptual system and the conceptual-intentional one. The Logical Form
(LF) of a linguistic expression is defined as the mental representation of it,
228 Chapter Four
derived from surface structure. According to Chomsky (1993), LF

captures those aspects of semantic representation that are strictly
determined by grammar, but abstracted from other cognitive systems (cf.
Collinge 1990). LF functions as the interface between grammar and
conceptual-intentional properties of language. The Phonetic Form (PF), in
turn, is the interface between grammar and the audio-perceptual properties
of utterances.
In the Minimalist Program, the syntactic computation occurs
repeatedly, and it is realised in strict chunks, or phases, while the structure
is built up (Chomsky 1998, 2000, 2001, 2008). The term “a phase” was
first introduced in Chomsky’s (1998, 2000) “Minimalist Inquiries.”
Building the concept of a phase on many previous principles related to
locality domains (e.g. cycles, barriers, islands, among others), Chomsky
(2000) characterises phases in terms of lexical subarrays (LAi), as
reproduced in (4.40a) below. On the level of a sentence, an array of lexical
items is chosen from the lexicon; and a subarray of lexical items is
selected from the array to construct a phase. When this is done, another
subarray is chosen up to the point when the whole array is used up.
(4.40) a. “A phase of a derivation is a syntactic object derived . . . by

choice of LAi.”
b. A phase is “the closest syntactic counterpart to a proposition:

either a verb phrase in which all theta roles are assigned or a
full clause including tense and force.”
(Chomsky 2000: 106)
In his definition of phases, reproduced in (4.40b), Chomsky (2000)

assumes what constituents may count as phases, and what their defining
properties are. Thus, CPs are phases, as are transitive and unergative vPs;
whereas TPs, as well as unaccusative and passive vPs are not.28 Providing
28
As summarised by Landau (2003: 2-3), T, v [little v], and C are core functional
categories. While v expresses transitivity, selects V, has ĳ-features (object
agreement), selects external argument, has optional EPP feature (second Merge)
for object shift; T expresses tense/event structure, has ĳ-features (subject
agreement), obligatory EPP feature. If T is selected by C, it has a full set of ĳ-
features, whether expressed (finite) or not (control). If selected by V
(raising/ECM), it has only a subset of ĳ- features. C, in turn, expresses force/mood,
has ĳ-features, and an optional EPP feature (for wh-phrases).
a number of diagnostics which distinguish phases from TPs and VPs,

Chomsky (2000, 2001) states that phases form natural semantic units,
since they comprise a clause including tense and force or a verb phrase
with all theta roles assigned. Thus, beside their propositional nature, also
syntactic (or semantic) completeness is a defining property of phases29 (cf.
Matushansky 2005). Moreover, following Citko (2014), it can be
presumed that the lack of external arguments (or / and the presence of an
internal argument) is the key factor which distinguishes verb phrases
which are not phases (unaccusative and passive vPs) from these which are
(transitive and unergative vPs).30
Interestingly, Chomsky (2000) does not characterize phases in terms of
“convergence,” justifying this by stating that phases are meant to reduce
computational complexity. On the other hand, phases can be defined in
terms of the interfaces, recognised as objects which determine points of
Transfer to the two interfaces, viz. PF and LF interfaces.
However, the defining characteristic of phases, which is most
commonly cited in the literature, refers to some specific property of phase
heads, i.e. that they are the loci of uninterpretable features (cf. Chomsky
2000, 2001; Gallego 2010; Legate 2012; and Citko 2014; among others).
Gallego (2010) states that “uninterpretable features signal phase boundaries”
Besides, “v* is the functional head associated with full argument structure,
transitive and experiencer constructions, and is one of several choices for v, which
may furthermore be the element determining that the selected root is verbal”
(Chomsky 2008: 142).
What is more, a sentence is usually said to comprise two phases, namely the
complementizer C (CP) and the light verb v (vP). Chomsky (2000) suggests that if
the light verb v represents a transitive verb, then it is ĳ [phi]-complete. Likewise,
finite and control C are recognised as ĳ-complete. Then, transitive vPs (ĳ-
complete v), and finite CPs and control CPs (with PRO subjects), represent strong
phases.
29
Chomsky (2000, 2001) argues that phases display some degree of phonological
and syntactic or LF independence. While semantic or LF independence is marked
by the fact that the edges of phases are reconstruction sites for quantifier and
operator movement; phonological independence is recognised by the phonological
isolation of phases.
30
However, Citko (2014) admits that it is difficult to explain the reason why the
argument structure of unaccusative or passive verbs is less complete than the
argument structure of transitive verbs. Likewise, unergative verbs may be treated
as incomplete in terms of their argument structure, when they lack an internal
argument (John ran vs. John ran a race), or many transitive verbs may also be
found incomplete in terms of their argument structure, because they form double
object constructions (John baked a cake vs. John baked Mary a cake) (cf. Citko
2014: 29-30; and Epstein 2007).
230 Chapter Four
(Gallego 2010: 151); whereas Legate (2012) specifies that “a C/v

possessing an unvalued ĳ-feature in the numeration is a phase head”
(Legate 2012: 239).
According to Chomsky (2000, 2001), ĳ-features are attribute-value
pairs that enter the derivation hosting a particular value (e.g., [Number:
+Pl]) or lacking a value (e.g., [Number: __ ]). The property of having or
lacking a value coexists with the property of (un)interpretability, and this
relation is biconditional (Chomsky 2001: 5), as formulated in (4.41a).
(4.41) a. A feature F is uninterpretable iff F is unvalued.

(Chomsky 2001: 5)
b. Types of features (Chomsky 2000, 2001):

(i) valued interpretable features iF [val],
(ii) unvalued uninterpretable features uF [ ]31
As can be seen in (4.41b), there are only two types of features visible to
the computational system (valued interpretable features iF[val], and
unvalued uninterpretable features uF[ ]). Val stands for any feature value
complex, while empty brackets signify the lack of value, and filled
brackets signify valued features. A feature is interpretable if it can be
interpreted at the interface level LF, making direct contributions to
meaning, while a feature is uninterpretable if it cannot be interpreted at the
interface level LF. In short, feature interpretability is determined in the
lexicon, being maintained throughout the derivation, including the LF-
branch of the derivation (Chomsky 2001: 6). The examples of interpretable
features include the ĳ-features of nouns (DPs) and the tense feature of T;
whereas uninterpretable features comprise, e.g. the ĳ-features of T and v,
the tense feature of V and the Case feature of DPs (cf. Willim 2012: 764).
Importantly, uninterpretable features have to be valued in the course of the
derivation. Valuing unvalued features occurs via the mechanism called
Agree. Unvalued features need to be deleted by the time of Transfer to the
interfaces, because uninterpretable features cannot be interpreted by the
interfaces.
Furthermore, within the Minimalist Program, it is uninterpretable ĳ-
features which are supposed to trigger computational operations (agreement
and movement). Moreover, only phase heads are said to have
31
Cf. Pesetsky and Torrego (2007), and Boškoviü (2011), who suggest
distinguishing other combinations, i.e. uF [val] an uninterpretable and valued
feature, and iF [ ] an interpretable and unvalued feature.
uninterpretable features, while non-phase heads may inherit them in the

derivation process by means of the process of Feature Inheritance (FI) (cf.
Chomsky 2007, 2008; Richards 2008, and Citko 2014; among others).
Consequently, phase heads (e.g. C or v), as the hosts of uninterpretable
features, trigger syntactic operations.
Once an expression no longer contains any uninterpretable features, it
inevitably spells out (Chomsky 2000, 2001, 2008; Svenonius 2001: 275,
2004). Since phase heads trigger Spell-Out, they can act as Probes, i.e.
value uninterpretable features. In other words, they trigger the transfer of
the Spell-out domain to the two interfaces (PF and LF), and are subject to
strong cyclicity, as formulated in (4.42) and (4.43).
(4.42) The head of a phase is inert after the phase is completed,

triggering no further operations.
(Chomsky 2000: 107)
(4.43) The Phase Impenetrability Condition:
In a Phase Į with head H, the domain of H is not accessible to

operations outside Į, only H and its edge are accessible to such
operations.
(Chomsky 1998, 2000: 108)
According to the so-called Phase Impenetrability Condition (PIC),

reproduced in (4.43), as soon as HP is complete, i.e. when H no longer
projects, the complement of H is spelled out. The fact that HP is
inaccessible means that any features which are unvalued (or unchecked)
are to remain so. The general configuration to which PIC refers is
reproduced in (4.44a), while in (4.44b), a more concrete clausal
configuration is presented. HP / vP is the phase and H / v is a phase head,
whereas YP/VP is the Spell-Out domain, and H/v with its specifier form
the phase edge.
232 Chapter Four
(4.44)
(Citko 2014: 32)
Movement out of the phase proceeds through the phase edge, and a
constituent is only permitted to move out of a phase (the Spell-Out
domain), providing the constituent has first moved to the phase “edge” (cf.
Citko 2014: 32). Hence, Chomsky (2000: 109) claims that phase heads
have the requisite features to trigger movement.
Furthermore, even though the generally cited form of the PIC is the
one reproduced in (4.43), various versions of the PIC have been proposed
in the literature. Indeed, all of them refer to the same general
configuration, presented in (4.45a) below, in which Z and H are phase
heads, while X is a non-phase head placed between them. A clausal
structure of this configuration is shown in (4.45b), in which C and v are
phase heads, while T is not.
(4.45) a. [ZP Z . . . [XP X [HP Į [ H YP ] ] ] ]

(cf. Chomsky 2001: 13)
b. [CP C . . .[TP T [vP DP [v VP ] ] ] ]
(cf. Citko 2014: 32)
As can be seen in (4.45a), where HP is a strong phase, and ZP is the next

higher strong phase, it is postulated that elements of HP are available to
operations within the smallest strong ZP phase, but not beyond. In other
words, elements inside the strong phase ZP can target the edge Į of HP.
Once the derivation reaches ZP, the complement of Z is sent to Spell-Out,
and it remains impenetrable (inaccessible) for further computation. The
domain of H (here, YP) is not accessible to operations at ZP, but only H
and its edge.
Besides, the successive-cyclic movement invariably targets the edge of
cyclic domains. The fact which stems from the cyclic determination of
strong phases and the Phase Impenetrability Condition is that accessibility
of the edge of a strong phase is only up to the next strong phase. Making a
supposition that the Spell-Out operation sends a syntactic object to both
PF and LF, the appropriate generalization can be formulated as in (4.46)
below, following Chomsky (2001).
(4.46) Interpretation / evaluation for Phase 1 is at the next relevant

(that is strong) Phase 2.
(Chomsky 2001: 13)
In (4.46), Phase 1 is strong and Phase 2 is the next higher strong phase,
where a strong phase is a CP or a vP.
In a nutshell, as noted by Radford (2000), Chomsky’s (2000) Phase
Impenetrability Condition32 is a natural consequence of the locality
constraint, referring to the phase boundary of syntactic and phonological
operations. Hence, the domain of a subordinate (strong) phase (i.e. vP, CP
or possibly DP) is not penetrable to the head of an immediately
superordinate phase. Syntactic movement operations, according to the
assumptions made within the Phase Theory, involve two sub-operations of
copying and deletion. Thus, the constituent which is moved is first copied
into the position to which it moves, and then the original one may be
32
As argued by Citko (2014: 33), Müller (2004), Richards (2011), among others,
there are at least two versions of the Phase Impenetrability Condition, i.e. (i) one
from Chomsky’s (2000) “Minimalist Inquiries,” generally referred to as “a Strong
PIC / PIC1;” and (ii) the other from Chomsky’s (2001) “Derivation by Phase”
(referred to as “a weak PIC / PIC 2”). They are reproduced below:
(i) PIC1: The domain of H is not accessible to operations outside HP; only H and
its edge are accessible to such operations.
(ii) PIC2:The domain of H is not accessible to operations at ZP; only H and its
edge are accessible to such operations.
(Chomsky 2001: 13-14)
234 Chapter Four
deleted, viz. given a null phonetic Spell-out. It is also typical of the Phase
Theory that “Spell-Out is cyclic, at the phase level” (Chomsky 2001: 9).33
4.5.2 The vP-phase boundary and psychological

idiomatic phrases (IdPs)
Recently, verbal idioms have become subject to locality constraints (cf.
Chomsky 1980, 1981, 2008). Marantz (1984, 1997) argues that the special
idiomatic interpretation evoked in idioms can be associated with a
functional head like v, which introduces the external argument of a verb.
In the Phase Theory, Chomsky (2000) introduces two important phase
boundaries within a clause, i.e. CPs and vPs. Svenonius (2005), and Stone
(2009), among others, predict that for some idioms idiomatic interpretation
can be dependent on constituents contained within a single phase.
According to Harwood et al. (2013, 2015, 2016), only idiomatic phrases
(IdPs) are restricted to a single phase (vP), while idiomatically combining
expressions (ICEs) can freely straddle multiple phase boundaries. In the
analysis below, we are to check if this claim can be confirmed by
psychological idioms.
To begin with, Harwood et al. (2016: 6) state that the material from the
TP-domain, thus beyond the vP-domain (e.g. tense, modality, and aspect),
in most cases does not contribute to the accessibility of the idiomatic
interpretation. That is, tense, modality or aspect may surface within the
idiom, and still the idiomatic reading may remain preserved (cf. Kitagawa
1986; Ifill 2002; and Svenonius 2005; among others), as exemplified in
(4.47a-c).
(4.47) Possible modifications of psychological idiomatic phrases (IdPs):

By wanting his own way the baby is upsetting the apple-cart.
(Google)
(tense of the V / progressive aspect)
33
In the Minimalist Program the thematic roles are assigned as a result of the
operation Merge. These roles are mapped according to the hierarchy of thematic
roles (cf. Kiparsky 1987; Baker 1989; and O’Grady 1998; among others) and
according to the UTAH (Baker 1988). Psych verbs do not satisfy the latter
requirement.

Now that you have a black-tie outfit, we can paint the town.
(Google)
(modality)
c. upset the applecart “to depress”
Old Jameson has upset my apple-cart. (Google)
(perfect aspect)
The idiomatic interpretation in fixed verbal idioms, instead, is expected

to be exclusively restricted to a single vP-phase (cf. Svenonius 2005;
Harwood 2013, 2015, 2016, 2017; and Corver et al. 2017).
To understand properly the line of our discussion in favour of a single
vP-phase, it is worth recalling the issue of idiom non-compositionality.
Each non-compositional idiomatic phrase (IdPs) is said to form one single
semantic unit, and as such, it is mapped to a single meaning, as illustrated
in (4.48).
(4.48)
What is more, following Fellbaum (1993), Nunberg et al. (1994),

McGinnis (2002), Everaert (2010), and Stone (2013), among others, who
treat idioms as constructed by means of regular structure building
mechanisms of syntax, it is assumed that the non-compositional/figurative
interpretation arises only at the syntax-semantics interface (SEM).34
Providing each phase is “shipped off” in an independent way to the
interfaces for pronunciation and interpretation (Chomsky 2000, 2001,
2008), an idiomatic phrase is expected to reach SEM as a wholesale unit
(one single phase) in one go if the idiomatic interpretation is to arise. But,
if an idiom has straddled the vP-phase boundary, then only some part of
the idiom will be sent off to SEM. Accordingly, this way of reasoning
seems to justify why fixed verbal idioms are confined to the material
included within one vP-phase.
In addition, idiomatic phrases (IdPs), recognised as idiomatic units
restricted to a single vP-phase, are comprised of the material (the verb and
34
Cf. Jackendoff (1997), Marantz (2001), Svenonius (2005), and Harwood et al.
(2016), among others, who provide various proposals concerning how idiomatic
meanings are primed at the syntax-semantics interface.
236 Chapter Four
its arguments) which forms a semantically coherent whole, entirely

contained within the vP-domain (Chomsky1980, 1981; Marantz 1984,
1997; Svenonius 2005; Citko 2014; and Harwood et al. 2016, 2017;
Corver et al. 2017; among others). Providing the first phase is an isolated
domain of meaning (Chomsky 2000, 2001), it is logical to treat this type of
verbal idioms and the clause-internal phase as aligned. In fact, fixed verbal
idioms can be smaller than or equal to the vP-phase, but they cannot
straddle the vP-phase boundary. Hence, the vP-phase boundary works as a
limit on the maximal size of the syntactic constituents which form a verbal
idiomatic phrase (cf. Harwood et al. 2016: 6). In brief, following the
assumption that IdP-idioms are indeed confined to a single vP-phase, the
size of verbal idioms can be determined.
Moreover, the most canonical idiomatic phrases (IdPs) comprise a
verbal predicate and a DP-object, as in those illustrated in (4.49a-i), which
correspond to psych-verbs. All these idiomatic phrases represent Type (A)
idioms and have a fixed syntactic pattern V + NP (cf. section 3.4 in
Chapter Three), in which both the idiomatic verb and the idiomatic NP
object are fixed. A noun object is mostly preceded by the definite article
“the,” as in (4.49a-f); however, the indefinite article “a/an” may also occur
in front of an NP, as in (4.49g); a null article [ø], as in (4.49h); or a fixed
adjective, e.g. white, as in (4.49f). The idiomatic noun object can also be
realised in a plural form, e.g. pins, as in (4.49i), or as a sequence of NPs
combined with a conjunction “and,” e.g. pins and needles, as in (4.49i).
(4.49) Examples of the fixed psychological idiomatic phrases (IdPs) with

no open position Type (A): V + NP

c. raise the roof “to enjoy”
d. fan the fire “to anger”
e. have the blues ”to worry”
f. show the white feather “to fear”
g. have a ball “to enjoy”
h. turn tail (and run) “to fear”
i. have / get pins and needles “to fear”
The psychological idiomatic phrases provided in (4.49) face the problem:

how can verbal idioms straddle the DP-phase boundary when they are
supposed to be restricted to a single vP-phase?
To address this question, the analysis provided by Harwood et al.

(2016) might be of some help. Articulating their phase-bound perspective,
Harwood et al. (2016) provide a clearly formulated explanation which is
to solve the puzzle with the apparent DP-phase straddling. They argue that
the solution to this problem lies in the very nature of the definite
determiner in IdPs, and, accordingly, in the structure of the object DP in
idiomatic phrases. More specifically, providing the constituents of an
idiom (e.g. kick the bucket “to die,” or raise the roof “to enjoy”) are
literally interpreted, the particular component elements contribute, then, to
the semantics and the overall meaning of the phrase. In terms of syntactic
functions, the verbal predicate (e.g. kick or raise) would occur with its
argument (e.g. the bucket, or the roof), carrying a theta role (Theme).
Whereas, as far as idiomatic phrases are concerned, the nominal
expressions (e.g. the bucket, or the roof) are not the true arguments
(Themes) of the verbs, and the verbs (e.g. kick or raise) do not act as
typical transitive predicates. Instead, the idioms under scrutiny should be
treated as a whole, entirely restricted to the vP-domain.
In addition, if the roof were a true argument (Theme) of the verb raise,
the definite determiner would be referential, with special reference made
to the unique object mentioned in the discourse (cf. Harwood et al. 2016:
12-13). As a result of the referential determiner (i.e. after binary Merge
has applied to D (the) and N (roof), yielding the Syntactic Object [Į D N]),
either D or N can project and label the whole Syntactic Object (cf.
Chomsky 2013, 2015). When D projects, the DP-phase is formed, as
illustrated in (4.50a).
(4.50)
Whereas in an IdP, it is the N (here roof) which projects, becoming the

head of the projection, which results in an NP, as in (4.50b). On the other
hand, Harwood et al. (2016) add that, if the idiomatic direct object in an
IdP were to act as the DP phase, then, when an idiom lacks a DP, there
will be no phase and no DP projection.
238 Chapter Four
Moreover, the nouns in IdPs (e.g. the bucket, or the roof) do not denote
generic and unique referents, but, instead, are “non-denoting nouns” for
Fellbaum (1993), or “no identifiable idiomatic referents” for Grégoire
(2009). Hence, the nominal expression (NP) in an idiomatic phrase (IdP) is
non-referential, and it cannot be spelled out self-sufficiently, but always as
a part of the vP-phase, to evoke the idiomatic non-literal interpretation.
Following the account provided by Harwood et al. (2016), a syntactic
representation for Type (A): V+ NP psychological idiomatic phrases,
listed in (4.49) can be as in (4.51) below.
(4.51)
As can be seen in the tree structure in (4.51), the nominal expression the
roof cannot become an argument, i.e. receive the theta-role, as in the case
of all IdPs, because only a referential category, i.e. only DPs, can occur in
argument positions (Stowell 1991; Longobardi 1994; and Harwood et al.
2016; among others). Likewise, the determiner of the direct object in IdPs
lacks a direct referent in the discourse, and thus it does not project. Since
there is no DP-phase in idiomatic phrases (IdPs), there is no DP-phase
boundary to be straddled by IdPs. Instead, the direct object of an idiomatic
phrase directly forms part of the vP-phase, rather than projecting its own
phasal domain (cf. Harwood et al. 2016: 2, 5).
Nonetheless, among the instances of psychological idiomatic phrases
(IdPs) listed in (4.49) above, there are some idioms which lack the definite
determiner, in contradistinction to the canonical example kick the bucket
“to die,” or raise the roof “to enjoy.” Instead, an NP is preceded by an
indefinite article (e.g. have a ball “to enjoy”), a null article [ø] (e.g. turn
tail “to fear”), a fixed adjective (e.g. show the white feather “to fear”),
takes a plural form (e.g. get pins and needles “to fear”), or is coordinated
with another NP (e.g. have / get pins and needles “to fear”). This fact does
not bring any consequences for the syntactic structure of this type of
psychological idioms, since all the idiom constituents are, anyway,
expected to be confined to a single transitive vP-phase, which can be
represented by means of the structure, as in (4.52a-b) below.
(4.52) A tree representation for Type (A): V+ NP psychological idiomatic

phrases with some variation in the structure of the fixed NP:
What is more, the analysis of nominal phrases outlined in Chomsky

(2007) and elaborated in Harwood et al. (2016), provides some more
justification, in favour of treating all these different variants of Type (A)
psychological idioms (discussed above) as a single vP-phase. Chomsky
(2007) assumes that indefinite nominals lack then* head;35 thus, they lack
referentiality. Even though referentiality is present in lexical definite
nominals, it is absent in idiomatic objects in IdPs, which share non-
35
The existence of the so-called “little-n” (by analogy with “little-v”) is justifiable
in the structure of complex head-initial nominal phrases. Even though an NP does
not need a theta-assigner for its external argument since there is no external
argument; the shell structure of complex head-initial NPs is the same type of shell
structure as that of complex head-initial VPs (cf. Haider 2012: 55). Introducing n,
Chomsky (2007) draws a parallel between VPs and NPs; thus, n* is like v* and D
is like N.
240 Chapter Four
referential properties with indefinites (cf. the discussion above), and thus
lack n* in their structure the same way indefinites do. Therefore, Chomsky
(2007) makes a proposal that only definite nominal phrases (which are in
fact n*Ps in this view) constitute a phase, while indefinite nominals are not
phases. Similarly, Adger (2003), Radford (2004), Hinzen (2012), Corver
et al. (2017), and Harwood et al. (2017) state that indefinite DPs are not
phases, but only complete referential expressions are phasal, and the
instances of idioms in (4.49) and (4.52) can be treated as “complete
referential expressions,” with an assigned idiomatic interpretation only if
their constituent elements (V+NP) form a single unit. In other words, the
object of these psychological IdPs will not be spelled out independently,
but only as part of the vP-phase.36
All in all, with the overall idiomatic interpretation not derived from
their constituent parts, IdPs do form single semantic units, necessarily
interpreted as a whole. Therefore, it is highly justifiable to treat fixed IdPs
as restricted to a single vP-phase boundary.37 This fact sheds light on
further consequences stemming from it, viz. vP-phase boundary accounts
for an upper limit of the size of verbal IdPs. To be more precise, the
syntactic material found within the vP-phase specifies the extent of a
verbal idiom (Svenonius 2005; Harwood 2013, 2015, 2016, 2017; Kim
2014; Harwood and Temmerman 2015; and Corver et al. 2017; among
others). What is more, having based our line of reasoning on psychological
verbal IdPs (e.g. paint the town / raise the roof “to enjoy”), the nominal
(NP) objects in these IdPs have been proved to be non-referential, and
always spell-out not as a separate DP-phase but as an NP of the vP-phase,
to evoke the idiomatic interpretation. Hence, with no DP-phase present in
IdPs, no problem of straddling the DP-phase boundary arises.
36
Following Harwood et al. (2016), it can be concluded that the determiner in a
NP of IdP idioms (e.g. raise the roof) is non-referential; whereas, the determiner
(definite article) in ICE-type idioms (e.g. spill the beans / fan the fire) is
referential, forming a DP.
37
Cf. Harwood and Temmerman’s (2015) findings concerning several idioms
which depend on material external to the vP-phase, i.e. modality and aspect. They
prove that if an idiom is deprived of the relevant aspectual form or modal verb, its
figurative interpretation is altogether lost. However, Harwood and Temmerman’s
(2015) claim is contradicted by Kitagawa (1986), Ifill (2002), and Svenonius
(2005), among others.
4.5.3 DP-phases and psychological ICE-idioms with a possessor

in NP objects
Beside idiomatic phrases (IdPs), confined to a single vP-phase, Nunberg et
al. (1994), and Harwood et al. (2016), among others, distinguish another
type of idioms, i.e. idiomatically combining expressions (ICEs) (cf. section
4.2 of this chapter). On the basis of well-recognised characteristics of
ICEs, i.e. that they are compositional in nature, and that their idiomatic
components are decomposed into smaller idiomatic chunks, ICEs are
assumed not to be confined to a single phase, the way IdPs are. Indeed,
ICEs do not form a single semantic unit; thus, they are not interpreted as a
whole, at one go at the syntax-semantics interface (SEM), but instead,
their idiomatic interpretation can be built up gradually in the course of the
entire derivation. In fact, ICEs are expected to, but do not have to, span
multiple phase boundaries (cf. Harwood et al. 2016; and Corver et al.
2017), which is to be discussed in the sections to follow, 4.5.3-4.5.7.
Originally only two “core” categories, i.e. CPs, and vPs, used to be
assumed to be phases (cf. Chomsky 2000, 2001). However, in the recent
literature devoted to phases, DPs have been proposed to be phases, as well.
Legate’s (2002), and Citko’s (2014) diagnostics for phasehood38 support
the treatment of DPs as phases.39 Likewise for Chomsky (2005: 9, 16), and
Svenonius (2004, 2005), Boškoviü (2005), Corver et al. (2017) among
others, DPs are claimed to act as phases (cf. Radford (2004), and Harwood
et al. (2016), who argue that indefinite / non-referential DPs are not
phases, but NPs).
The core of a DP-phase is largely dependent on the internal syntax of
DP (cf. Citko 2014: 108). Cross-linguistically, the relative ordering of the
internal DP elements follows the hierarchy as in (4.53), developed by
Svenonius’ (2007) study of the order of articles (Art), demonstratives
38
Since phasehood is often characterized in terms of interface considerations, the
diagnostics are usually classified into two major groups: LF diagnostics and PF
diagnostics, with reference given to the two interfaces. However, some researchers,
e.g. Matushansky (2005), Citko (2014), among others, divide phasehood
diagnostics into more types, adding a separate class of syntactic (or
morphosyntactic) diagnostics to the two based on the interfaces.
39
According to Citko’s (2014) syntactic diagnostic evidence, DPs are phases since,
e.g. DP is a domain for feature valuation; D is the locus of uninterpretable features;
movement out of DP proceeds through the edge; and D determines Spell-Out (cf.
Citko 2014: 123).
242 Chapter Four
(Dem), plural markers (Pl), classifiers, numerals (Num), and adjectives

(Adj) (cf. Alexiadou, Haegeman and Stavrou 2007).40
(4.53) The hierarchy of DP elements:
Dem > Art >Num >unit >Pl / sort41 >Adj > n42 > N
(Svenonius 2007: 26)
In the recent literature, as argued by Adger (2003), Radford (2004),

Chomsky (2007), Hinzen (2012), and Corver et al. (2017), the phasehood
of DPs is associated with referentiality and definiteness. As discussed
earlier for IdPs (cf. section 4.5.2), both indefinite DPs and non-referential
expressions lack the functional head n*, which marks referentiality (cf.
footnote 30; Chomsky 2007; Harwood et al. 2017: 50). Therefore NPs in
IdPs do not form DP-phases, but indefinite DPs (cf. (4.54) for the IdPs,
e.g. kick the bucket “to die,” or paint the town “to enjoy”).
40
Cf. Citko (2014: 109), who proposes a DP structure with more common nominal
functional projections, suggesting to accommodate the DP-elements. Nonetheless,
according to Citko (2014), nP is placed below DP, which is in contradistinction to
Chomsky (2007), Corver et al. (2017), and Harwood et al. (2016, 2017), for whom
nP marks referentiality and is placed higher than DP (cf. (4.54)-(4.55)).
41
In Svenonius’ (2007) hierarchy, the categories “Pl/sort” and “unit” apply to such
cases in which plurals cooccur with classifiers, involving either “unit” classifiers or
noun classifiers, not “sort” classifiers. Classifiers individuate masses for
quantification and counting (cf. Svenonius 2007: 23).
42
The “n” here is a gender or noun class marker (cf. Svenonius 2007: 26).
(4.54) Indefinite/non-referential DP in IdP idioms
Complete referential expressions with the functional head n*, instead,

form a DP-phase, as illustrated on the example of ICEs (spill the beans “to
divulge a secret,” or fan the fire “to anger”), in (4.55).
(4.55) Referential DP and a DP-phase in ICE idioms
As can be seen in (4.55), idioms which are ICEs, and which follow the
pattern V + the + NP, as in Type (A) of psychological idioms, comprise a
244 Chapter Four
DP-phase provided the phrase is referential. If not, the expression is not

phasal.43
Having analysed the very nature of a DP-phase, and the difference
between referential and non-referential determiner in V+NP idioms;
further discussion concerning an open position realised by a possessor in
an idiomatic DP-object, can be embarked on. In their phase-bound
account, Corver et al. (2017) and Harwood et al. (2017: 54) contribute to
the analysis of possessors in idiomatic NPs.44 They argue that ICE
possessive idioms, which express the possessive relation in their DP-
object, all involve cases of alienable possession (e.g. style in the idiom
cramp x’s style “to make x look uncool,” or sorrows in drown x’s sorrows
“to drink to relieve depression”), in contradistinction to inalienable
possession, typical of IdPs (e.g. ear in the idiom bend x’s ear “to talk
excessively to x,” or balls in break x’s balls “to give x a hard time”).45
43
Canonical V-Object IdPs do not form a DP-phase, but give rise to the idiomatic
V+NP structure, which is confined to a single vP-phase, and spelled out at one go,
to evoke an idiomatic interpretation, as illustrated in (4.51), and repeated in (4.54).
44
It is worth mentioning here the structural constraint on idiomatic interpretations,
introduced by Koopman and Sportiche (1991: 224), who deal, among other issues,
with idiomatic constituents with open positions realised in possessors, such as the
ones found in our dataset of psychological idioms, e.g. get Y’s goat “to annoy,” try
Y’s patience “to annoy,” ruffle Y’s feathers “to annoy,” catch Y’s eye “to
fascinate,” win Y’s heart “to fascinate,” or turn Y’s head “to fascinate.” Koopman
and Sportiche (1991) argue that an open position never bears a direct
correspondence to the structural head of the idiomatic constituent, since no
idiomatic interpretation should be based on, for example, a pair of co-occurring
words viz. a modifier of the subject or a modifier of the VP. Therefore, they
propose the following structural constraint on idiomatic interpretations, formulated
in (i) below.
(i) The constituency-based constraint on idiomaticity:

If X is the minimal constituent containing all the idiomatic material, the head
of X is part of the idiom.
(Koopman and Sportiche 1991: 224)
The constituency-based constraint on idiomaticity specifies that the head of the
smallest constituent that contains all of the idiomatic material must also be
idiomatic. Regardless of the fact that Koopman and Sportiche (1991) fail to
contribute to set limits on the size of the constituent which can retrieve idiomatic
meaning, their constraint does underline the importance of local relations between
idiomatic constituents (cf. Harwood 2013, 2015).
45
Cf. Harwood et al. (2017) for further discussion concerning alienable possession
in ICEs and inalienable possession in IdPs.
Psychological idioms of Type (A’), e.g. raise Y’s hackles, kindle Y’s
wrath, and ruffle Y’s feathers, also include the instances of alienable
possession, such as hackles, wrath, or feathers.46 They have been checked
against syntactic diagnostics, applied by Harwood et al. (2017), i.e.
passivization, topicalization and modification, to check if these idioms are
ICEs or IdPs. As a result, the psychological idioms which have passed the
diagnostics are ICEs and are presented in (4.56a-c); whereas an example
of a psychological idiom (with inalienable possession eye), which has not
met the requirements of the syntactic diagnostics, is shown in (4.56d).
Nonetheless, the idiom in (4.56d) is not an IdP.47
(4.56) Type (A’): V + possessor + N
a. raise Y’s hackles “to annoy”
After 1976, Western European hackles were raised by the tendency

of President Jimmy Carter to make the application of detente
contingent. (Google) (passivization)
46
I would like to thank Prof BoĪena Cetnarowska for her suggestion that if hackles
and feathers in the psychological idioms raise Y’s hackles, and ruffle Y’s feathers
“to annoy” are personified, then these nouns may be treated as the parts of one’s
human body, which results in making them the instances of inalienable possession
not the alienable one.
47
Cf. a counterexample of what Corver et al. (2017) postulate about no ICEs with
inalienable possession, viz. a psychological idiom in (i), which includes inalienably
possessed heart, and still passes the diagnostics to be an ICE; and a psychological
idiom in (ii), which comprises alienable possession boats, and fails the diagnostics
to be an ICE:
(i) gladden Y’s hearts “to amuse”
His heart has been gladdened because of the presence of God’s Holy Spirit.
(Google) (passivization)
Her heart, the painting gladdened. (native speakers’ judgement)
(topicalization)
A lovelier June morning never gladdened innocent hearts. (Google)
(modification)
(ii) float Y’s boats “to appeal to”
# His boats were floated onto cradles.(native speakers’ judgement)
(passivization)
*Definitely, her boats, comedies float most.(native speakers’ judgement)
(topicalization)
# A calm sea can float some of her boat. (native speakers’ judgement)
(modification)
246 Chapter Four
The teacher’s hackles, this naughty boy’s behaviour raised.

(native speakers’ judgement) (topicalization)
Children’s health insurance bill in the current Congress has also

raised conservative hackles. (COCA) (modification)
b. kindle Y’s wrath “to anger”
God’s wrath has been kindled. (Google) (passivization)

My wrath, not yours, his expectations kindled.
(topicalization)
You have kindled the Fire of God’s Wrath. (Google)
(modification)
c. ruffle Y’s feathers “to annoy”
My feathers were ruffled and the more I thought about it the more
ruffled they got. (Google) (passivization)
Her feathers, the misbehaving child ruffled. (topicalization)
This love story ruffles my emotional feathers every time I see it.
(Google) (modification)
d. catch Y’s eye “to fascinate”48
*His eyes were caught by her look. (native speakers’ judgement)

(passivization)
* My eyes, this dress has caught. (native speakers’ judgement)
(topicalization)
*His calm soothing look has caught her big eyes.
(native speakers’ judgement) (modification)
In addition to Corver et al.’s (2017) and Harwood et al.’s (2017)

proposals concerning alienable and inalienable possession, Alexiadou
(2003), and Lin (2007) argue that the relation between a possessor and an
alienable noun (as in ICEs in (4.56)-(4.57)) requires more functional
48
The fact that the idioms catch Y’s eye “to fascinate” fails the canonical syntactic
diagnostics does not make simultaneously the idiom an IdP, because of the open
slot realised in the position of the possessor, which is not expected in IdP-fixed
idioms. Instead, the idiom is expected to be a “mismatch” of ICE-idioms. An
example of an IdP-idiom with a possessor in its object can be open Pandora’s box
“to start something that causes many new and unexpected problems,” which does
not belong to our dataset of psychological idioms.
structure than inalienable possession (in IdPs). Hence, the relation in

alienable nouns is constructed by means of a higher functional head n*
introducing the possessor, as illustrated in (4.57) for the psychological ICE
idiom Type (A’): kindle Y’s wrath “to anger.”
(4.57) Alienable possession in ICE idioms, e.g. kindle Y’s wrath “to
anger:”
The relation between a possessor and an inalienable noun (“possessum”),

instead, is more direct. The inalienable noun (presumably in IdPs but also
in some ICEs) subcategorises for a possessor argument, with which the
noun composes to form an XP that behaves like a small clause (cf.
footnote 44; Harwood et al.’s 2017: 59), as shown in (4.58) for the
psychological idiom Type (A’): catch Y’s eye “to fascinate.”49
45
An interested reader may also find some phase-bound account, proposed by
Svenonius (2005), which is left here aside for the lack of space. Svenonius (2005),
referring to multidominant trees, suggests structures called Banyan trees, which
include more than one undominated node. Banyan trees are created when first
Merge as well as internal Merge can target subconstituents of structures already
built by Merge. Thus, a psychological idiom with an open possessor position, e.g.
raise Y’s hackles “to annoy,” would be illustrated by means of a Banyan tree, as in
(i).
248 Chapter Four
(4.58) Inalienable possession in some ICE-idioms, e.g. catch Y’s eye “to
fascinate”
(i)
(cf. Svenonius 2005: 250)

According to Svenonius (2005), as presented in (i), the possessor can be merged
with possP, and the external argument with VP. The licensing material outside vP
is responsible for case-licensing the larger object DP, together with the possessor.
While DP-internal functional heads must be responsible for case-licensing the
possessor. Therefore, it is probable that the possessor DP spells out before the DP
is complete, and the lower (idiomatic) N may as well. Besides, some higher
structure must unify the Banyan
tree structure so that both vP and possP are contained under the same node (cf.
Svenonius 2005: 250). This way, Svenonius (2005) extends Chomsky’s (1993)
Extension Condition for discontinuous idioms (in Chomsky’s (1993: 22-23) terms,
the Extension Condition is a basic assumption about the Merge operation, namely
that it always “extends its target”).
As for possessors, which typically are not recognised as part of an idiom,
Svenonius (2005) argues that if the genitive case on the possessor is assigned by
the possessive head, then the possessor DP is to spell out as soon it is merged, and
is not to belong to idiom components (cf. Svenonius 2005: 252).
All in all, as shown in (4.57) and (4.58), the higher functional head is
absent in inalienable possession structures (cf. Alexiadou 2003; and Lin
2007; among others). Interestingly, Alexiadou (2003) notes that a
subject/agent relation between the possessor and the possessed noun is
reflected in the case of inalienable possession in ICE idioms, which
accounts for the presence of n* in alienable possession structures. Only
this type of idioms contains a DP-phase. Inalienably possessed nominals,
in turn, do not project n*P, and thus do not exhibit a DP-phase. This kind
of phase-bound analysis correlates well with the explanation provided in
section 4.5.2 above for nominal phrases in IdP fixed idioms, which lack
n*, and thus do not constitute a DP-phase, but belong to the vP-phase the
whole IdP idiom is confined to. However, idiomatic nouns in ICE-idioms
which are alienably possessed build independent n*Ps and form phases.
4.5.4 Psychological idiomatically combining expressions (ICEs)

with Double Object Constructions in the Phase Theory
In this section, the psychological idioms with Double Object
Constructions (DOC) are to be analysed in terms of possible phases that
these idioms may contain.50 The idiomatic expressions under scrutiny
50
I refer an interested reader to some previous approaches to DOC constructions,
e.g. to Larson’s (1988) DOC framework, who posits a hierarchical structure for the
VP, involving two VP-shells, i.e. higher VP (V1P) and lower VP (V2P). The
structure is known as a “VP-shell” structure since the highest VP contains the
lower VP as a complement. In his proposal, Larson (1988) claims that the subject
is generated in SpecV1P; the Theme, i.e. the indirect object (IO) is generated as the
specifier of the lower VP (V2P); while the direct object (DO) as the complement of
V2. The verb starts out in the lower head position V2’ and moves to the higher
position V1’. The psychological idiom with a DOC, e.g. give Y the blues “to
depress,” is supposed to be illustrated in Larsonian shell structure, as in (i) below:
(i) Larsonian shell structure for the psychological idiom with a DOC
e.g. give Y the blues “to depress”
250 Chapter Four
represent Type (E), in which an Experiencer Y, realises an open position,
(cf. Larson 1988: 342-343)

For Larson (1988), the “PP-frame” structure (give a book to John) is basic, in
which the DO (a book) precedes, and c-commands the IO (to John). Whereas the
“double object frame” (give John a book) is derived by a syntactic operation, in
which the IO (John) precedes and c-commands the DO (a book) (cf. Pesetsky
1995; Harley 2002; and Lechner 2006).
Moreover, in his analysis of idiom chunks, Larson (1988: 340) argues that the
existence of idioms, like those listed in (ii), is an argument for a D-structure
constituent comprising a verb and a PP, with the direct object NP generated
external to that constituent. The fact that a verb may form a “discontinuous idiom”
with its outer arguments became Larson’s (1988) initial evidence for an articulated
VP-shell structure. The idiomatically interpreted constituents in (ii) are marked in
italics.
(ii) a. Larson’s examples of VP + NP+ PP idioms:
send NP to the showers (Lasorda sent his starting pitcher to the
showers)
take NP to task
throw NP to the wolves (Larson 1988: 340)
b. Psychological VP + NP+ PP idioms:

frighten / scare NP to death “to horrify”
drive NP to distraction ”to anger”
put / send / throw NP into a (blue) funk “to depress”
Finally, double complement idioms should be expected to freely shift.

Nonetheless, this does not turn out to be true, e.g. * frighten death NP. Therefore,
Larson’s (1988) approach to the double object alternation has been found
unsustainable.
refers to some direct referent in the discourse, and takes the role of an
indirect object. The role of a direct object is associated with a fixed idiom
component, as shown in (4.59a-f).
(4.59) Psychological idioms of Type (E): V + NP + NP, e.g.
a. give Y the blues “to depress”

b. throw Y a curve (ball) “to surprise”
c. give Y a turn “to horrify”
d. give Y (quite) a (bit of) fright “to scare”
e. give Y goose pimples “to scare”
f. rub Y (up) the wrong way “to annoy”
As shown in (4.59a-f), the fixed idiom element, in the role of the direct
object, can be represented either by a definite noun (e.g. the blues, the
wrong way), by an indefinite nominal unit, both singular (e.g. a curve, a
ball, a turn, a fright) and plural (e.g. pimples), by a compound noun (e.g.
goose pimples), or by a noun modified with an adjective (e.g. the wrong
way), or a quantifier (e.g. quite, a bit of). The verbal predicate of this type
of psychological idioms is exemplified by an agentive transitive verb give,
throw, or rub (up), the last of which can be accompanied by a particle up.
Nunberg et al. (1994), and Espinal and Mateu (2010) assume that open
object positions are one of the characteristics ICEs do allow for, which
makes them more compositional, while fixed slots are typical of canonical
IdPs. Following this claim, the psychological idioms in (4.59a-f), can be
treated as ICEs, composed of a lexical verb and a theme, with an open
position for the thematic goal.
The starting point in our phase-based analysis of psychological ICEs
with DOCs is McGinnis’ (2001) claim that certain types of double object
constructions may be phases, while the relationship between the direct and
the indirect object in a DOC is mediated by a functional head, called an
applicative head51 (cf. Pylkkänen 2008). Pylkkänen (2002, 2008),
developing a suggestion made by Marantz (1993), distinguishes high and
low applicative structures, as presented in (4.60) and (4.61), respectively.
51
The applicative head conveys a large number of meanings: benefactive,
malefactive, instrumental, locative and comitative meanings, among others (cf.
Citko 2014: 152).
252 Chapter Four
(4.60) high applicative (ApplH)
(4.61) low applicative (ApplL)
As demonstrated in (4.60), in a high applicative structure (ApplH), the

direct object is the complement of the verb, while the indirect object is the
specifier of a high applicative head, which takes a VP as a complement.
The relationship between the indirect object and the event which is
described by the VP including the direct object, is established by a high
applicative head. In a low applicative structure (ApplL), presented in
(4.61), the indirect object c-commands the direct one, so the relationship is
the same as in a high applicative, with a difference that this relationship is
more direct. The applicative phrase, having the indirect object as its
specifier and the direct object as its complement, is a complement of the
verb.
Furthermore, the differences between these two applicatives involve

compatibility of an applicative structure with intransitive and stative verbs,
and the ability of a depictive modifier to modify the indirect object. In
other words, high applicatives are possible within transitive and stative
verbs, and it is possible for depictive modifiers to modify an indirect
object; whereas low applicatives are immune to both of these tests (cf.
Citko 2014: 152). Moreover, English fails these diagnostics, because, first,
applicative objects are disallowed with intransitive verbs (*Maria ran Jan,
or *Maria died Jan); second, applicative objects (indirect objects) are
disallowed in English with stative verbs (*I held Maria the door); and
third, the indirect object cannot be modified by depictives (*I gave Mariai
the book curiousi). Therefore, it is difficult to assume which applicative
structures the psychological idioms with DOC, listed in (4.59a-f)
represent.
Significantly, McGinnis (2001), Kim (2014, 2015), and Citko (2014),
among others, argue that a high applicative is a phase head but a low
applicative is not. Therefore, high applicatives can have the EPP52 feature,
which makes movement of the direct object over the indirect object
possible, as shown in (4.62) below.
(4.62)
52
The Extended Projection Principle (EPP), proposed by Chomsky (1982), is a
linguistic hypothesis about subjects. The EPP refers to clauses which are required
to contain a NP or DP in the subject position viz. in the specifier of TP or IP (cf.
Chomsky 1982: 10). Importantly, the EPP-feature forces an item equipped with it
to project a specifier.
254 Chapter Four
Accordingly, the phase head, ApplH, becomes a structural boundary to

define idiomatic interpretations (4.62), but a non-phase head, ApplL,
instead, is not able to delimit the idiomatic interpretation. However, Kim’s
(2014: 220; 2015: 633) assumptions concerning the structural boundaries
for idiomatic interpretations in English ditransitive idioms differ from the
aforementioned ones (cf. the discussion below and the syntactic
representations for idioms in (4.64), (4.65), and (4.66)).
Indeed, in his accounts of the distribution of idioms in English
ditransitives, Kim (2014, 2015) offers a different view than Bruening
(2010) does in his idiom-as-selection hypothesis.53 While Bruening’s
(2010) proposal lacks an explanation of how the specifiers of functional
heads are treated as regards idiomatic interpretation, Kim (2014, 2015)
offers an account for the distribution of idioms in English DOC and
Postpositional Dative (PD) in terms of phases. Following Bruening’s
(2010) patterns of idioms, Kim (2014, 2015) predicts both possible idioms
and patterns which are non-existent, as illustrated in (4.63), with
exemplary English psychological idioms fitting these patterns (idiomatic
53
In his idiom-as-selection proposal, built on O’Grady’s (1998) work, Bruening
(2010) argues that in order for two syntactic constituents, X and Y, to form an
idiom, one must select the other; and it is via selection that two elements are
combined together and interpreted contextually (cf. Bruening’s (2010) Constraint
on Idiomatic Interpretation, formulated in (4.39), and repeated in (i) for
convenience.
(i) If X selects a lexical category Y, and X and Y are interpreted idiomatically, all
of the selected arguments of Y must be interpreted as part of the idiom that
includes X and Y.
Lexical categories are V, N, A, and Adv.
(Bruening 2010: 532 (25-26))
To be more specific, in V-NP idioms (e.g. pull strings, or paint the town), the
verbs select direct objects, and then, taken together, they can be interpreted
idiomatically. In the case of verb-theme English ditransitive idioms (e.g. give NP
the boot, give NP a turn), they are proposed by Bruening (2010) to have ApplHP
and the following structure [ApplHP NP [ApplH' [ApplH [VP [V give] theme]]]].
ApplH selects V, which selects the theme. ApplH and V are interpreted
idiomatically, and thus all selected arguments of V have idiomatic interpretation.
Nonetheless, Bruening’s (2010) analysis lacks an explanation why the specifiers of
ApplH, i.e. the NPs in e.g. give NP the boot, give NP a turn, are excluded from the
idiomatic interpretations. Kim (2014: 216; 2015: 648) argues that it is not clear
why the specifier of ApplH is excluded from the idioms, provided they are also
selected by ApplH. Therefore, Bruening’s (2010) idiom-as-selection approach
seems not to rule out the specifiers of functional heads in a sufficiently explanatory
way.
parts are underlined; while X is a Causer, and Y is an Experiencer, and

both of them realise an open slot in an idiom).
(4.63) Distribution of possible ditransitive idioms, with relevant data of

psychological idioms with DOCs and a PD:
a. Class I: Verb NP NP
(give Y the blues “to depress,” give Y a turn “to
horrify”)
b. Class II: Verb NP to NP
(take a liking to X “to love;” pay court to Y “to
fascinate”)
c. Class III: Verb NP to NP
(send Y into a funk “to depress”54)
d. Class IV: V NP NP (non-existent)

(cf. Bruening 2010: 532; Kim 2014: 216; Kim 2015: 646)
The possible patterns for psychological ditransitive idioms, in the light of

Kim’s (2014, 2015) phase-based proposal, would be visually presented as
in (4.64)-(4.65). According to Kim’s (ibid.) idea, Class I (and a non-
existent class IV) idioms are DOCs which do have ApplHP, but the
boundary for idiomatic interpretation is placed below the ApplHP, to
exclude the open slot of the specifier, as shown in (4.64).
54
The psychological idiom send Y into a funk “to depress” seems to be of the same
PD (postpositional Dative) pattern as Bruening’s (2010) original example for class
III ditransitive idioms (Class III: Verb NP to NP (send X to the showers “to remove
from a job”). However, the status of other idioms with the preposition “to” is
ambiguous. For instance, some psychological idioms of this type appear to be
regular VP-NP+PP idioms (e.g. drive Y up the wall “to annoy”), while others seem
to be resultative structures (e.g. frighten / scare Y to death “to horrify”).
256 Chapter Four
(4.64)
Moreover, according to Kim (2014, 2015), Class II and III ditransitive

idioms, demonstrated in (4.65) and (4.66), are PDs which have a VP
complement that embeds a PP.55
55
Svenonius (2005), in his idioms-listed in the lexicon approach, also predicts a set
of permissible idioms. Illustrating his patterns of permissible idioms with
psychological examples, they would be as follows: (i) idioms with free Goal
objects in a double object structure (give Y a turn “to scare”); (ii) idioms with open
possessor positions (raise Y’s hackles “to annoy”); (iii) but no idioms with free
Theme objects in a double object structure (*give a fright Y), which is also
confirmed by the data of our study on psychological idioms.
(4.65)
(4.66)
As suggested by Kim (2014, 2015), ApplH in English ditransitives can

become a phase boundary for idiomatic interpretations; excluding the
specifier of ApplH, which realises an open slot (Class I). Class II and III
ditransitives do not include ApplH, and thus cannot become phases, but
258 Chapter Four
their idiomatic parts comprise verbs and objects that appear in the domain
of the VPs. Indeed, Kim’s (2015) proposal accounts for DOC-idioms,
including psychological idioms, but the puzzle why the phase boundary in
phase-bound Class I idioms is restricted to VP (which is commonly not
treated as a phase), and not to ApplHP (which is assumed to be a phase)
remains unsolved.
4.5.5 PP-phases in psychological idiomatically combined

expressions (ICEs)
The starting point in the discussion concerning the possible phasehood of
PPs refers to the common assumption that only functional categories are
capable of being phases (cf. Marantz 1984, 1997; Chomsky 2000; and
Citko 2014; among others). Thus, the phase-like status of PPs depends on
whether prepositions are functional or lexical.
Prepositions are modifiers that indicate location, origin or direction.
For some researchers, prepositions are principally lexical, but possibly
dominated by a functional layer (e.g. Jackendoff 1977; den Dikken 2010).
Other linguists consider prepositions as functional (e.g. Grimshaw 1991;
Kayne 2004), while some others treat some prepositions as lexical and
others as functional (e.g. Yadroff 1999; Abraham 2010). In his systematic
discussion related to the differences between the two types of prepositions,
Yadroff (1999) argues that functional prepositions are mainly unstressed,
monosyllabic and polysemous, since their meaning is more abstract (e.g.
for, down, in, up, off, out of); whereas lexical prepositions tend to be more
complex, i.e. stressed, polysyllabic (often polymorphemic), and have a
more fixed meaning (e.g. during, while, instead of, except, above) (cf.
Citko 2014: 140). In short, functional elements, when compared to lexical
categories, lack “descriptive content,” they perform certain functions, they
are phonologically and morphologically dependent, and they do not
undergo derivational processes (cf. Abney 1987; and Citko 2014: 138-
139).
However, for Citko (2014: 141), the existence of two types of
prepositions does not contradict the assumption that all PPs can be phases,
on condition that lexical prepositions (as not-phase heads) may be
dominated by functional heads, which are phases. Indeed, according to the
phasehood diagnostics, prepositions are the source of uninterpretable
features; a PP is a domain for feature valuation; movement out of PP has
to proceed through the edge of PP; Quantifier Raising cannot target PP;
the complement of P cannot be deleted; and a PP can be a binding domain
(cf. Citko (2014: 141-151) for a more detailed discussion). Regarding the
phasehood status of PPs, Citko (2014) states that, nonetheless, the

behaviour of PPs with respect to phasehood diagnostics is mixed.
In the same vein, Koopman (2010), Aelbrecht and den Dikken (2013),
and Corver et al. (2017), suppose that PPs represent the type of structure
that may potentially project a phase. As far as idioms containing a PP are
concerned, all of these expressions are expected to display ICE properties
since they allow syntactic modification, such as passivization, topicalization
and quantification, among others (cf. section 4.3). A PP-phase in ICE
idioms does not contradict a phase-bound analysis since ICE-type idioms,
being not confined to a single phase, are able to span multiple phase
boundaries and freely straddle phase boundaries (cf. Corver et al. 2017:
17-18).
Consequently, the seven patterns of psychological idioms with a PP-
goal/oblique found in our dataset have the following syntactic phase-based
representations, as in (4.67b) for Type (B), Type (B’), Type (C), and Type
(D’), and in (4.68) for Type (C’), Type (D), and Type (D’’).
(4.67) a. Patterns of ICE-idioms which have the representation as in

(4.67b):

carry a torch for X “to love”
Type (B’): V + NP + preposition+possessor +N (complement of a P)

have a yellow streak / belly down Y’s back “to fear”

hold X in abomination / contempt “to hate”
drive Y up the wall “to annoy”
Type (D’): V + NP + preposition+possessor +N (complement of a P)

sweep Y off Y’s feet “to fascinate”
260 Chapter Four
(4.67) b. A syntactic representation for Types (B), (B’), (C), and (C’)
As shown in (4.67b), in Type (B), Type (B’), Type (C), and Type (D’) ICE
idioms, a phase boundary for idiomatic interpretation is marked by a PP,
which is the complement of a VP. Since the prepositions (for, down, in,
up, off, etc.) in all the idiom types are functional, the PPs are capable of
being phases (cf. Marantz 1984, 1997; Chomsky 2000; and Citko 2014;
among others). The complements of the Ps are either fixed nominals (e.g.
in contempt, or up the wall) or nominals that comprise possessors (down
Y’s back, off Y’s feet). The object DPs in the aforementioned types are
represented either by fixed nominals with indefinite articles “a” (a torch, a
belly) or realising an open position (X or Y), and they belong to the VP.
Regardless of whether or not the DPs in these ICE-type idioms are
referential or not; and hence whether there are DP-phases formed or not,
these idioms do not pose a problem, since ICEs can straddle a number of
different phases.
Furthermore, the idiomatic phase-based structure in (4.68b) represents
Type (C’), Type (D), and Type (D’’) psychological ICE idioms, with the
patterns as in (4.68a).
(4.68) a. Patterns of ICE-idioms which have the representation as in

(4.68b):
Type (C’): V+ possessor+N + preposition + NP (complement of a P)

put Y’s nose out of joint “to annoy”
Type (D): V +possessor +N +preposition +NP (complement of a P)

lose Y’s heart to X “to love”
Type (D’’): V+possessor +N + preposition +possessor +N (complement

of a P)
have Y’s heart in X’s mouth “to fear”
262 Chapter Four
(4.68) b. A syntactic representation for Types (C), (D), and (D’’):
As shown in (4.68b), the syntactic representation for Types (C), (D), and
(D’’) Type (B), differs from the one in (4.67b) only in the type of a DP-
object, which is a specifier of the VP. In (4.68b), the DP would be given a
chance to become a DP-phase if only the nominals were alienably
possessed as in (4.57). Instead, the DPs in these idioms refer to parts of
one’s body, which makes them inalienable; hence, no DP-phase is attested
in a VP in (4.68b). Nonetheless, these ICE-idioms build PP-phases,
formed by functional Ps (e.g. out of, to, in) and the complements of the Ps,
i.e. fixed NPs (joint), or NPs which realise an open position (e.g. X), or
NPs which comprise a possessor realising an open position (e.g. X’s
mouth).
In short, assuming that PPs can work as phases as long as they, among
other requirements, include functional Ps (cf. Kim 2014, 2015; and
Harwood et al. 2016), the psychological ICE-idioms which represent the
seven patterns presented in this section do comprise PP-phases. Besides,

these idioms may also contain DP-phases, provided the nominals in the
idiomatic DPs are alienably possessed. The coexistence of the two phases
does not give rise to any problems since ICEs can straddle more than one
phase (cf. Harwood et al. 2016).
4.5.6 Psychological idioms with particles

Particles have a variety of uses, such as idiomatic, directional and
aspectual. Svenonius (2005), in his account of verb-particle expressions,
claims that particles can be idiomatic (eat Y”s heart out “to worry,”
put/get Y’s back up “to annoy”), in contra-distinction to verb preposition
constructions which usually are not (worry about X). Since Chomsky
(1957), it has been proposed that a verb and a particle form a constituent,
excluding the direct object at some underlying level of representation,
which may be disrupted later on as a result of movement (cf. Chomsky
1957; Johnson 1991; and Svenonius 2005: 237; among others).
In general, in the literature, verb particle constructions are considered
to be base generated, as part of a complex V, without projecting a P-type
syntactic projection (cf. Koster 1975; and Johnson 1991). The projection
which contains the particle is assumed to be either a projection of an
intransitive P with no complement (cf. Emonds 1976, 1985), or a
projection of some type of small clause. In the latter option, the argument
of the verb particle combination may originate, either in the subject
position of the particle (cf. Kayne 1985), or in the complement position of
the particle (cf. Guéron 1986; den Dikken 1992; and Koopman 1991,
2010; among others). For the purpose of our analysis of idiomatic verb
particle constructions, it is assumed that the complement originates within
the PP. Moreover, idiomatic particles, as thematically complex structures
with V, are said to lack autonomous theta-properties, and similarly to
unaccusative verbs, do not assign accusative Case (cf. Kayne 1985).
Providing that idiomatic particles have no independent lexical properties,
they are believed to project a PP without any functional layers, as
suggested by Koopman (2010: 62), and displayed in (4.69) below.
264 Chapter Four
(4.69) Verb particle constructions: V takes a bare PP complement
As shown in (4.69), particles P are complements of VPs, while the DP

complements of Ps can escape the PP via Spec, PP (cf. Koopman 1991,
2010).
In our analysis of psychological V-particle idioms of Type (A’’), e.g.
eat Y’s heart out “to worry,” or knock Y’s socks off “to surprise,” the
idioms have been first tested as regards their ICE or IdP status, by means
of the three syntactic diagnostics applied by Harwood et al. (2017), i.e. the
possibility of idioms for passivization, topicalization and modification.
These are summarised in (4.70a-b) below.
(4.70) a. eat Y’s heart out “to worry”
Is it strange that her judgment was perverted, and her heart was
eaten out. (Google) (passivization)
Her heart, the disappointed love for him was eating out.
Napoleon at St. Helena is eating his proud heart out with rage.
b. knock Y’s socks off “to surprise”
If his socks had been knocked off, she couldn’t tell a word.
(Google) (passivization)
My socks, the news knocked off.
What I read in the files knocked off my proverbial socks

Once the idioms have been confirmed to be of ICE-type, their DP

arguments have been checked if they are alienably possessed or not. In
fact, it has turned out that some of the particle-idioms have alienably
possessed DPs (Y’s socks)56 and are confined to a phase boundary, as
presented in (4.71a). Other particle-idioms, in turn, have inalienably
possessed DPs (Y’s heart),57 which makes them unable to form phases, as
represented syntactically in (4.71b) below.
(4.71) a. The syntactic representation of ICE-idioms with particles and

alienably possessed DPs:
Type (A’’): V+ possessor + N + particle
(e.g. knock Y’s socks off “to surprise”)
56
Cf. the structure in (4.57) for alienably possessed DPs in ICE-type idioms.
57
Cf. the structure in (4.58) for inalienably possessed DPs in ICE-type idioms.
266 Chapter Four
(4.71) b. The syntactic representation of ICE-idioms with particles and

inalienably possessed DPs:

(e.g. eat Y’s heart out “to worry”)
All in all, as shown in (4.71a) and (4.71b), both the inalienable or

alienable DPs in the aforementioned psychological idioms originate as
complements of the Ps, and then they are moved into the Spec, PP
position. However, only psychological particle-ICE idioms with alienably
possessed DPs (e.g. knock Y’s socks off “to surprise”) contain DP-phases,
in contradistinction to ICEs with inalienably possessed DPs that do not act
as phases.
4.5.7 Small Clauses as Predication Phases in psychological

idioms
In this section, the last type of psychological idioms is to be analysed, viz.
Type (E’), in which the idiomatic verb is followed by a [NP XP] sequence.
While NP realises an open position, representing the Experiencer (Y), XP
is a fixed idiomatic element, which can be filled in either by an AP, AdvP,
PP, or NP, as illustrated in (4.72).
(4.72) Psychological idioms of Type (E’): V + [NP XP]:
a. V + [NP AP]:
drive Y batty “to annoy”
leave Y open-mouthed “to surprise”
strike Y dumb “to surprise”
b. V + [NP AdvP]:
knock Y sideways “to scare”
bring Y low “to depress”
catch Y unawares “to surprise”
c. V + [NP PP]:
strike Y with wonder “to surprise”
strike Y with awe “to surprise”
leave Y at a loss for words “to surprise”
d. V + [NP NP]:
drive Y nuts/ bananas / bonkers “to annoy”
The [NP XP] sequence occurring after a verb, as exemplified on the

instances of psychological idioms in (4.72) above, has been given various,
often competing, interpretations in the literature. Within the framework of
Government and Binding Theory (Chomsky 1981), two major proposals
have received some linguistic significance, viz. the Small Clause Theory
and the Predication Theory. The former recognises the mentioned string as
a clause and a single constituent, and the NP and the XP in a subject-
predicate relationship but lacking tense inflection (Stowell 1981, 1983;
Chomsky 1981; and Safir 1983; among others). The latter treats the NP
and the XP as two distinct complements of the main verb, which do not
form a single constituent (Williams 1980, 1983). These two approaches to
the postverbial [NP XP] sequence, are summarised in (4.73a) and (4.73b),
for the Small Clause Theory and the Predication Theory, respectively.
(4.73) a. I [VP consider [SC John silly]]

b. I [VP consider [NP John] [AP silly]]
(Safir 1983: 731)
c. I [VP consider [John to be silly]]
d. I [VP consider [ that John is silly]
268 Chapter Four
Interestingly, the subject in the two aforementioned theories is defined in

different ways. In the Small Clause Theory, it is structurally characterised,
i.e. as an NP is dominated by S. In the Predication Theory, in turn, the
non-structural NP subject is an external argument, which combines with
any predicative phrase (cf. Williams 1983).58 Yet, the [NP XP] sequence,
commonly associated with a term “small clause” (henceforth SC), refers to
“not fully fledged” clause constructions. Moreover, SCs are usually
distinguished from their full clause counterparts, e.g. the one given in
(4.73c), which includes tense specification, albeit a non-finite one, in
addition to the subject and the predicate, and the one in (4.73d), which is a
full tense-inflected clause, and a clausal complement of the VP consider
(for a more detailed discussion concerning the internal structure, category
and typology of small clauses cf. Chomsky 1981: 106; Radford 1988a,b;
Aarts 1992; Kim, J.-B. 2013: 76; and Tomacsek 2014; among others).
Undoubtedly, with so many competing analyses of SCs offered in the
literature so far (Stowell 1981, 1983; Kitagawa 1985; Aarts 1992; Bowers
1993, 2001; Haegeman 1994; den Dikken 2006; and Citko 2008, 2011;
among many others), the very nature of small clauses is far from being
uncontroversial, as is their categorial status and internal structure.
Nonetheless, for the purpose of our discussion undertaken in the book, the
focus is put on the idea that there is a relationship between predication and
phasehood. Staying in line with Chomsky’s (2007, 2008) view that phasal
properties should be extended to other phrases, in addition to CPs, vPs and
DPs, Citko (2014) argues that SCs may receive a phase-based analysis, as
well. Yet, it is the approach proposed by Bowers (1993), towards
predication and phasehood, which is the one adopted here for the case of
psychological idioms with small clauses. Bowers (1993) suggests
expanding the inventory of functional projections to include a Predication
Phrase (PrP), which corresponds to a SC.59 The head of PrP is recognised
58
Cf. Bruening’s (2016) discussion concerning resultative constructions analysed
as small clauses, and other Small Clause approaches to the argument structure.
59
Jiménez-Fernández and Spyropoulos (2013), similarly to Cinque (1990) and
Haegeman (2010), assume that SC structure includes a functional projection F
above the projection of the category that functions as the predicate. Jiménez-
Fernández (2000) recognises some more possibilities of including other functional
categories in the structure of SCs, e.g. Aspectual Phrase. Bowers (1993, 2001)
posits a Predication Phrase above the SC; while Starke (1995) holds that SCs are
full clauses, and hence project a CP. For reasons of space, it is not possible to
discuss these different proposals related to the functional spine of SCs in detail.
Thus SCs are analysed here, after Bower (1993) and Citko (2014), as Predication
Phrase. Importantly, SCs are treated as reduced clauses, which lack a CP and a TP.
by Citko (2014) as a phase head, which heads a small clause (SC), as

illustrated in (4.74) below.
(4.74) A structural representation for a SC

I [VP consider [SC parrots smart]]:
(Citko 2014: 132)
In her structural representation for SCs, presented in (4.74), Citko (2014)

recognises PrPs as phases. Citko (2014) provides some convincing
evidence for the phasehood of PrPs, at least in some languages with
respect to most phasehood diagnostics. Consequently, she argues that (i) a
Pr can be the source of uninterpretable features; (ii) movement out of a
PrP has to proceed through the edge of the PrP; (iii) Quantifier Raising can
target a PrP; (iv) the complement of a Pr cannot be deleted; and (v) a PrP
is semantically complete (propositional) since it consists of a subject and a
predicate (cf. Citko 2014: 137, for more details relevant to phasehood
diagnostics). What is more, Bowers (2002), Harves (2002), Matushansky
(2000), and Tanaka and Yokogoshi (2010), among others, also argue
explicitly for the phasehood of Predication Phrases.
Once the phasehood of PrPs is established, and the inventory of
functional projections is to include a Predication Phrase (PrP), as
suggested by Bowers (1993, 2002) and Citko (2014), the PrP structure can
become a possible structure for small clauses, also those embedded in the
270 Chapter Four
psychological idioms from our dataset with SCs listed in (4.72). By

analogy with the small clause I [VP consider [SC parrots smart]], offered
by Citko (2014: 132), and represented structurally in (4.74), psychological
idioms under scrutiny are believed to have the following structure, as in
(4.75a-d).
(4.75) Psychological idioms of Type (E’): V + [NP XP]:

272 Chapter Four
As shown in (4.75a-d) for Type (E’) psychological idioms, the subject of

the embedded SC is always represented by an Experiencer (Y), whereas
the predicate comprises either an AP, AdvP, PP, or NP. Importantly, all
predication relations are mediated by a Pr head, as proposed by Bowers
(1993, 2002); and the Predication Phrases in all the cases are certainly
semantically complete, since they contain a predicate with all its
arguments, which is the behaviour expected of phases (cf. Citko 2014).
Furthermore, as noted by Citko (2014: 132), the Spell-Out of the
complement of Pr occurs as soon as the next head up (v) is merged, as
shown in (4.76).
(4.76) The Spell-Out of the complement of Pr
In brief, a PrP structure is not the only possible structure for small
clauses (cf. Citko (2011) for an overview of possibilities), but it seems to
work efficiently for the instances of psychological idioms of Type (E’).
To conclude, section 4.5 has offered an analysis of psychological

idioms in the light of the Phase Theory, which has been widely adopted by
many researchers working in the field (e.g. Svenonius 2005; Stone 2009;
Harwood 2013, 2017; Harwood et al. 2016; Harley and Stone 2013; Kim
2014, 2015; and Corver et al. 2017; among others). It has been noted that,
in addition to CPs and vPs, also DPs, ApplHP, or PrPs may act as phases,
which has contributed significantly to our analysis of psychological
idioms.
This section of the chapter has started with a claim that the vP-phase
imposes an upper bound on verbal IdP idioms. Indeed, IdPs are confined
to a single phase, namely the clause-internal phase; whereas ICEs are able
to span multiple phase boundaries and can depend on the material beyond
274 Chapter Four
them. Moreover, nearly all idioms from our dataset can be characterized as
having an open position located, e.g. in the direct object, small clause
subject, or possessor.
4.6 An Experiencer in psychological idioms

It is commonly accepted in the literature that the defining characteristic of
psych-verbs is to express (a change in) mental or/and emotional state and a
relation between the two arguments: an Experiencer and the Cause /
Theme of such a psychological condition (cf. Belletti and Rizzi 1988;
Dowty 1991; Pesetsky 1995; and Landau 2010; among others). There is
also quite a great number of phraseological units which may refer to one’s
psychological condition, as has been confirmed by the results of our
corpus research in Chapter Three. In other words, 161 idiomatic
expressions elicited in the corpus study have become the counterparts of
the chosen top psych-verbs. The psychological VP-idioms, similarly to the
psych predicates, comprise a participant who experiences some emotional
or mental state, i.e. an Experiencer, and a Stimulus / Causer / Cause /
Target, which has contributed to this specific state or become a target of it.
With this in mind, the three questions to be addressed in this section
concern: (i) the position of an Experiencer in psychological idioms; (ii) an
Experiencer in psych-idioms analysed as a mental location, following
Landau’s (2010) proposal; and (iii) the place of an Experiencer in Baker’s
(1989) Thematic Hierarchy.
4.6.1 The position of an Experiencer in psychological idioms

To start with, one of the basic criteria specified at the beginning of the
corpus search for psychological idioms was to elicit only those
phraseological entities which can be used, instead of common
psychological predicates. Therefore, it has been assumed that SE psych-
verbs (class I, following Belletti and Rizzi’s (1988) tripartite
classification) are to be prototypes for those idioms, with the main
argument (Experiencer) in the subject position, as presented in (4.77).60
60
If an Experiencer is the trigger of emotion, it behaves like an agent, then “the
experiencer (…) resembles the agent to the extent that his involvement in the
action is volitional and he has control over the execution of the mental process”
(Dąbrowska 1997: 94). Therefore, the performer of the process appears rather
“active than passive, conscious rather than lacking consciousness, rational rather
OE psych-verbs (class II and class III), in turn, are to correspond to such

psych idioms, which have an Experiencer realised in an object position, as
exemplified in (4.78).
(4.77) Examples of SE psych-verbs and their idiomatic counterparts:

a. Mark loves his girlfriend Lucy. (SE psych-verb)
Mark carries a torch for his girlfriend Lucy.
(Type (B) idiom)
Mark has lost his heart to his girlfriend Lucy.
(Type (D) idiom)
b. Mark was enjoying himself in the evening. (SE psych-verb)

Mark was painting the town in the evening.
(Type (A) idiom)
Mark was kicking his heels in the evening.
(Type (A’) idiom)
Mark was getting a buzz out of dancing in the evening.
(Type (B) idiom)
c. Mark worries about his debts. (SE psych-verb)

Mark has the blues because of his debts. (Type (A) idiom)
Mark eats his heart out because of his debts.
(Type (A’’) idiom)
Mark has lost his sleep over his debts. (Type (B) idiom)
Mark has a bee in his bonnet because of his debts.
(Type (B’) idiom)
(4.78) Examples of OE psych-verbs and their idiomatic counterparts:
a. The comedy amused Mark.

(OE psych-verb–class II)
The comedy gladdened Mark’s heart.
(Type (A’) idiom)
The comedy tickled Mark to death. (Type (C) idiom)
than irrational, [and] capable of referring to the objective world and seeing things
that are really ‘out there’” (Dąbrowska 1997: 97).
276 Chapter Four
b. The screaming children annoyed Mark.

(OE psych-verb–class II)
The screaming children raised the hump, so that Mark couldn’t
sleep.
(Type (A) idiom)
The screaming children raised Mark’s hackles.
(Type (A’) idiom)
The screaming children stirred up a hornet’s nest in Mark.
(Type (B))
The screaming children brought a hornet’s nest round Mark’s
ears. (Type (B’) idiom)
The screaming children put Mark’s nose out of joint.
(Type (C’) idiom)
The screaming children thumbed their noses at Mark. (Type
(D) idiom)
The screaming children gave Mark the pip.
(Type (E) idiom)
The screaming children drove Mark batty / bananas.
(Type (E’) idiom)
c. This film appeals to Mark. (OE psych-verb–class III)

This film tickles Mark’s fancy. (Type (A’) idiom)
This film sets Mark on his ear. (Type (D’) idiom)
As can be seen in (4.77) and (4.78), the position of the Experiencer in the
aforementioned psychological idioms corresponds to the position an
Experiencer takes in psych-verbs. In all the cases the parallelism goes in a
straight line, viz. if an Experiencer (Mark for all these instances) occupies
the preverbal position, having the role of a subject, it plays the same role
and takes a preverbal position in the case of psychological idioms. On the
other hand, if the Experiencer has the role of an object, its position both in
psych-predicates and their idiomatic counterparts is postverbal.
What is more, the Experiencer, both for SE and OE psychological
idioms, is realised mostly overtly as a DP. However, in the case of the
idioms corresponding to OE psych-verbs, the Experiencer is realised either
as a DP-direct object (e.g. gave Mark the pip), or as a possessor of the DP-
object (e.g. tickles Mark’s fancy), or as a possessor of the DP-complement
of P (e.g. brought a hornet’s nest round Mark’s ears), or as a PP
complement (e.g. thumbed their noses at Mark). The Experiencer may be
also covert, so that only the context of the discourse reveals who the
Experiencer is (e.g. The screaming children raised the hump, so that Mark
couldn’t sleep).
4.6.2 An Experiencer in psychological idioms

as a mental location
The objective of this part of the chapter is to check whether English verbal
idioms, which denote a psychological condition of an Experiencer and
occur with locative Prepositional Phrases (PPs), may serve as an evidence
for Landau’s (2010) hypothesis of “Experiencers as mental locations, that
is, locative” (Landau 2010: 6; cf Landau’s analysis discussed in section
2.5.2 of Chapter Two). Consequently, all psych-idioms corresponding to
class II and III psych-verbs can belong to the dataset to be investigated,
providing these OE idioms comprise a PP with an open slot, realised by an
Experiencer (Y). Thus, these are all OE idioms of Type (B), (B’), (D), and
(D’), as illustrated in (4.79).
(4.79) Examples of OE (class II and III) psych-idioms

with an Experiencer placed in a PP:
In TOTAL: 35 of these psych-idioms ĺ 32%

Type (B): V + NP + preposition + NP (complement of a P):
26 idioms ĺ 23%
cast a gloom / a shadow over Y “to depress Y”
blow a gasket on Y “to anger Y”
scare the bejesus out of Y “to horrify Y”
Type (B’): V + NP + preposition + possessor +N

(complement of a P)
3 idioms ĺ 3%
take the wind out of Y’s sails “to depress Y”
bring tears to Y’s eyes “to depress Y”
bring a hornet’s nest round Y’s ears ”to annoy Y”
Type (D): V + possessor + N + preposition + NP

(complement of a P)
4 idioms ĺ 4%
vent X’s spleen at Y “to anger Y”
thumb X’s nose at Y “to annoy Y”
shake X’s / a fist at Y “to scare Y”
278 Chapter Four
Type (D’): V + NP + preposition + possessor + N

(complement of a P)
2 idioms ĺ 2%
catch Y off (Y’s) guards “to surprise Y”
sweep Y off Y’s feet “to fascinate Y”
set / put Y on Y’s ear “to appeal to Y”
As can be seen in (4.79), four types out of twelve elicited in Chapter Three
consist of a PP, with an Experiencer overtly placed, which comprises
altogether 35 psych-idioms out of 112 OE class II and III (ĺ 32%). Types
(B)-(B’) are the most numerous, covering 26%. These types include a verb
and a fixed object, followed by a Prepositional Phrase realising an open
position in a complement of the P, i.e. either in an NP (V+NP+P +NP), or
in a possessor modifying the fixed noun (V+NP+P +possessor +N). Types
(D)-(D’), in turn, realise two open positions, i.e. in the object position and
within a PP, hardly ever occur in this idiomatic dataset (6%).
Interestingly, in 5 SE idioms out of 49 SE idioms (which comprises
10% of Type (B’) and (D’’)), the Experiencer is located in a PP, viz. the
Experiencer is realised as a possessor modifying the fixed noun placed in a
PP. The instances of these SE idioms are shown in (4.80) below.
(4.80) Examples of SE (class I) psych-idioms with an Experiencer

placed in a PP:
In TOTAL: 5 of these psych-idioms ĺ 10%
Type (B’): V + NP + preposition + possessor +N

(complement of a P)
have a bee in Y’s bonnet “Y worries about X”
have ants in Y’s pants “Y worries about X”
have butterflies in Y’s stomach “Y worries about X”
have a yellow streak / belly down Y’s back “Y fears X”
Type (D’’): V + possessor + N +preposition + possessor +N

(complement of a P)
have Y’s heart in Y’s mouth ”Y fears X”
Nonetheless, 10% in the entire set of SE psych-idioms sounds insignificant

to provide evidence for Landau’s (2010) and Fábregas and Marín’s (2015)
(cf. Arad 1998) assumptions about the locative nature of SEs. In fact, the
proposal that SE psych-verbs should be treated as mental locations has not
be confirmed by the relevant evidence available in the literature so far (cf.

section 2.5.2.1 in Chapter Two).
Consequently, only OE psych-idioms with a PP, such as those in
(4.79), can work as a dataset to check whether it is justified to treat these
idioms as “periphrastic variants” of psych-verbs with locative
Prepositional Phrases (PPs). Since in Landau’s (2010: 6) account,
Experiencers are regarded as mental locations–locatives, i.e., containers or
destinations of mental states or effects, Landau makes a claim that (i) all
Object Experiencers are oblique (or dative); (ii) Experiencers undergo
“locative inversion” (Landau 2010: 6). Landau’s (2010) standpoint is
compatible with Jackendoff’s (1990: 300 n.4) decomposition approach, in
which the relation between an Experiencer, and a non-Experiencer
argument (called Stimulus, Trigger of emotion, Causer or Target/ Subject
Matter, or Theme) (Landau, 2010: 10), can be presented in terms of a
conceptual representation. According to Jackendoff (1990), the
correspondence rules relating Conceptual Structures (CS) to Syntactic
Structures (SS) are directly related. Following Jackendoff (1990), any
psych-sentence can be represented by means of the functions BE, INCH,
and AT, as illustrated in (4.81b), roughly read as in (4.81c).61
(4.81) a. X frightens Y
b. [CS+ ([X]Į, [INCH [BE ([FEAR ([Į])], [AT [Y]])]])]
c. X causes fear of X to come to be IN Y.
(Jackendoff 1990: 300 n.4, cited in Landau 2005: 7)
Jackendoff (1990) indicates the semantic representation in (4.81) to

express the paraphrase “X causes fear to come to be in Y.” FEAR in
(4.81b), indicating a mental state, is somehow extracted from the verb,
becoming a thematic argument, a co-argument of the Experiencer. This
paraphrase implies the locational property of the Experiencer (Y), since
fear (X) is described as a moving Object and Y as a container for the
emotion. The Experiencer seems to be the object of a preposition (AT /
IN), which locates the mental state within it.62 Even though in
61
Following Jackendoff (1990), the meaning of each sentence relies on the
universal semantic categories that the relative construction is made up of. The
inventory of such categories include EVENT, STATE, THING, PATH, PLACE,
PROPERTY, and TIME. All these universal semantic categories can combine with
each other by means of functions such as IN, AT, BE, INCH, and so on.
62
Cf. Baker (1997) for a similar suggestion, Iwata’s (1995) “reversed” option for
the Experiencer to be placed within the mental state.
280 Chapter Four
Jackendoff’s (1990) analysis the target of fear equals its cause, Pesetsky
(1995) does not find this equation necessary.
Moreover, Bouchard (1995) treats one’s mental state as an independent
semantic argument, called psy-chose, which he names also a syntactic
argument since “in mental space, the psy-chose is somehow put in contact
with the argument it affects” (Bouchard 1995: 272). This “argument” can
absorb the emotion or feeling that the psy-chose denotes, as shown in the
periphrastic psych construction63 provided in (4.82a), or illustrated by
means of the OE in (4.82b).
(4.82) a. Cela a éveillé en Pierre une rage terrible.

“That awoke in Pierre a terrible rage.”
b. Cela a enragé Pierre.

“That enraged Pierre.”
(Bouchard 1995: 275 (ex. 35a,c))
Landau (2010: 10) further assumes that despite the fact that psych-
verbs are decomposed conceptually into an “action” light verb plus a
mental state (psy-chose), this does not imply that this decomposition
happens on the syntactic level as well.64 Instead, the locative preposition is
syntactically active no matter if the Experiencer is a bare nominal or not,
while syntactic activity happens in the mental state only when it is visible,
i.e. in periphrastic constructions.
In short, according to Landau’s (2010) localist approach, the
Experiencer designates a mental location, thus it is placed within the
structure of a PP. This PP may be headed by a lexical P (as in English
obliques) or a null P (as in Latin obliques); nevertheless, both cases are
structurally distinct from bare DPs (Landau 2010: 21-22). On the ground
of these fundamental assumptions of Landau’s (2010) localist and
decomposition theory of psych-verbs, the verbal idiomatic units that
63
Cf. section 2.5.2.1. in Chapter Two for more details.
64
Landau (2010: 137 fn. 2) mentions that, actually, some evidence has been
provided to justify the claim that periphrastic and synthetic psych constructions
have different forms in some semantic aspects that cannot be ascribed as the single
factor of incorporation. While in non-agentive contexts, periphrastic forms are telic
(as in (ii)), synthetic forms are not (as in (i)). Thus, simple N-to-V incorporation
does not imply such aspectual shifts.
i. The movie horrified / enraged Mary for/*in 15 minutes.
ii. The movie filled Mary with horror / awoke rage in Mary in/*for 15 minutes.
(Landau 2010: 137 fn. 2)
comprise a PP might be treated as a periphrastic construction related to

standard psych-verbs, in which the Experiencer is a mental location.
However, the data show that a lexical P with the Experiencer as an object
appears only in one third, i.e. 35 (31%) idiomatic expressions out of the
112 OE psych-idioms, whereas 77 items (69%) include an Experiencer
preceded with no P. The latter might be treated as exhibiting an oblique
Experiencer with a null preposition. However, no relevant syntactic
evidence can be found in support of the claim that there is a covert P in
this type of phrase. Therefore, the results obtained as regards the OE
psych-verbs with explicit PPs do not provide enough evidence in favour of
Landau’s (2010) theory of Experiencers as mental locations, placed either
in a covert or overt PP.
4.6.3 An Experiencer in Baker’s (1989) Thematic Hierarchy

According to Kiparsky’s (1987) and Baker’s (1989) Thematic Hierarchies,
reproduced in (4.37a) and in (4.37b), and repeated in (4.83) and (4.84) for
the sake of convenience, arguments which are part of the fixed portion of
an idiom, are expected to be lower on the thematic hierarchy than
arguments which are not part of the idiom.
(4.83) Thematic Hierarchy Constraint:

Agent > Theme > Goal / Location
(Kiparsky 1987: 35-36)
(4.84) Extended Thematic Hierarchy Constraint:

Agent > Instrument >|Experiencer>|Patient /Theme >
Goal/Location
(Baker 1989: 544)
Following the Hierarchy Constraint, certain arguments may or may not be

part of a verbal idiom, as it is encoded in a thematic hierarchy, i.e. Agent >
Theme > Goal / Location. In fact, all psychological idioms comply with
Baker’s (1989) Thematic Hierarchy, formulated in (4.84), since the
Experiencer (Y) is situated higher in the hierarchy than the Theme (for V-
O idioms, V-DOC idioms, and V-small clause idioms) or Goal (for V-O-
PP idioms). Moreover, scholars, such as Larson (1988), believe that the
hierarchy reflects the order of composition of arguments with the verb.
The argument lowest in the hierarchy is expected to combine with the verb
first, whereas the argument which is highest in the hierarchy should
combine last.
282 Chapter Four
To sum up, the position of the Experiencer in the psychological idioms

under scrutiny corresponds to the position that an Experiencer takes in
typical psych-verbs. The OE (class II and III) psych-verbs with explicit
PPs, found in our databank, do not provide enough evidence in favour of
Landau’s (2010) theory of Experiencers as mental locations. Finally, all
the SE and OE psychological idioms comply with Baker’s (1989)
Thematic Hierarchy.
4.7 Aspectual properties of psychological idioms

This section is not meant to offer a comprehensive analysis of the
aspectual properties of either psych-verbs or idiomatic expressions
corresponding to them. The objective is, however, both to recognise some
existing correlation between psych-verbs and psych-idioms, and to note
some challenging dissimilarities between these two, as regards aspect.
To begin with, verbal predicates related to the psychological and/or
mental /emotional condition of an Experiencer and their aspectual
properties have attracted a great deal of attention in the literature cross-
linguistically. Indeed, the fact that the meaning of psych-verbs is often
ambiguous between states and events, and that they are likely to display
subtle aspectual distinctions, not observed with non-psych verbs, makes
these predicates even more special and challenging (cf. Grimshaw 1990;
van Voorst 1992; Tenny 1994; Pesetsky 1995; Arad 1998; Rozwadowska
2005, 2012; Marín and McNally 2011; and Alexiadou and Iordăchioaia
2014; among many others). They have even been argued to comprise a
special grammatical class with individual characteristics which are not
found with other verb classes, or classified according to core / peripheral
psych properties (cf. Landau 2010).
Therefore, with regards to psychological idioms, the question arises
what aspectual properties this dataset of idioms exhibits. Should the aspect
of psychological idioms be investigated along the diagnostics applied to
usual psych-verbs, or rather should more idiom-bound rules related to
aspectuality be followed? Importantly, in this section of the book, the
stress is to be put more on principles related to idiomaticity of the
psychological expressions under scrutiny than to canonical psych-verbs.
In section 2.4.2 of Chapter Two, some space has been devoted to the
aspectual properties of both SE and OE psych-verbs, and the syntactic
tests to distinguish between stative, eventive agentive, and eventive non-
agentive readings of particular verbs. With this in mind, it is worth
comparing the aspectual characteristics of psych-predicates and their
idiomatic counterparts. Therefore, for the sake of convenience, some
observations related to the aspectuality of psych-verbs, made in (2.23) in

Chapter Two, are repeated here.
In (4.85a)-(4.87a) below, some examples of SE and OE psych-verbs
are displayed, whereas in (4.81b-c)-(4.84b-c) idioms corresponding to
these predicates are tested against the basic syntactic diagnostics, adopted
from Dowty (1979, 1991), Arad (1998, 1999), Rothstein (2004, 2008),
Biaáy (2005), Guidi (2011), Rozwadowska (2012), and Grafmiller (2013),
among others. In all the exemplary sentences Mary is an Experiencer,
while children or dogs represent a Theme.
(4.85) Aspectual properties of SE psych-verbs and idioms corresponding

to them:
a. SE psych-verb “TO LOVE” (stative / eventive reading

(cf. (2.23a) in Chapter Two)
Mary loves children.
b. SE psych-idiom: carry a torch for X (eventive agentive /

non-agentive reading)
Test 1: progressive tenses

Mary is carrying a torch for children.
Test 2: imperatives
Carry a torch for children!
Test 3: an adverb unintentionally

Mary carries a torch for children unintentionally.
Test 4: an adverb on purpose / deliberately / intentionally

Mary carries a torch for children *on purpose.
but Mary carries a torch for children intentionally /
deliberately.65
65
Even though the native speakers’ judgements vary concerning the
grammaticality of this sentence with the adverbs on purpose / deliberately /
intentionally, the most natural in this sentence is the adverb intentionally.
284 Chapter Four
c. SE psych-idiom: have eyes for X (stative reading)

*Mary is having eyes for children.
Test 2: imperatives
*Have eyes for children!

Mary has eyes for children unintentionally.

Mary has eyes for children *on purpose/
*deliberately/
but Mary has eyes for children intentionally.
The SE psych-verb “to love” is stative, but shares some properties with
eventive verbs, e.g. the imperative Love children! sounds good (cf. tests in
(2.23a) in Chapter Two). As shown in (4.85), the predicate “to love” can
be replaced by idioms corresponding to it, which do not always show the
same aspectual property. To be precise, e.g. the idiom carry a torch for X
has a non-stative viz. eventive agentive or eventive non-agentive reading,
since it satisfies the progressive tense test and the imperative test for
eventivity, and it can be used non-agentively (with the adverb
unintentionally) or agentively (with the adverb intentionally), as shown in
(4.85b). In turn, the idiom have eyes for X, fails the tests for eventive
(agentive vs. non-agentive) reading, and satisfies the tests for stativity (it
does not appear in the progressive or the imperative), which confirms its
stative reading. Thus, psychological idioms with an Experiencer Y in the
subject position can retain the same aspectual property (i.e. stative
reading) as the usual psych-verb which corresponds to them. However,
some SE idioms give rise to an eventive (agentive or non-agentive)
reading, in contradistinction to the corresponding psych-verb, as illustrated
in (4.85b).
Furthermore, the sentences in (4.86)-(4.88) below are to check the
aspectual property of psychological idioms, in which the Experiencer
Mary is realised in an object position of class II and III psych-verbs. In
(4.86), the meaning of the idioms is “to depress;” in (4.87) the idioms
mean “to annoy;” while in (4.88) the idioms mean “to matter to.”
(4.86) Aspectual properties of class II OE psych-verbs and idioms

corresponding to them:
a. OE psych-verb “TO DEPRESS”

(stative reading for unaccusative verbs of class II cf. (2.23b
(i)) in Chapter Two)
Dogs depress Mary.
b. OE psych-idiom: give Y a hard time

(eventive agentive / non-agentive reading)

These dogs are giving Mary a hard time.
but Dogs (in general) give Mary a hard time.
Test 2: imperatives
Give Mary a hard time!

Dogs give Mary a hard time unintentionally.

These dogs / children give Mary a hard time on
purpose/ deliberately/ intentionally.
c. OE psych-idiom: break Y’s heart (eventive agentive/non-agentive

reading)

These children are breaking Mary’s heart.
but Small children (in general) break Mary’s heart.
Test 2: imperatives
Break Mary’s heart!

Dogs break Mary’s heart unintentionally.

These dogs / children break Mary’s heart on
286 Chapter Four
(4.87) Aspectual properties of class II OE psych-verbs and idioms

a. OE psych-verb “TO ANNOY” (eventive non-agentive reading for

stative or transitive verbs of class II or eventive agentive for
regular transitive verbs of class II cf. (2.23b (ii)-(iii)) in Chapter
Two)
These children usually annoy Mary.
(non-agentive)
These children annoyed Mary yesterday. (agentive)
b. OE psych-idiom: drive Y bananas


These dogs are driving Mary bananas.
but Dogs (in general) drive Mary bananas.
Test 2: imperatives
Drive Mary bananas!

Dogs drive Mary bananas unintentionally.

These dogs / children drive Mary bananas on
c. OE psych-idiom: bring a hornet’s nest round Y’s ears


These children are bringing a hornet’s nest round
Mary’s ears.
but Small children (in general) bring a hornet’s
nest round Mary’s ears.
Test 2: imperatives
Bring a hornet’s nest round Mary’s ears!

Dogs/children bring a hornet’s nest round Mary’s
ears unintentionally.

These dogs / children bring a hornet’s nest round
Mary’s ears on purpose/ deliberately/ intentionally.
(4.88) Aspectual properties of class III OE psych-verbs and idioms

a. OE psych-verb “TO MATTER”

(stative reading for unaccusative verbs, cf. (2.16c) in Chapter
Two)
Dogs matter to Mary.
b. OE psych-idiom: carry some weight with Y (stative reading)

#These dogs are carrying some weight with Mary.
Test 2: imperatives
# Carry some weight with Mary!

Dogs / children carry some weight with Mary
unintentionally.

These dogs / children carry some weight with Mary *on
purpose/ *deliberately/ *intentionally.
c. OE psych-idiom: make a difference to Y (stative reading)

#These dogs are making a difference to Mary.
Test 2: imperatives
# Make a difference to Mary!
288 Chapter Four

Dogs / children make a difference to Mary
unintentionally.

These dogs / children make a difference to Mary *on
purpose/ *deliberately / *intentionally.
As shown in (4.86)-(4.87), the aspectual ambiguity of class II OE psych-

verb between stative, eventive agentive, and eventive non-agentive reading
is reflected in the aspectual properties of psychological idioms which
correspond to these predicates. Interestingly yet, the idioms give Y a hard
time or break Y’s heart, which relate to the stative class II psych-verb “to
depress,” evoke non-stative, i.e. the eventive (agentive or non-agentive)
reading. Whereas class III psych-verbs (e.g. “to matter to”) and their
idiomatic counterparts (e.g. carry some weight with Y or make a difference
to Y) are stative, as demonstrated in (4.88). Therefore, it is more a matter
of context in a particular sentence in which an idiom occurs, which defines
the aspectual status of the idiom, than the psych-predicate which shares
the same meaning.
Indeed, it was Nunberg (1978), who initially noticed the fact,

developed later by McGinnis (2002), that the aspectual class of an idiom
usually matches that of its non-idiomatic phrasal counterpart. Similarly,
Stone (2016: 51) argues that the observation concerning the analogous
event structure between figurative and literal readings seems to be
confirmed for all verb phrase idioms; however, this sounds unexpected if
idioms are simply recognised as listed lexical entries. An opposing
viewpoint is held by Glasbey (2003; 2007), as is exemplified below (but
cf. Ruwet 1991).
Additionally, Krifka (1992, 1998) and McGinnis (2002) argue that
aspect should be compositional in idioms, in the same way that it is in
literal expressions. Thus, taking that assumption as a starting point,
McGinnis (2002) finds representatives of idioms from all four of
Vendler’s (1957, 1967) aspectual classes. Besides, McGinnis (2002) states
that if a verb occurs with a count noun, a telic idiomatic predicate results.
Likewise, combing the same verb with a mass/ uncountable noun results in
obtaining an atelic idiomatic predicate. McGinnis’ (2002) assumption can
be confirmed by the examples of psychological idioms illustrated in
(4.89a-d), but simultaneously, contradicted by other possible sentences
with the same idioms. The predicates have been subjected to the most
widely used test for distinguishing telic and atelic predicates (the “in-for
test”), since Verkuyl (1972), Dowty (1979), Hinrichs (1985), Levin and
Rappaport (1995), Hay, Kennedy and Levin (1999), among others. It is
said that in-adverbials constitute a test for telicity (Mary painted the wall
in two hours/*for two hours), while for-adverbials for atelicity (Mary
walked for two hours/*in two hours).
(4.89) Psych-idioms confirming vs. contradicting McGinnis’ (2002)

assumption:
a. These children gave the pip to Mary in two hours/*for two

hours.ĺ “to annoy”
These children were giving the pip to Mary (*in two hours
/ for two hours) “to annoy”
(the pip ĺ a count noun and “to annoy”ĺ a telic / an
atelic idiomatic predicate)
b. Mary was raising the roof *in two hours / for two hours.
ĺ “to enjoy”
Mary raised the roof in two hours / *for two hours. ĺ “to
enjoy”
(the roof ĺ a count noun vs. “to enjoy” ĺ a telic / an
atelic idiomatic predicate)
c. Mary was carrying a torch for her children * in two hours /

for all these years. ĺ “to love”)
Mary carried a torch for her children in a long time / *for
all these years. ĺ “to love”)
(a torch ĺ a count noun vs. “to love” ĺ an atelic / a telic
idiomatic predicate)
d. These dogs gave Mary a lot of fright in two hours/*for two

hours. ĺ “to scare”)
These dogs were giving Mary a lot of fright *in two hours
/ for two hours. ĺ “to scare”)
(fright ĺ an uncountable noun vs. “to scare / horrify” ĺ a
telic / an atelic idiomatic predicate)
As shown in (4.89), distinguishing between telic and atelic interpretation

of the psychological idioms is dubious. All these instances of idioms do
not reveal a direct correlation between their (a)telicity and the nominal
290 Chapter Four
(non)count components of these idioms. Therefore, psychological idioms

under scrutiny cast a serious doubt on McGinnis’ (2002) assumption.66
Significantly, Stone (2016), in her recent account, explains that this
correlation between lexical verbal idiom components and idioms as a
whole is justified so long as “the functional elements make the same
semantic contribution, and hence affect event structure composition, in the
same way in both literal and figurative interpretations” (Stone 2016: 51-
52). Undoubtedly, this statement with regard to idiom aspect has
challenged the fundamental understanding of idiomaticity in terms of non-
compositionality. This hypothesis concerning the aspectual contributions
of functional items as being computed equally in literal and figurative
readings also seems to confirm the compositional polysemy view of
treating idioms as compositional units (cf. Stone 2016: 52).
Finally, Glasbey (2003) argues that, even though aspect in idioms
complies in so many cases with their lexical component verbs, there are a
number of interesting exceptions, in which aspect is not derived
compositionally. Indeed, it is Glasbey’s (2003) standpoint which complies
with the results obtained for the psychological idioms, which in so many
cases are aspectually non-compositional. Moreover, Glasbey (2003: 43)
states that many (but not all) of these exceptions fall into a class described
by Jackendoff (1997) as “fake object resultatives,” and exhibit apparent
non-compositionality.67
66
An interested reader is referred to McGinnis’ (2002) detailed analysis in favour
of her claim that “the meaning of idioms is not entirely arbitrary: the structural
component of meaning (specifically, aspect) is both systematic and compositional.
This observation supports the claim of Distributed Morphology that structural
meaning, but not idiosyncratic meaning, is built in the syntax” (McGinnis 2002:
671).
67
Glasbey (2003) suggests regarding these mismatching idioms as aspectually
compositional, provided the aspectual composition is to include Krifka’s (1992)
“thematic relations” as part of its input.
Krifka (1989, 1992) argues that all thematic relations are cumulative in the
sense that if a predicate applies to x and y, it applies to the sum x + y. What is
more, Krifka (1989, 1992) states that telicity is a function of the structure of the
“incremental theme” argument of the verb. Telicity refers to the relation between
the structure of the argument and the described event (cf. Dowty 1991; Filip 1999;
Jackendoff 1996; Tenny 1987, 1994; Verkuyl 1993; Ramchand 1997; and
Rothstein 2004, 2008; among others). For Krifka (1989, 1992), a defining
characteristic of the incremental theme role is that it can satisfy “Mapping to
Objects” and “Mapping to Events” (cf. Punske and Stone 2015).

The behaviour of English idioms has been the focus of much linguistic
research in the past few decades. In the generative linguistic literature in
many studies idioms have been used to support various syntactic analyses,
e.g. syntactic transformations, thematic hierarchy, compositional asymmetry
between subjects and objects (cf. Chomsky 1980; Marantz 1984; Larson
1988; and Kiparsky 1987; among others). Nevertheless, much less linguistic
research has been undertaken into the very structure of idioms (e.g. O’Grady
1998; Bruening 2010). The results of the syntactic examination have
revealed some challenging properties of idioms, which can be
syntactically transformed in various ways with their parts undergoing
modification. For instance, parts of idioms can be passivized, focused,
quantified, omitted, used as antecedents for anaphoric expressions, or
replaced with other homomorphic lexical counterparts.
Since the variation within idioms can be indicative of some kind of
internal structure, it is inevitably expected that there are numerous
constraints which restrict both the size of idioms, their semantic and
discourse productivity, lexical flexibility, and lexical substitution. But the
most essential for our research is the fact that idiomatic constructions are
subject to syntactic constraints. While some linguists argue that various
constraints on idioms are better explained in terms of semantic properties
or several rules applied in the figurative use of language; others point out
to the syntax-semantic interface as the way out to solve the constraint
problem.
Hopefully, the most current research done within the confines of the
Phase Theory (cf. Svenonius 2005; Stone 2009; Harwood 2013, 2016,
2017; Harley and Stone 2013; Kim 2014, 2015; and Corveret al. 2017;
among others) has shed light both on idiomaticity, and their constrained
variability. The particular subset of idioms, namely the psychological
idioms elicited in Chapter Three, have been used in this chapter to
examine the validity and predictions made by the phase-bound approach.
Having provided some insight into the problem of idiom identification,
categorization, (in)flexibility, aspectual properties, and syntactic
constraints proposed for them in the literature so far, it can be concluded
that there are still far more specific issues that should be addressed in
future research, especially concerning the domain of psychological
condition and emotions. One of the specific questions that needs to be
addressed, for sure, refers to the status of English ditransitive idioms
examined in the light of the Phase Theory.
SUMMARY AND CONCLUSIONS
The knowledge of individual words, as stated by CieĞlicka (2010: 149), is

not sufficient to understand and produce a language. Instead, a great
number of formulaic utterances, or fixed expressions, such as:
collocations, phrasal verbs, proverbs, idioms, slogans, common quotations,
or sayings, is required to grasp the very core of the language. Only then
can one’s thoughts and feelings, national features and worldwide truths
embodied in a language be communicated to give it life and richness by
taking the existing words, combining them in a new sense and creating
new meanings, just like in work of art (cf. Lennon 1998).
It seems futile to define and classify formulaic language in general;
likewise, reaching an agreement as to what definition of an idiom to adopt
appears to be impossible. However, despite all potential arrays of idiom
diversity, some basic principles which most of the definitions share have
been outlined:
1. an idiom is recognized as an expression that contains more than one

word, and whose meaning is usually different from the sum of the
literal meanings of its components;
2. some subset of idioms, classified by Nunberg et al. (1994), as
idiomatic phrases (e.g. kick the bucket “to die”), has a fixed word
order, a restricted set of variants and modifications (if any), and should
be treated as comprising single entities since the link between the form
of these idioms and their meaning has not been recognized (cf.
Chomsky 1980; Cruse 1986; Fraser 1970; and Katz 1973; among
others);
3. recently, with a great input of psycholinguistic (cf. Fernando 1996),
and corpus studies (cf. Moon 1998; Grafmiller 2013; among others), a
great number of idioms has been found to be far from being dead or
frozen, but marked, instead, with possible alterations, metaphoricity
and perceptible origins of their meanings. This subset of idioms is
referred to as idiomatically combining expressions (e.g. pull strings “to
use connections”), following the typology offered by Nunberg et al.
(1994).
294 Summary and Conclusions
Moreover, the models of idiomaticity presented in this book have been

based on the results of several novel empirical studies, reported in the first
part of this work. These are:
1. The traditional non-compositional model, typical of generative

grammar (e.g. Swinney and Cutler 1979; Gibbs 1980), with a view of
idioms as compositionally non-derivable by the morpho-syntactic rules
of a language, e.g. the meaning of the canonical idiom kick the bucket
“to die” is not derived from the meanings of its constituents, i.e.
“kick,” “the,” and “bucket” (van der Linden 1992: 223; cf. Katz and
Postal 1963: 275). Idioms are held to be arbitrarily stipulated in
memory (cf. Chomsky 1980; Cruse 1986; Fraser 1970; and Katz 1973;
among others); and treated as “frozen” expressions “not predictable
from the composition” (Machonis 1985: 306), since their structure is
“non-compositional” (Chomsky 1980: 149; Kiparsky 1976: 79).
Consequently, hardly any modification is possible in the case of this
type of “frozen” idioms (called idiomatic phrases (IdP), after Nunberg
et al. 1994).
2. The recent non-traditional compositional model (e.g. Cacciari and
Tabossi 1988; Gibbs, Nayak, and Cutting 1989), which objects to the
standard view of idioms as non-compositional strings. Hence, not all
idioms are “frozen,” and as opaque or unanalysable as kick the bucket.
There are far more idioms, called idiomatically combining expressions
(ICE) (cf. Nunberg et al. 1994), which are syntactically flexible in
various ways, with their parts being modified and semantically
productive (cf. Makkai 1972; Nunberg et al. 1994; O’Grady 1998; and
Ifill 2002; among others). In these compositional idioms some relations
between their overall meaning and form can be established. In fact, the
meanings of particular components of idioms do play a role in the way
idioms are used and understood (cf. Cacciari 1993; Cacciari and
Glucksberg 1991; Gibbs 1990; and Keysar and Bly 1995, 1999).
3. The hybrid approaches with the Model of Dual Idiom Representation
(e.g. Titone and Connine 1999), which try to combine the two previous
accounts of viewing idioms.
For the purpose of the book, the compositional model of idioms, with a
bipolar division of idioms into IdP and ICEs (cf. Nunberg et al. 1994;
Harwood et al. 2016), has been adopted.
Furthermore, the type of idioms that has been chosen as the research
dataset refers to emotional and mental condition of a human participant
(called an Experiencer). Thus, some semantic correspondence between
psychological verbs and psychological idioms has been be established.

Undoubtedly, to reveal one’s emotional and psychological state is a
challenging task, which requires openness and accuracy to name what is
invisible. Nonetheless, an attempt has been made to find both in
dictionaries, the online corpora and via the Google Search any verbal
idiomatic expressions corresponding to the top psych-predicates, also
fitting in the research criteria. As a result, the database of psychological
idioms, obtained in the study, has contained a number of 161 English
verbal idioms. The idioms under scrutiny have fallen into 12 distinct types,
depending on the surface syntactic pattern the core of an idiom has formed
(which is underlined in an idiom pattern) and the position in which an
argument has been placed (e.g. Type (A): V + NP (no open position):
paint the town (red); Type (B): V + NP + preposition + NP (complement
of a P): carry a torch for X; Type (C): V + NP + preposition + NP
(complement of a P): hold X in abomination / contempt; Type (D): V +
possessor + N + preposition + NP (complement of a P): lose Y’s heart to
X; Type (E): V + NP + NP (Double Object Construction): give Y the blues;
and Type (E’): V + a complement small clause: drive Y batty / nuts/
bananas / bonkers). In these patterns of idioms, X (an NP argument)
functions as a Target / a Subject Matter or a Causer, while Y (an argument
corresponding to an Experiencer) is placed either in a subject or an object
position. The arguments also represent open slots, i.e. positions which are
not fixed, but occupied by X or Y (cf. WitkoĞ and Dziemianko 2006).
The idioms under scrutiny have comprised almost always agentive
verbs, which occur necessarily, with an object, followed, or not, by a
Prepositional Phrase. The open position is located either within the object
complement or/and within a PP. The open position can be realised as
either an NP or a possessive modifier. Some idiomatic phrases are opaque
and have all their constituents (NP, PP, or a small clause) fixed, while in
the case of other idioms, their component elements contribute to the
overall meaning of the idiom, which makes the idiomatic meaning more
predictable.
What is more, having analysed the structure of idioms referring to
psychological states in more detail, and having discussed the constraints
on the way they are built, let us now bring together the most relevant
aspects of this analysis.
1. Idiom analysability, non-compositionality, and idiom semantic

decomposition refer to the extent to which idiom constituent parts
contribute to the idiom overall interpretation (cf. Nunberg 1978;
296 Summary and Conclusions
Cacciari and Tabossi 1988; Glucksberg 1991; and Gibbs 1994; among
others).
2. Both the syntactic productivity and the lexical creativity of idioms are
matters of degree, depending on the idiom compositional properties.
The results of the syntactic examination of idioms (on the example of
psychological idioms) have revealed their challenging properties.
These are: the syntactic and lexical variability of ICEs, viz. their
possibility to undergo quantification, topicalization, ellipsis, and
anaphora, among others. As exemplified by psychological idioms,
ICEs, thanks to their more compositional nature, tend to also exhibit a
degree of lexical substitution; thus, parts of idioms may be replaced
with other homomorphic lexical counterparts. IdPs, in turn, are usually
entirely resistant to any such alterations, as summarised in Table 4-1.
from section 4.3 in Chapter Four, repeated here for convenience.
Compositionality
Conventionality
Open Slots
Flexibility
Variation
Syntactic
Opacity
Lexical
Idiom
Non-
IdP High Opaque No No No
compositional
More
ICE Low Trans-parent Yes Yes Yes
compositional
Table 4-1. Properties IdPs and ICEs

(Corver et al. 2017: 10; cf. also Harwood et al. 2016)
3. The most current research performed within the scope of the Phase
Theory (cf. Svenonius 2005; Stone 2009; Harwood 2013, 2016, 2017;
Harley and Stone 2013; Kim 2014, 2015; and Corver et al. 2017;
among others) has shed light both on idiomaticity, and idiom
constrained variability. As a result of the analysis of psychological
idioms in the light of the Phase Theory, it has been noted that, in
addition to CPs and vPs, also DPs, ApplHP, or PrPs may act as phases.
4. The vP-phase imposes an upper bound on verbal IdP idioms. In other
words, IdPs are said to be confined to a single phase, namely the
clause-internal phase;
5. ICEs are able to span multiple phase boundaries and can depend on the
material beyond them. Moreover, nearly all idioms from our dataset
can be characterized as having an open position located, e.g. in the
direct object, small clause subject, or possessor.
Regarding the aspectual properties which psychological idioms

display, the following conclusions have been reached.
1. Following Glasbey’s (2003) stance, and on the basis of the results

obtained in this book for the psychological idioms, we argue that in a
far bigger number of idioms aspect is not derived compositionally.
Indeed, the aspectual status of some psychological idioms corresponds
to the aspect of the lexical component verbs.
2. In this sense, it is difficult to find a direct correlation between the
aspectual properties of psych-idioms and the aspectual structure of
psych-verbs, to which the idioms under scrutiny correspond. From this
perspective and in agreement with Arad’s (1998), and Alexiadou and
Iordachioaia’s (2014) stance, it has been noted that what makes OE
psych-verbs special is their aspectual ambiguity, (between stative,
eventive non-agentive and eventive agentive reading), rather than their
Experiencer argument. Indeed, it is the diversity of aspectual readings,
and more precisely, the non-agentive reading of some OE psych-verbs,
in which these predicates reveal special “psych-effects” (Arad 1999;
and Landau 2010; among others).
In a nutshell, the feature that most psychological idiomatic expressions

share, even though the boundaries are sometimes overextended, is their
metaphorical or figurative nature (Mäntylä 2004: 28-29). What is more,
the ambiguity of many idiomatic expressions has been pointed out, as
some idioms have one interpretation (the literal meaning) derived from the
meanings of the words involved and / or the other (idiomatic meaning). To
conclude, the heterogeneity of these idiomatic expressions indicates the
necessity of learning at least some constraints and irregularities related to
idioms. Having provided some perspectives on the overall picture of idiom
identification, categorization, (in)flexibility, aspectual properties, and
syntactic constraints proposed for them in the literature, the conclusion can
be drawn that there are still far more specific issues that should be
addressed in future research, such as the status of English ditransitive
idioms in the light of the Phase Theory. Finally, so far too little linguistic
research has been done into the very syntactic structure of idioms (e.g.
O’Grady 1998; and Bruening 2010).
APPENDIX 1
The set of all idioms elicited for the 5 top SE psych-verbs under scrutiny
(3.13) love - feel deep affection (13 idioms)
a) carry a torch for X
You carry a torch for a girl you dated in high school.
(COCA)
b) fall head over heels in love with X
He fell head over heels in love with your sister Alice.
(Google)
c) set store by X
Her father had taught her not to set store by icons.
(COCA)
d) think the world of X
We both thought the world of my dad.
(COCA)
e) take (great) delight/ interest / joy / satisfaction in X
God likes us, takes delight in us, in fact.
(COCA)
f) take a fancy / a liking / a shine to X
John began to take a fancy to Sally late last August at the picnic.
(Google)
g) have a soft spot for X
Harry, I've got a soft spot for you, in my heart.
(Google)
h) have a thing about X
I have a thing about Maggie. I guess I'm in love.
(Google)
i) have a weakness for X
John has a weakness for Mary. I think he's in love.
(Google)
j) have eyes for X
“That may be true but I've only got eyes for you, Pretty Girl,” he says.
(Google)
k) go a bundle on X
I don't go a bundle on Anne's new haircut.
(Google)
300 Appendix 1
l) lose Y’s heart to X

James had already lost his heart to the nearest choice, Jane Beaufort.
(Google)
m) set Y’s heart on X
I am sorry you didn't get to pick the one you wanted; I know you had set
your heart on Fred.
(Google)
(3.14) enjoy - receive pleasure or satisfaction from something

(11 idioms)
a) paint the town (red)

After the show, we went out to paint the town red. We'd been sitting at an
outdoor cafe, drinking rum.
(COCA)
b) raise the roof
The whole college is ready to raise the roof at next weekend's homecoming
celebrations.
(Google)
c) have a ball
I'm having a ball interviewing the people.
(COCA)
d) blow / let off (some) steam
To blow off steam after training sessions, we'd go to the terrain park.
(COCA)
e) kick (up) Y’s heels
She was kicking her heels while he slept.
(Google)
f) make the most of X
When you're being raised, you want to make the most of your life.
(COCA)
g) derive/gain/get pleasure from X
He derives pleasure from even the most trivial occupations bringing his
talent into play.
(COCA)
h) take/find pleasure in X
I am taking pleasure in the simple act of her reading me a book.
(COCA)
i) get a buzz out of X
I hope you get a buzz out of Ted. He's a funny guy.
(Google)
j) get a kick / a charge / a bang out of X

This book is just the kind you like and you'll get a real kick out of it.
(Google)
k) reap the benefits of X
God blessed me to be in a position to reap the benefit of the performance.
(COCA)
(3.15) hate - feel strong dislike for or hostility toward (10 idioms)
a) pour scorn on X
He poured scorn on the defendants: And then came this cynical part of the
alleged plot.
(COCA)
b) bear ill will toward X
Do you bear ill will toward Lady Di?
(COCA)
c) bear / owe a grudge against X
Dugliss has borne a grudge against me since that day he lost his first son,
the Fourth of July.
(COCA)
d) bear / feel aversion / malice / hostility / repugnance toward (to) X
But those with the large rifts persist in their slanders and bear malice
towards one another out of anger.
(Google)
e) have no stomach / taste / use for X
Pat has no stomach for violent movies.
(Google)
f) show contempt for X
Art critics for years ignored or showed contempt for De, especially for
political reasons.
(COCA)
g) hold X in abomination / contempt
His created beings hold him in contempt and despise him.
(Google)
h) turn Y’s back on X
But I don't think now is the time to turn your back on each other. So,
therefore, I support him.
(COCA)
i) turn Y’s nose up at X
John turned his nose up at Ann, and that hurt her feelings.
(Google)
302 Appendix 1
j) look down (Y’s) nose at X

Seniors have a way of looking down their noses at juniors.
(Google)
(3.16) fear - feel apprehensive, afraid or frightened of something /

someone (9 idioms)
a) give / raise an alarm

Next you will raise an alarm that we are threatened by creatures of human
size.
(COCA)
b) show the white feather
In the way of destruction, I do not turn tail or show the white feather at
calamity's first notice. I try to stand up and put my face.
(COCA)
c) get cold feet
California's San Bernardino County got cold feet when representatives of
the securities and banking industries sent threatening letters.
(COCA)
d) have goose bumps
He was leaving. We all had goose bumps. I mean, I saw exultation in the
cast and crew.
(COCA)
e) have / get pins and needles
I had pins and needles until I got my job.
(Google)
f) turn tail (and run)
The pair turned tail and ran deeper into the mine, the Germans close on
their heels. (COCA)
g) take dim view of X
They take dim view of those they consider outsiders. The public, however,
seems willing.
(COCA)
h) have a yellow streak / belly down X’s back
Tex has a yellow streak down his back a mile wide. He's afraid to cross the
street! (Google)
i) have Y’s heart in Y’s mouth
I had my heart in my mouth when I waited for the news of my son.
(Google)
(3.17) worry – to feel uneasy or concerned about something; to fret or be

anxious about the welfare of someone or something (SE)
(6 idioms)
a) have the blues
After seeing the old house in such bad shape, I had the blues for weeks
(Google)
b) eat Y’s heart out
She is still eating her heart out over being fired.
(Google)
c) lose sleep over X
Akiko Murakami, meanwhile, is losing sleep over the worst-case
scenario.
(COCA)
d) have a bee in Y’s bonnet
My colleague Dan Buirski had a bee in his bonnet about eating
disorders.
(Google)
e) have ants in Y’s pants
I have ants in my pants. I'm a jitterbug, a worrywart, a wiggle worm.
(COCA)
f) have butterflies in Y’s stomach
Her mouth was dry, there were butterflies in her stomach, and her
knees were shaking so much it was hard to walk on stage.
(Google)
APPENDIX 2
The set of all idioms elicited for the 8 top OE (class II) psych-verbs under
scrutiny, illustrated with the help of sentences extracted from the COCA /
via the Google Search.
(3.18) annoy - make (someone) a little angry; irritate, upset, irritate,

aggravate
(26 idioms)
a) play (a game of) cat and mouse
Enemy warplanes have been playing a deadly game of cat and mouse,
trying to bring American fighter planes into range of their missiles.
(Google)
b) get the hump
Tony got the hump because he thought we hadn’t invited him to the party.
(Google)
c) get Y’s goat
Do you ever hear annoying phrases that get your goat?
(COCA)
d) raise Y’s hackles
Jim could raise her hackles quickly, but she enjoyed being with him
anyway.
(Google)
e) rattle Y’s cage / chain
I rattled his cage by telling him I hated his art.
(Google)
f) try Y’s patience / try the patience of Y
People were always disappointing her, or trying her patience, or boring
her into a state of stupefaction.
(COCA)
g) ruffle Y’s feathers
I didn't mean to ruffle his feathers. I just thought that I would remind him
of what he promised us.
(Google)
h) get Y’s dander /hacklers/ Irish up
I insulted him and really got his hackles up.
(Google)
306 Appendix 2
i) put/get Y’s back up

She put my back up immediately by interrupting everything I said.
(Google)
j) get a rise out of Y
Ignore him - he's just trying to get a rise out of you.
(Google)
k) take it out of Y
This construction job really takes it out of me.
(Google)
l) stir up a hornet’s nest (of something) amongst/ in Y
An unfortunate remark made by a member stirred up a hornet’s nest in the
Club.
(Google)
He certainly stirred up a hornet's nest of angst amongst the Welsh-
speaking fraternity.
(Google)
m) upset the apple cart against Y
Dundee will be hoping they can upset the apple cart against Rangers.
(Google)
n) bring a hornet’s nest round Y’s ears
However, the chief offenders for the time were flogged and kept in bounds;
but the victorious party had brought a nice hornet's nest about their
ears.
(Google)
o) cut Y to the quick / bone
Your heartless comments cut me to the quick.
(Google)
p) drive Y up the wall
These slovenly workmen drive me up the wall.
(Google)
q) put Y in a hole /a bind
I campaigned against two opponents from this state, and I think it put me
in a hole.
(Google)
r) put Y on the spot
She asked if I would vote for her, which really put me on the spot because
I had decided not to.
(Google)
s) put Y out of countenance

They perfectly put her out of countenance by staring at her through their
spying glasses.
(Google)
t) put Y through wringer
They really put me through the wringer in my interview.
(Google)
u) throw/send Y into a tizzy / tizz
The idea of producing a meal for fifty people threw her into a tizzy.
(Google)
v) put Y’s nose out of joint
Don't disappoint Little Mabel; though he has put her nose out of joint.
(Google)
w) thumb X’s nose at Y
Walter thumbed his nose at Fred and asked the gang to send someone else
to do the job.
(Google)
x) give Y the pip
That sort of talk gave Jimmy the pip.
(Google)
y) rub Y (up) the wrong way
One can see how [his] expression of his ideals and intentions must have
rubbed many people the wrong way.
(Google)
z) drive Y batty/nuts/ bananas / bonkers / crazy
You’re certainly annoying! You’re going to drive me batty!
(Google)
(3.19) fascinate – attract the strong attention and interest of (someone)
(11 idioms)
a) catch Y’s eye / catch the eye of Y

We need a new advertising campaign that really catches the eye of
younger demographics. (Google)
b) win Y’s heart / win the heart of Y
Good thing I had probably won her heart by tumbling backward over my
chair at our first meeting.
(COCA)
c) turn Y’s head
Success has never turned her head – she’s still the same simple unaffected
girl she always was.
(Google)
308 Appendix 2
d) tickle Y’s fancy

This doesn't tickle my fancy at all. This is dull and boring.
(Google)
e) stir the /Y’s blood
The Olympics as a concept, as a package, doesn’t stir my blood, and I
don't greatly care as such whether Australians win things or not.
(Google)
f) cast a spell on Y
She is a real beauty. She cast a spell on every man she met.
(Google)
g) make a hit with Y
The dessert you served really made a hit with the guests.
(Google)
h) pay court to Y
The lobbyist paid court to all the influential members of Congress.
(Google)
i) put Y in a trance
He put me in a trance with those brown eyes.
(Google)
j) set Y on fire
Ted’s presentation didn't exactly set me on fire, but it was a good summary
of the project.
(Google)
k) sweep Y off Y’s feet
She was hoping that some glamorous young Frenchman would come
along and sweep her off her feet.
(Google)
(3.20) amuse – cause (someone) to find something funny; provide

interesting and enjoyable occupation for (someone); entertain
(4 idioms)
a) gladden Y’s hearts / the heart of Y

Wine gladdens human hearts. (Google)
b) make the grade for Y
I find your number doesn’t make the grade for most women.
(Google)
c) play the fool for Y
I still play the fool for them, and from time to time I am called upon to
amuse them.
(COCA)
d) tickle Y to death / to pieces / pink

What you told her just tickled her to death!
That story just tickles me to pieces. Bill told a joke that really tickled us all
pink.
(Google)
(3.21) scare – cause great fear or nervousness in; frighten

(11 idioms)
a) curdle / chill Y’s blood
The terrible scream was enough to curdle my blood.
(Google)
b) frighten / scare the life / the hell out of Y
She frightened the life out of me, shouting like that.
(Google)
c) put the fear of God into Y
They'd put the fear of God into the most hardened criminal.
(COCA)
d) put/get the wind up Y
Say you’ll take him to court if he doesn’t pay up – that should put the wind
up him.
(Google)
e) strike terror into Y
These are words that strike terror into the hearts of artisan bakers and
advocates.
(COCA)
f) chill Y to the bone/marrow

The sound of scraping at the window chilled me to the bone.
(Google)
g) send Y into a cold sweat
With the discovery came a sudden nervous reaction that sent him into a
cold sweat.
(Google)
h) throw /send Y into a panic
The lack of knowledge threw her into a panic.
(Google)
i) shake (a / X’s) fist at Y
I shake a fist at my childhood friend in mock anger.
(COCA)
j) give Y goose pimples/ bumps
That truth seemed to creep beneath her clothes, give her goose pimples.
(Google)
310 Appendix 2
k) give Y (quite) a (bit of) turn / a fright

I have to tell you, it gave me a turn. I saw a man die
. (COCA)
(3.22) depress – make (someone) feel utterly dispirited or dejected

(20 idioms)
a) upset the applecart

I don't want to upset the applecart now by asking you to change the date
for the meeting.
(Google)
b) break Y’s heart / spirit
The children looked so desperately sad it broke my heart to see them
sweltering.
(COCA)
c) damp Y’s hopes
Notwithstanding your endeavours, too, to damp my hopes, I comfort
myself with expectations of their promised return.
(Google)
d) dampen/damp Y’s spirits
She was in the hospital for a while, and that was just dampening my spirit.
(COCA)
e) dash / wither Y’s hopes
Mary dashed my hopes when she said she wouldn’t marry me.
(Google)
f) cast a gloom / a shadow over Y
For some time the presence of the Peer seemed destined to cast a gloom
over the society.
(Google)
g) do a number on Y
The editor did a number on me because of my term paper.
(Google)
h) knock the stuffing out of Y
It was their third defeat in a row and it really knocked the stuffing out of
them.
(Google)
i) put a damper on Y
We both wanted kids, but it didn't happen for us, which put a damper on
the marriage.
(COCA)
j) bring tears to Y’s eyes

The mention of my child's name may bring tears to my eyes, but it never
fails to bring music to my ears.
(Google)
k) take the wind out of Y’s sails
She took the wind out of his sails by turning down his offer to marry him.
(Google)
l) cut Y down to size
He started he thought he knew everything, but we soon cut him down to
size.
(Google)
m) take/ knock Y down a peg / notch (or two)
The teacher’s scolding took Bob down a notch or two.
(Google)
n) put / send / throw Y into a (blue) funk
Having to change her menu threw the whole day off schedule and put her
into a blue funk.
(Google)
o) leave Y in the / a lurch

I hope they can find someone to replace me at work. I don’t want to leave
them in the lurch.
(Google)
p) bring Y into disrepute
My bankruptcy brought me into disrepute.
(Google)
q) knock Y for six
It really knocked me for six when my ex-boyfriend announced he was
getting married.
(Google)
r) give Y a bad /hard time/ the blues / the run around
The coach really gave me a bad time when I missed that catch last night.
(Google)
The rain sometimes gives me the blues.
(COCA)
s) bring Y low / to naught
Dave was boasting so much about getting an A on the test that I finally
had to bring him low and tell him that Miss Hayes is a really easy
grader.
(Google)
312 Appendix 2
t) knock Y sideways
The news of her brother’s death knocked her sideways.
(Google)
(3.23) anger - fill (someone) with anger; provoke anger in

(10 idioms)
a) fan the fire / flame (of something)

She already found him attractive, but his letters really fanned the flames.
(Google)
b) kindle Y’s wrath
This insufferable mutability of the king's temper kindles my wrath.
(Google)
c) get Y’s back / dander up
Bill’s arrogance really got my back up.
(Google)
d) put / set Y’s back (up)
The foolish delays at the bank only put her back up.
(Google)
e) blow a gasket on Y
We blew a gasket on our French press – we debated the day’s itinerary.
(COCA)
f) breathe fire over Y
The bishop was breathing fire over the press release made a few days ago.
(Google)
g) drive Y out of mind
You are driving me out of my mind with your nagging. (Google)
h) drive Y to distraction
The problems I am having with my boss are driving me to distraction.
(Google)
i) drive Y round the bend / twist
You're driving me round the bend with your constant complaining.
(Google)
j) vent X’s spleen at Y
Jack vented his spleen at his wife whenever things went badly at work .
(Google)
(3.24) surprise – cause (someone) to feel mild astonishment or shock.

(15 idioms)
a) take Y’s breath away

The first time I saw Tyler’s studio operation, it took my breath away.
(COCA)
b) knock Y’s socks off
“Your story knocked my socks off,” I said, not adding that ever since I’d
heard it.
(COCA)
c) knock Y down / over with a feather
It really knocked me down with a feather to hear that my ex-wife had
already remarried. (Google)
d) blow Y out of the water
Her singing blew me out of the water – I haven’t heard anyone sing like
that since Sarah Vaughn.
(Google)
e) catch Y off balance / up short / napping
I didn’t expect you so soon. You caught me off balance.
The thieves caught the security guard napping.
(Google)
f) catch Y on the hop
I’m afraid you’ve caught me on the hop – I wasn’t expecting your call
until this afternoon.
(Google)
g) throw Y off balance
Your last question sort of threw me off balance.
(Google)
h) strike Y with awe
Being treated by persons who were so much my superiors struck me with
awe.
(Google)
i) strike Y with wonder
Lord’s great love struck me with wonder and astonishment at His free and
undeserved love.
(Google)
j) catch Y off (Y’s) guards
He’s used to being interviewed and it seems that no question catches him
off guard.
(Google)
314 Appendix 2
k) throw Y a curve (ball)

John threw me a curve when we were making our presentation, and I
forgot my speech.
(Google)
l) catch Y unawares
The question caught me unaware as I stood there on the corner in a daze
searching for some sort of an answer.
(Google)
m) leave Y open-mouthed
Such courage left me open mouthed in wonder.
(Google)
n) leave Y at a loss for words
Father’s tirade left us all at a loss for words.
(Google)
o) strike Y dumb
And all we learned about the Fair filled my heart to overflowing and
struck me dumb with dread.
(COCA)
(3.25) horrify – fill with horror; shock greatly

(8 idioms)
a) freeze Y’s (the) blood

Believe me, every meal at their table freezes my blood.
(COCA)
b) put the screws on Y
They put the screws on him until eventually he was forced to resign.
(Google)
c) scare the shit / the wits out of Y
The door blew shut and scared the shit out of me.
(Google)
d) frighten / scare the pants off Y
The piano lid fell and scared the pants off my parents.
(Google)
e) scare the bejesus out of Y
I look at what she does and it scares the bejesus out of me.
(Google)
f) frighten / scare Y to death
The thought of standing in front of a huge crowd frightens me to death.
(COCA)
g) frighten / scare Y out of their wits

Don’t sneak up behind me like that – you frightened me out of my wits!
(Google)
h) give Y a turn
You gave her quite a turn, suddenly appearing like that!
(Google)
APPENDIX 3
The set of all idioms, exemplified with the sentences taken from the
COCA / via the Google Search elicited for the 2 top OE (class III) psych-
verbs under scrutiny.
(3.26). appeal – be attractive or interesting

(4 idioms)
a) float Y’s boat

I think this new job in the lab will really float Isabel’s boat.
(Google)
b) tickle Y’s fancy
I have an interesting problem here that I think will tickle your fancy.
(Google)
c) whet Y’s appetite

Her work on this show has whetted her appetite to do theatrical sets for
other shows.
(Google)
d) set / put Y on Y’s ear
The presence of the movie star set the whole town on its ear.
(Google)
(3.27) matter to – be important or significant for someone

(3 idioms)
a) carry some / a lot of weight with Y

Your argument does not carry a lot of weight with me.
(Google)
b) cut (no) ice with Y
The party's stance on taxation no longer cuts ice with voters.
(Google)
c) make a difference to Y
Health and safety over the past decade had made a difference to their
business.
(COCA)
REFERENCES
Aarts, Bas. 1992. Small Clauses in English: The Nonverbal Types. Berlin,
New York: Mouton de Gruyter.
Abel, Beate. 2003. English idioms in the first language and second
language lexicon: A dual representation approach. Second Language
Research 19, 329-358.
Abeillé, Anne. 1995. The flexibility of French Idioms: A representation
with lexicalized tree adjoining grammar. In Everaert, M.; and van der
Linden, E.-J.; Schenk, A.; and Schreuder, R. (eds), Idioms: Structural
and Psychological Perspectives, 15-42. Hillsdale, NJ: Lawrence
Erlbaum Associates.
Abney, Steven P. 1987. The English Noun Phrase in its Sentential Aspect.
PhD Dissertation. Cambridge, MA: MIT Press.
Abraham, Werner. 2010. Misleading Homonymies, Economical PPs in
Microvariation, and P as a Probe. In Cinque, G.; and Rizzi, L. (eds.),
Mapping Spatial PPs, 261-294. New York: Oxford University Press.
Adger, David. 2003. Core Syntax: A Minimalist Approach. Oxford:
Oxford University. Press.
Adkins, Patricia G. 1968. Teaching Idioms and Figures of Speech to
Non-Native Speakers of English. The Modern Language Journal 52
(3), 148-152.
Aelbrecht, Lobke; and den Dikken, Marcel. 2013. Preposition doubling
in Flemish and its implications for the syntax of Dutch PPs. The
Journal of Comparative Germanic Linguistics 16, 33-68.
Aisenstadt, Esther. 1979. Collocability restrictions in dictionaries. ITL:
Review of Applied Linguistics 45 (6), 71-74.
Alexander, Robin J. 1987. Problems in understanding and teaching
idiomaticity in English. Anglistik und Englischunterricht 32, 105-122.
Alexiadou, Artemis. 2003. Some notes on the structure of alienable and
inalienable possessors. In Coene, M.; and D’hulst, Y. (eds.), From NP
to DP. Volume 2: The Expression of Possession in Noun Phrases, 167-
188. Amsterdam: John Benjamins Publishing Company.
Alexiadou, Artemis; Anagnostopoulou, Elena; and Everaert, Martin
(eds.). 2004. The Unaccusativity Puzzle: Explorations of the Syntax-
Lexicon Interface. Oxford: Oxford University Press.
320 References
Alexiadou, Artemis; Haegeman, Liliane M. V.; and Stavrou, Melita.

2007. Noun Phrase in the Generative Perspective. Berlin, New York:
Mouton de Gruyter.
Alexiadou, Artemis; and Iordăchioaia, Gianina. 2014. The psych
causative alternation. Lingua 148, 53-79.
Anagnostopoulou, Elena. 1999. On Experiencers. In Alexiadou, A.; and
Horrocks, G. C.; and Stavrou, M. (eds.), Studies in Greek Syntax, 67-
93. Dordrecht: Kluwer.
Anagnostopoulou, Elena. 2008. Psych-verbs. Handout from Athens
reading group in linguistics, April 11.
Anastasiou, Dimitra. 2010. Idiom Treatment Experiments in Machine
Translation. PhD Dissertation. Luxembourg: Luxembourg Institute of
Science and Technology.
Anudo, Cellyne Nelly; and Kodak, Benard. 2017. A Conceptual
Analysis of Love Metaphors in Selected Dholuo Popular Music.
International Journal of Innovative Research and Development 6 (3),
168-176.
Arad, Maya. 1996. A minimalist view of the syntax-lexical semantics
interface. UCL Working Papers. Linguistics 8, 215-242.
Arad, Maya. 1998. Psych-notes. UCL Working Papers in Linguistics 10.
London: University College.
Arad, Maya. 1999. What Counts as a Class? The Case of Psych Verbs.
MIT Working Papers. Linguistics 35, 1-23.
Arad, Maya. 2000. Psych-verbs and the Syntax-Lexicon Interface. Ms.,
Geneva: University of Geneva.
Arad, Maya. 2002. Universal features and language-particular
morphemes. In Alexiadou, A. (ed.), Theoretical Approaches to
Universals, 15-39. Amsterdam: John Benjamins Publishing Company.
Bach, Emmon. 1981. On time, tense, and aspect: An essay in English
metaphysics. In Cole, P. (ed.), Radical Pragmatics, 63-81. New York:
Academic Press.
Bach, Emmon. 1986. The Algebra of Events. Linguistics and Philosophy
9, 5-16.
Baker, Mark C. 1988. Incorporation: A Theory of Grammatical Function
Changing. Chicago, IL: University of Chicago Press.
Baker, Mark C. 1989. Object Sharing and Projection in Serial Verb
Constructions. Linguistic Inquiry 20, 513-553.
Baker, Mark C. 1997. Thematic roles and syntactic structure. In
Haegeman, L. (ed.), Elements of Grammar: Handbook in Generative
Syntax, 73-137. Dordrecht: Kluwer.
Becker, Joseph D. 1975. The phrasal lexicon. In Webber, B.; and Schank,
R. (eds.), Theoretical Issues in Natural Language Processing, 70-73.
Cambridge, MA: Association for Computational Linguistics.
Beeman, Mark. 1998. Coarse semantic coding and discourse comprehension.
In Beeman, M.; and Chiarello, C. (eds.), Right Hemisphere Language
Comprehension: Perspectives From Cognitive Neuroscience, 255-284.
Mahwah, NJ: Erlbaum.
Benjafield, John G. 1992. Cognition. Englewood Cliffs, New Jersey:
Prentice-Hall.
Belletti, Adriana; and Rizzi, Luigi. 1988. Psych-verbs and ș theory.
Natural Language and Linguistic Theory 6, 291-352.
Bennis, Hans. 2000. Adjectives and argument structure. In Coopmans, P.;
Everaert, M; and Grimshaw, J. (eds.), Lexical Specification and
Insertion, 27-67. Amsterdam: John Benjamins Publishing Company.
Berman, Arlene. 1973. A constraint on tough-movement. In Corum, C.;
Smith-Stark, T. C.; and Weiser, A. (eds.), Papers from the Ninth
Regional Meeting of the Chicago Linguistic Society, 34-43. Chicago,
IL: University of Chicago.
Bever, Thomas G.; Lackner, James R.; and Kirk, R. 1969. The
underlying structures of sentences are the primary units of immediate
speech processing. Perception and Psychophysics 5 (4), 225-234.
Biaáy, Adam. 2005. Polish Psychological Verbs at the Lexicon-Syntax
Interface in Cross-linguistic Perspective. Frankfurt am Main: Peter
Lang.
Bobrow, Samuel A.; and Bell, Susan M. 1973. On catching on to
idiomatic expressions. Memory and Cognition 1 (3), 342-346.
Bos, Johan; and Spenader, Jeniffer. 2011. An annotated corpus for the
analysis of VP ellipsis. Language Resource and Evaluation 45 (4),
463-494.
Boškoviü, Željko. 2005. On the locality of left branch extraction and the
structure of NP. Studia Linguistica 59, 1-45.
Boškoviü, Željko. 2011. On Unvalued Interpretable Features. In Lima, S.;
Mullin, K.; and Smith, B. (eds.), NELS 39: Proceedings of the 39th
Annual Meeting of the North East Linguistic Society, 109-120.
Amherst, MA: GLSA.
Boškoviü, Željko; and Lasnik, Howard. 2006. Minimalist Syntax: The
Essential Readings. Oxford: Blackwell.
Botelho da Silva, Teresa; and Cutler, Anne. 1993. Ill-formedness and
transformability in Portuguese idioms. In Cacciari, C.; and Tabossi, P.
(eds.), Idioms: Processing, Structure, and Interpretation, 129-143.
Hillsdale, NJ: Erlbaum.
322 References
Bouchard, Denis. 1992. Psych constructions and linking to conceptual

structures. In Hirschbühler, P.; and Koerner, K. (eds.), Romance
Languages and Modern Linguistic Theory, 25-44. Amsterdam: John
Benjamins Publishing Company.
Bouchard, Denis. 1995. The Semantics of Syntax: A Minimalist Approach
to Grammar. Chicago, IL: University of Chicago Press.
Bowers, John. 1993. The Syntax of Predication. Linguistic Inquiry 24,
591-656.
Bowers, John. 2001. Predication. In Baltin, M.; and Collins, C. (eds.), The
Handbook of Contemporary Syntactic Theory, 299-333. Oxford:
Blackwell.
Bowers, John. 2002. Transitivity. Linguistic Inquiry 33, 183-224.
Brannon, Linda L. 1975. On the Understanding of Idiomatic
Expressions. PhD Dissertation. Texas: University of Texas.
Bresnan, Joan. 1982. The Passive in Lexical Theory. In Bresnan, J. (ed.),
The Mental Representation of Grammatical Relations, 3-86. MIT Press
Series on Cognitive Theory and Mental Representation. Cambridge,
MA; London: The MIT Press.
Bruening, Benjamin. 2010. Ditransitive asymmetries and a theory of
idiom formation. Linguistic Inquiry 41, 519-562.
Bruening, Benjamin. 2016. Depictive Secondary Predicates and Small
Clause Approaches to Argument Structure. Ms., Delaware: University
of Delaware.
Burt, Jennifer S. 1992. Against the lexical representation of idioms.
Canadian Journal of Psychology 46, 582-605.
Burzio, Luigi. 1986. Italian Syntax: a Government-Binding Approach.
Dordrecht: Kluwer.
Burzio, Luigi. 1989. On the non-existence of disjoint reference principles.
Rivista di Grammatica Generativa 14, 3-27.
Bybee, Joan. 2006. From usage to grammar: the mind’s response to
repetition. Language 82 (4), 529-551.
Cacciari, Cristina. 1993. The place of idioms in a literal and
metaphorical world. In Cacciari, C.; and Tabossi, P. (eds.) Idioms:
Processing, Structure and Interpretation, 27-55. Hillsdale, NJ:
Lawrence Erlbaum Associates.
Cacciari, Cristina; and Glucksberg, Sam. 1991. Understanding
idiomatic expressions: The contribution of word meanings. In
Simpson, G. B. (ed.), Understanding Word and Sentence, 217-240.
Amsterdam: Elsevier.
Cacciari, Cristina; and Tabossi, Patrizia. 1988. The comprehension of
idioms. Journal of Memory and Language 27, 668-683.
Calzolari, Nicoletta; Fillmore, Charles; Grishman, Ralph; Ide, Nancy;

Lenci, Alessandro; Macleod, Catherine; and Zampolli, Antonio.
2002. Towards best practice for multiword expressions in
computational lexicons. In Proceedings of the Third International
Conference on Language Resources and Evaluation (LREC 2002),
1934-1940. Las Palmas, Canary Islands.
Carine, Marret. 2005. Idioms. In Strazny, P. (ed.) Encyclopedia of
Linguistics. New York: Fitzory Dearborn. Retrieved from
http://www.gigapedia.com
Carlson, Gregory N. 1977. Reference to Kinds in English. PhD
Dissertation. Amherst, MA: University of Massachussetts.
Carlson, Gregory N.; and Tanenhaus, Michael K. 1988. Thematic
Roles and Language Comprehension. In Wilkins, W. (ed.), Syntax and
Semantics, Vol. 21: Thematic Relations, 263-289. New York:
Academic Press.
Carlson, Lauri. 1981. Aspect and quantification. Syntax and Semantics
14, 31-64.
Carter, Richard J. 1976. Some constraints on possible words. Semantikos
1, 27-66.
Cedar, Payung S. 2004. Transferability and Translatability of Idioms by
Thai-Speaking Learners of English. PhD Dissertation. Boston: Boston
University.
Chafe, Wallace. L. 1968. Idiomaticity as an anomaly in the Chomskyan
Paradigm. Foundations of Language 4, 109-125.
Chomsky, Noam. 1957. Syntactic Structures. ’s-Gravenhage: Mouton.
Chomsky, Noam. 1965. Aspects of the Theory of Syntax. Cambridge, MA:
The MIT Press.
Chomsky, Noam. 1970. Remarks on nominalization. In Jacobs, R. A.; and
Rosenbaum, P. S. (eds.), Readings in English Transformational
Grammar, 184-221. Waltham, MA: Ginn.
Chomsky, Noam. 1980. Rules and Representations. New York: Columbia
University Press.
Chomsky, Noam. 1981. Lectures on Government and Binding.
Hawthorne, Westchester, New York: Foris Publications
Chomsky, Noam. 1982. Some Concepts and Consequences of the Theory
of Government and Binding. Cambridge, MA: MIT Press.
Chomsky, Noam. 1993. A minimalist program for linguistic theory. In
Hale, K.; and Keyser, S. J. (eds.), The View from Building 20: Essays
in Linguistics in Honor of Sylvain Bromberger, 1-52. Cambridge, MA:
MIT Press [reprinted in Chomsky, Noam. 1995. The Minimalist
Program. Cambridge, MA: MIT Press].
324 References
Chomsky, Noam. 1998, 2000. Minimalist inquiries. In Martin, R.;

Michaels, D.; and Uriagereka, J. (eds.), Step by Step: Essays on
Minimalist Syntax in Honor of Howard Lasnik, 89-155. Cambridge,
MA: MIT Press.
Chomsky, Noam. 2001. Derivation by phase. In Kenstowicz, M. (ed.),
Ken Hale: A life in Language, 1-52. Cambridge, MA: MIT Press.
Chomsky, Noam. 2005. On phases. In Freidin, R. et al. (eds.),
Foundational Issues in Linguistic Theory: Essays in Honor of Jean
Roger Vergnaud, 133-166. Cambridge, MA.: MIT Press.
Chomsky, Noam. 2007. Approaching UG from Below. In Gartner, H.-M.;
and Sauerland, U. (eds.), Interfaces + Recursion = Language?
Chomsky’s Minimalism and the View from Syntax-Semantics, SGG89,
1-30. Berlin: Mouton de Gruyter.
Chomsky, Noam. 2008. On phases. Current Studies in Linguistics Series
45, 133-166.
Chomsky, Noam. 2013. Problems of projection. Lingua 130, 33-49.
CieĞlicka, Anna. 2004. On Processing Figurative Language: Towards a
Model of Idiom Comprehension in Foreign Language Learners.
PoznaĔ: Motivex.
CieĞlicka, Anna. 2006. Literal salience in on-line processing of idiomatic
expressions by second language learners. Second Language Research
22, 115-144.
CieĞlicka, Anna B. 2010. Formulaic language in L2: Storage, retrieval
and production of idioms by second language learners. In Pütz, M.; and
Sicola, L. (eds.), Cognitive Processing in Second Language
Acquisition: Inside the Learner's Mind, 149-168. Amsterdam /
Philadelphia: John Benjamins Publishing Company.
CieĞlicka, Anna B. 2011. Suppression of literal meanings in L2 idiom
processing: does context help? Studies in Second Language Learning
and Teaching 1 (1), 13-36.
CieĞlicka, Anna B. 2013. Do non-native language speakers chew the fat
and spill the beans with different brain hemispheres?: Investigating
idiom decomposability with the divided visual field paradigm. Journal
of Psycholinguistic Research 42 (6), 475-503.
CieĞlicka, Anna B.; and Heredia, Roberto R. 2011. Hemispheric
asymmetries in processing L1 and L2 idioms: effects of salience and
context. Brain and Language 116, 136-150.
CieĞlicka, Anna B.; Heredia, Roberto R.; and Olivares, Marc. 2014.
It’s all in the eyes: how language dominance, salience, and context
affect eye movements during idiomatic language processing. In
Pawlak, M.; and Aronin, L. (eds.), Essential Topics in Applied
Linguistics and Multilingualism. Studies in Honor of David Singleton,

21-41. New York: Springer.
CieĞlicka, Anna B.; O’Rourke, Breffini; and Singleton, David. 2009.
Salience or context in the processing of L1 and L2 multi-word units?
In Wysocka, M. (ed.), On Language Structure, Acquisition and
Teaching, 293-304. Katowice: Wydawnictwo Uniwersytetu ĝląskiego.
CieĞlicka, Anna B.; Rataj, Karolina; Maáaszuk, Justyna; and
Jaworska, Dorota. 2008. Figurative Language Impairment in Aphasic
Patients: the Effects of Task Type and the Type of Figurative Trope.
PoznaĔ: Adam Mickiewicz University.
Cinque, Guglielmo. 1990. Types of A-bar Dependencies. Cambridge,
MA: MIT Press.
Citko, Barbara. 2008. Small Clauses Reconsidered: Not So Small and
Not All Alike. Lingua 118, 261-295.
Citko, Barbara. 2011. Small Clauses. Language and Linguistics Compass
5, 748-763.
Citko, Barbara. 2014. Phase Theory: An Introduction. Cambridge:
Cambridge University Press.
Clore, Gerald L.; Ortony, Andrew; and Foss, Mark A. 1987. The
psychological foundations of the affective lexicon. Journal of
Personality and Social Psychology 53 (4), 751-766.
Colin, Nathalie. 2005. English and Swedish Animal Idioms - A Study of
Correspondence and Variation in Content and Expression. PhD
Dissertation. Engelska: C-uppsats.
Collinge, Neville Edgar. 1990. An Encyclopedia of Language. London,
Angleterre: Routledge.
Collins, Peter C. 1991. Cleft and Pseudo-Cleft Constructions in English.
London: Routledge.
Condoravdi, Cleo. 1992. Individual level predicates in conditional
clauses. Talk given at the LSA Annual Meeting, Philadelphia, PA.
Corver, Norbert; Temmerman, Tanja; Leufkens, Sterre; van
Cranenbroeck, Jeroen; and Hladnik, Marko. 2017. Idioms:
phasehood and compositionality. Ms., Utrecht: Utrecht University.
Cowie, Anthony P. 1981. The treatment of collocations and idioms in
learners’ dictionaries. Applied Linguistics 2 (3), 223-235.
Cowie, Anthony P. 1998a. Introduction. In Cowie, A. P., Phraseology:
Theory, Analysis, and Applications, 1-20. Oxford: Clarendon Press.
Cowie, Anthony P. 1998b. Phraseology: Theory, Analysis, and
Applications. Oxford: Clarendon Press.
326 References
Cowie, Anthony P.; and Mackin, Ronald. 1975. Oxford Dictionary of

Current Idiomatic English, Volume 1: Verbs with Prepositions and
Particles. London: Longman.
Cowie, Anthony P.; Mackin, Ronald; and McCaig, Isabel R. 1983.
Oxford Dictionary of Current Idiomatic English: Phrase, Clause and
Sentence Idioms (Vol. 2). Oxford: Oxford University Press.
Cresti, Diana. 1990. A unified view of psych-verbs in Italian. In
Dziwirek, K.; Farrel, P.; and Mejias-Bikandi, E. Grammatical
relations: A Cross-theoretical Perspective, 59-81. Stanford, CA: CSLI
Publications.
Croft, William. 1991. Syntactic Categories and Grammatical Relations.
Chicago, IL: University of Chicago Press.
Croft, William. 1993. Case marking and the semantics of mental verbs. In
Pustejovsky, J. (ed.), Semantics and the Lexicon, 55-72. Dordrecht:
Kluwer.
Croft, William. 1998. Event structure in argument linking. In Butt, M.;
and Geuder, W. (eds.), The Projection of Arguments: Lexical and
Compositional Factors, 1-43. Stanford, CA: CSLI Publications.
Croft, William. 2012. Verbs: Aspect and argument structure. Oxford:
Oxford University Press.
Croft, William; and Cruse, Alan D. 2004. Cognitive Linguistics.
Cambridge: Cambridge University Press.
Cronk, Brian C.; Lima, Susan D.; and Schweigert, Wendy A. 1993.
Idioms in sentences: Effects of frequency, literalness, and familiarity.
Journal of Psycholinguistic Research 22 (1), 59-82.
Cronk, Brian C.; and Schweigert, Wendy A. 1992. The comprehension
of idioms: the effects of familiarity, literalness, and usage.
Applied Psycholinguistics 13, 131-146.
Cruse, Alan D. 1986. Lexical Semantics. Cambridge: Cambridge
University Press.
Cutler, Anne. 1982. Idioms: the colder the older. Linguistic Inquiry 13
(2), 317-319.
Cutting, J. Cooper; and Bock, Kathryn. 1997. That’s the way the
cookie bounces: Syntactic and semantic components of experimentally
elicited idiom blends. Memory and Cognition 25 (1), 57-71.
ýermák, František. 2001. Substance of idioms: perennial problems, lack
of data or theory? International Journal of Lexicography 14, 1-20.
ýermák, František. 2007. Frazeologie a Idiomatika ýeská a Obecná.
Praha: Karolinum.
Davidson, Donald. 1967. The logical form of action sentences. In

Rescher, N. (ed.), The Logic of Decision and Action, 81-95. Pittsburgh,
PN: University of Pittsburgh Press.
Dąbrowska, Ewa. 1997. Cognitive Semantics and the Polish Dative.
Berlin: Mouton de Gruyter.
De Groot, Annette M. B. 2002. Lexical representation and processing in
the L2 user. In Cook, V. (ed.), Portraits of the L2 user, 29-64.
Clevedon: Multilingual Matters.
Dell, Gary. 1986. A spreading activation theory of retrieval in language
production. Psychological Review 93, 283-321.
den Dikken, Marcel. 1992. Particles. Leiden: Holland Institute of
Generative Linguistics.
den Dikken, Marcel. 2006. Relators and Linkers: The Syntax of
Predication, Predicate Inversion, and Copulas. Cambridge, MA: MIT
Press.
den Dikken, Marcel. 2010. On the Functional Structure of Locative and
Directional PPs. In Cinque, G.; and Rizzi, L. (eds.), Mapping Spatial
PPs: The Cartography of Syntactic Structures, 74-126. Oxford: Oxford
University Press.
den Dikken, Marcel; Meinunger, André; and Wilder, Chris. 2000.
Pseudoclefts andellipsis. Studia Linguistica 54, 49-89.
Dobrovol'skij, Dmitrij. 1997. Idiome im mentalen Lexikon: Ziele und
Methoden der kognitivbasierten Phraseologieforschung. Fokus,
Linguistisch-Philologische Studien 18. Trier: Wissenschaftlicher
Verlag Trier.
Dobrovol’skij, Dmitrij; and Piirainen, Elisabeth. 2005. Cognitive
theory of metaphor and idiom analysis. Jezikoslovlje 6 (1), 7-35.
Doron, Edit. 2003. Agency and voice: the semantics of the Semitic
templates. Natural Language Semantics 11, 1-67.
Dowty, David. 1979. Word meaning and Montague Grammar. Dordrecht:
D. Reidel Publishing Company.
Dowty, David R. 1991. Thematic Proto-roles and Argument Selection.
Language 67, 547-619.
Dowty, David. 2007. Compositionality as an Empirical Problem. In
Barker, C.; and Jacobson, P. (eds.), Direct Compositionality, 23-101.
Oxford: Oxford University Press.
Drijkoningen, Frank. 2000. Experiencer Objects: Two Types of
Ergativity. In Coopmans, P.; Everaert, M.; and Grimshaw, J. (eds.),
Lexical Specification and Insertion, 69-89. Amsterdam: John
328 References
Egan, Andy. 2008. Pretense for the complete idiom. Noûs 42 (3), 381-
409.
Ekman, Paul. 1992. An argument for basic emotions. Cognition and
Emotion 6 (3), 169-200.
Ekman, Paul. 1999. Basic emotions. In Dalgleish, T.; and Power, M.
(eds.), Handbook of Cognition and Emotion, 45-60. Chichester: Wiley.
Emonds, Joseph. 1976. A Transformational Approach to English Syntax.
New York: Academic Press.
Emonds, Joseph. 1985. A Unified Theory of Syntactic Categories.
Dordrecht: Foris Publications.
Epstein, Samuel. 2007. On I(nternalist)-Functional Explanation in
Minimalism. Linguistic Analysis 33, 20-53.
Ernst, Thomas. 1981. Grist for the linguistic mill: Idioms and “extra”
adjectives. Journal of Linguistic Research 1 (3), 51-68.
Espinal, M. Teresa; and Jaume, Mateu. 2010. On classes of idioms and
their interpretation. Journal of Pragmatics 42 (5), 1397-1411.
Estill, Robert B.; and Kemper, Susan. 1982. Interpreting idioms.
Journal of Psycholinguistic Research 11, 559-568.
Evans, Vyvyan. 2009. How Words Mean. Oxford: Oxford University
Press.
Evans, Vyvyan; and Green, Melanie. 2006. Cognitive Linguistics: An
Introduction. Mahwah, NJ: Erlbaum.
Everaert, Martin. 1993. Verbal idioms, participant idioms and theta
theory. In Smith, L. (ed.), Papers from the third annual meeting of the
formal Linguistics Society of Midamerica, 43-59. Bloomington, IN:
Indiana Linguistics Club.
Everaert, Martin.2010. In Rappaport Hovav, M; Doron, E.; and Sichel, I
(eds.), Lexical Semantics, Syntax, and Event Structure, 76-98. Oxford:
Everaert, Martin; van der Linden, Erik-Jan; Schenk, Andre; and
Schreuder, Rob (eds.). 1995. Idioms: Structural and Psychological
Perspectives. Mahwah, NJ: Lawrence.
Fábregas, Antonio; Marín, Rafael; and McNally, Louise. 2012. From
psych verbs to nouns. In Demonte, V.; and McNally, L. (eds.), Telicity,
Change, and State: A cross-categorial View of Event Structure, 162-
184. Oxford: Oxford University Press.
Fábregas, Antonio; and Marín, Rafael. 2015. Deriving individual-level
and stage-level psych-verbs in Spanish. The Linguistic Review 32 (2),
227-275.
Fellbaum, Christiane. 1993. The determiner in English idioms. In

Cacciari, C; and Tabossi, P. (eds.), Idioms: Processing, Structure and
Interpretation, 271-296. Hillsdale, NJ: Erlbaum.
Fellbaum, Christiane. 2011. Idioms and collocations. In von Heusinger,
K.; Maienborn C.; and Portner, P. (eds.), Semantics: An International
Handbook of Natural Language Meaning, 439-454. Part 1. Berlin:
Mouton de Gruyter.
Fernando, Chitra. 1996. Idioms and Idiomaticity. Oxford: Oxford
University Press.
Fernando, Chitra; and Flavell, Roger. 1981. On Idiom, Critical Views
and Perspectives. Exeter: University of Exeter.
Fiengo, Robert W. 1974. Semantic Conditions on Surface Structure. PhD
Dissertation. Cambridge, MA.: MIT.
Filip, Hana. 1996. Psychological predicates and the syntax-semantics
interface. In Goldberg, A. (ed.), Conceptual Structure, Discourse and
Language, 131-147. Stanford, CA: CSLI Publications.
Filip, Hana. 1999. Aspect, Eventuality Types, and Nominal Reference.
New York: Garland Press (1993 Berkeley, CA PhD Dissertation).
Fillmore, Charles J. 2001. Mini-grammars of some time-when expressions
in English. In Bybee, J. L.; and Noonan, M. (eds.), Complex Sentences
in Grammar and Discourse, 31-59. Amsterdam and Philadelphia, PA:
John Benjamins Publishing Company.
Fillmore, Charles J.; and Kay, Paul. 1995. A Construction Grammar
Coursebook. Unpublished Ms., Berkeley, CA: University of California.
Fillmore, Charles J.; Kay, Paul; and O’Connor, Mary Catherine.
1988. Regularity and Idiomaticity in Grammatical Constructions: The
Case of Let Alone. Language 64 (3), 501-538.
Fillmore, Charles J., and Baker, Collin F. 2001. Frame semantics for
text understanding. Proceedings of WordNet and Other Lexical
Resources Workshop, NAACL, 59-64. Pittsburgh.
Flores d’Arcais, Giovanni, B. 1993. The comprehension and semantic
interpretation of idioms. In Cacciari, C.; and Tabossi, P. (eds.), Idioms:
Processing, Structure, and Interpretation, 79-98. Hillsdale, NJ:
Folli, Raffaella; and Harley, Heidi. 2007. Causation, obligation, and
argument structure: On the nature of little v. Linguistic Inquiry 38 (2),
197-238.
Forrester, Michael A. 1995. Tropic implicature and context in the
comprehension of idiomatic phrases. Journal of Psycholinguistic
Research 24, 1-22.
330 References
Franco, Jon. 1990. Towards a typology of psych-verbs: Evidence from

Spanish. In Green, T.; and Sigal, U., MITWPL 12: Proceedings of the
2nd Meeting of SCIL, 46-62. Cambridge, MA: MIT Press.
Fraser, Bruce. 1970. Idioms within a transformational grammar.
Foundations of Language 6 (1), 22-42.
Frege, Gottlob. 1884 [1980]. The Foundations of Arithmetic (2nd ed.).
Austin, J. L. (trans.). Evanston, IL: Northwestern University Press.
Frijda, Nico H. 1986. The Emotions. Cambridge: Cambridge University
Press.
Frijda, Nico H. 2007. What might emotions be? Comments on the
Comments. Social Science Information 46, 433-443.
Gallego, Ángel J. 2010. Phase Theory. Amsterdam and Philadelphia:
John Benjamins Publishing Company.
Gasser, Michael; and Dyer, Michael. 1986. Speak of the devil:
Representing deictic and speech act knowledge in an integrated
memory. In Proceedings of the Sixth Annual Meeting of the Cognitive
Science Society, 388-398. Hillsdale, NJ: Erlbaum.
Gavelin, Emma. 2015. Conceptual Metaphors: a Diachronic Study of
Love Metaphors in Mariah Carey’s Song Lyrics. Umeå: Umeå
University.
Gazdar, Gerald; Klein, Ewan; Pullum, Geoffrey K; and Sag, Ivan A.
1985. Generalized Phrase Structure Grammar. Oxford: Basil
Blackwell.
Gernsbacher, Ann Morton. 1984. Resolving 20 years of inconsistent
interactions between lexical familiarity and orthography, concreteness,
and polysemy. Journal of Experimental Psychology: General 113,
256-281.
Gibbs, Raymond W. 1980. Spilling the beans on understanding and
memory for idioms in conversation. Memory and Cognition 8 (2), 149-
156.
Gibbs, Raymond W. 1984. Literal meaning and psychological theory.
Cognitive Science 8, 275-304.
Gibbs, Raymond W. 1985. On the process of understanding idioms.
Journal of Psycholinguistic Research 14 (5), 465-472.
Gibbs, Raymond W. 1986. What makes some speech acts conventional?
Journal of Memory and Language 25, 181-196.
Gibbs, Raymond W. 1989. Understanding and literal meaning. Cognitive
Science 13, 243-251.
Gibbs, Raymond W. 1990. Psycholinguistic studies on the conceptual
basis of idiomaticity. Cognitive Linguistics 1, 417-451.
Gibbs, Raymond W. 1992. What Do Idioms Really Mean? Journal of

Memory and Language 31 (4), 485-506.
Gibbs, Raymond W. 1993. Why Idioms Are Not Dead Metaphors. In
Cacciari, C.; and Tabossi, P. (eds.), Idioms: Processing, Structure, and
Interpretation, 57-78. Hillsdale, NJ: Lawrence Erlbaum Associates.
Gibbs, Raymond W. 1994. The Poetics of Mind: Figurative Thought,
Language, and Understanding. Cambridge: Cambridge University
Press.
Gibbs, Raymond W. 1995. Idiomaticity and human cognition. In
Everaert, M.; van der Linden, E.-J.; Schenk, A.; and Schreuder, R.
(eds.), Idioms: Structural and Psychological Perspectives, 97-116.
Hillsdale, NJ: Lawrence Erlbaum Associates.
Gibbs, Raymond W. 2002. A new look at literal meaning in
understanding what is said and implicated. Journal of Pragmatics 34,
457-486.
Gibbs, Raymond W.; Bogdanovich, Josephine M.; Sykes, Jeffrey R.;
and Barr, Dale J. 1997. Metaphor in idiom comprehension. Journal of
Memory and Language 37, 141-154.
Gibbs, Raymond W.; and Gonzales, Gayle P. 1985. Syntactic
frozenness in processing and remembering idioms. Cognition 20, 243-
259.
Gibbs, Raymond W.; and Nayak, Nandini P. 1989. Psycholinguistic
studies on the syntactic behaviour of idioms, Cognitive Psychology 21,
100-138.
Gibbs, Raymond W.; Nayak, Nandini; and Cutting, Cooper. 1989
[Gibbs et al. 1989a]. How to kick the bucket and not decompose:
Analyzability and idiom processing. Journal of Memory and Language
28, 576-593.
Gibbs, Raymond W.; Nayak, Nandini; Bolton, John; and Keppel,
Melissa. 1989 [Gibbs et al. 1989b]. Speaker’s assumptions about the
lexical flexibility of idioms. Memory and Cognition 17 (1), 58-68.
Gibbs, Raymond W.; and O’Brien, Jennifer. 1990. Idioms and mental
imagery: the metaphorical motivation for idiomatic meaning.
Cognition 36, 35-68.
Gibbs, Raymond W.; O'Brien, Jennifer; and Doolittle, Shelly. 1995.
Inferring meanings that are not intended: Speakers’ intentions and
irony comprehension. Discourse Processes 20, 187-203.
Ginzburg, R. S.; Khidekel, S. S.; Knyazeva, G. Y.; and Sankin, A. A.
2004. A Course in Modern English Lexicology. Moscow: Moscow
Vysšaja Škola.
332 References
Giora, Rachel. 1997. Understanding figurative and literal language: The

graded salience hypothesis. Cognitive Linguistics 7, 183-206.
Giora, Rachel. 1999. On the priority of salient meanings: studies of literal
and figurative language. Journal of Pragmatics 31 (7), 919-929.
Giora, Rachel. 2002. Literal vs. figurative language: Different or equal?
Journal of Pragmatics 34, 487-506.
Giora, Rachel. 2003. On Our Mind: Salience, Context, and Figurative
Language. Oxford: Oxford University Press.
Gläser, Rosemarie. 1988. The grading of idiomaticity as a presupposition
for a taxonomy of idioms. In Htillen, W.; and Schultz, R. (eds.),
Understanding the Lexicon. Meaning Sense and World Knowledge in
Lexical Semantics, 264-279. Tübingen: Max Niemeyer Verlag.
Gläser, Rosemarie. 1998. The Stylistic Potential of Phraseological Units
in the Light of Genre Analysis. In Cowie, A. P. (ed.), Phraseology:
Theory, Analysis, and Applications, 125-143. Oxford: Clarendon Press.
Glasbey, Sheila R. 2003. Let’s paint the town red for a few hours:
Composition of aspect in idioms. LexFig ’03: Proceedings of the ACL
2003 Workshop on Lexicon and Figurative Language, Volume 14, 43-
49. Stroudsburg, PA: Association for Computational Linguistics.
Glasbey, Sheila R. 2007. Aspectual composition in idioms. In Saussure,
L. de; Moeschler, J.; and Puskás, G. (eds.), Recent advances in the
syntax andsemantics of tense, aspect and modality, 71-88. (Trends in
Linguistics. Studies and Monographs [TiLSM] 185). Berlin: Mouton
de Gruyter..
Glass, Arnold L. 1983. The comprehension of idioms. Journal of
Psycholinguistic Research 12, 429-442.
Glucksberg, Sam. 1991. Beyond literal meanings: The psychology of
allusion. Psychological Science 2, 146-152.
Glucksberg, Sam. 1993. Idiom Meanings and Allusional Content. In
Cacciari, C.; and Tabossi, P. (eds.), Idioms: Processing, Structure, and
Interpretation, 3-26. Hillsdale, NJ: Lawrence Erlbaum Associates.
Glucksberg, Sam. 2001. Understanding Figurative Language: From
Metaphors to Idioms. Oxford: Oxford University Press.
Glucksberg, Sam; Brown, Mary; and McGlone, Matthew S. 1993.
Conceptual metaphors are not automatically accessed during idiom
comprehension. Memory and Cognition 21, 711-719.
Goldberg, Adele. E. 1995. Constructions: A Construction Grammar
Approach to Argument Structure. Chicago, IL: University of Chicago
Press.
Goldberg, Lotus Madelyn. 2005. Verb-stranding VP Ellipsis: A Cross-
linguistic Study. PhD Dissertation. Montreal: McGill University.
Grafmiller, Jason. 2013. The Semantics of Syntactic Choice: An Analysis

of English Emotion Verbs. PhD Dissertation. Stanford, CA: Stanford
University.
Gramley, Stephan; and Pátzold, Michael. 1992. A Survey of Modern
English. London; New York: Routledge.
Grant, Lynn; and Bauer, Laurie. 2004. Criteria for re-defining idioms:
Are we barking up the wrong tree? Applied Linguistics 25, 83- 61.
Grégoire, Nicole. 2009. Untangling Multiword Expressions: A Study on
the Representation and Variation of Dutch Multiword Expressions.
PhD Dissertation. Utrecht: University of Utrecht.
Grimshaw, Jane. 1990. Argument structure. Cambridge, MA: MIT Press.
Grimshaw, Jane. 1991. Extended Projection. Ms., Waltham, MA:
Brandeis University.
Guéron, Jaqueline. 1986. Clause Union and the Verb-Particle
Construction in English. Paper presented at NELS 17.
Guidi, Luiz Guilherme. 2011. Old English psych-verbs and quirky
experiencers. York Papers in Linguistics Series 2 (11), 29-48.
Haegeman, Liliane. 1994. Introduction to Government and Binding
Theory. Oxford: Blackwell.
Haegeman, Liliane. 2010. The internal syntax of adverbial clauses.
Lingua 120, 628-648.
Haider, Hubert. 2012. Symmetry Breaking in Syntax. Cambridge, UK:
Cambridge University Press (Cambridge Studies in Linguistics 136).
Hamblin, Jennifer L.; and Gibbs, Raymond W. 1999. Why you can’t
kick the bucket as you slowly die: Verbs in idiom understanding.
Journal of Psycholinguistic Research 39, 25-39.
Handel, Steven. 2011. Classification of Emotions. Retrieved from
http://www.theemotionmachine.com/classification-of-emotions
Harley, Heidi. 2002. Possession and the Double Object Construction.
Linguistic Variation Yearbook 2, 29-68.
Harley, Heidi; and Stone, Schildmier Megan. 2013. The “No Agent
Idioms” hypothesis. In Folli, R.; Sevdali, C. and Truswell, R. (eds.),
Syntax and Its Limits, 283-311. Oxford: Oxford University Press.
Hartshorne, Joshua, K.; O’Donnell, Timothy, J.; Sudo, Yasutada;
Uruwashi, Miki; and Snedeker, Jesse. 2010. Linking meaning to
language: linguistic universals and variation. In Proceedings of the
Thirty-Second Annual Conference of the Cognitive Science Society,
1186-1191. Portland, OR.
Harves, Stephanie. 2002. Where Have All the Phases Gone? (Non-
)Defective Categories and Case Alternations in Russian. In Toman, J.
334 References
(ed.), Formal Approaches To Slavic Linguistics 10: The Second Ann

Arbor Meeting, 97-118. Ann Arbor, MI: Michigan Slavic Publications.
Harwood, William. 2013. Being Progressive is Just a Phase: Dividing
the Functional Hierarchy. PhD Dissertation. Gent: Universiteit Gent.
Harwood, William. 2015. Being progressive is just a phase: celebrating
the uniqueness of progressive aspect under a phase-based analysis.
Harwood, Will; Hladnik, Marko; Leufkens, Sterre; Temmerman,
Tanja; Corver, Norbert; and van Craenenbroeck, Jeroen. 2016.
Idioms: Phasehood and Projection. Ms., Brussel: KU Leuven.
Harwood, William; Hladnik, Marko; Leufkens, Sterre; and
Temmerman, Tanja. 2017. Idioms and Phases. Paper presented at
The syntax of idioms workshop, on 20 January 2017. Utrecht: Utrecht
University.
Harwood, William; and Temmerman, Tanja. 2015. Pushing the
boundaries: idioms and phases in Dutch dialects and English. Papers
presented at CGSW 30. Chicago, IL: University of Chicago.
Hay, Jennifer; Kennedy, Christopher; and Levin, Beth. 1999. Scalar
Structure Underlies Telicity in Degree Achievements. In Mathews, T.;
and Strolovitch, D. (eds.). SALT IX, 127-144. Ithaca: CLC Publications.
Hicks, Glyn. 2009. Tough-constructions and their derivation. Linguistic
Inquiry 40 (4).
Higgins, Roger F. 1974. On the use of idioms as evidence for movement:
A cautionary note. Paper presented at the Linguistics Society of
America Annual Meeting, New York, December 1974.
Higgins, Roger F. 1981. On what are here called proxy clauses.
Unpublished Ms.
Hinrichs, Erhard W. 1985. A Compositional Semantics for Aktionsarten
and NP Reference. PhD Dissertation. Ohio, Columbus: Ohio State
University.
Hinzen, Wolfram. 2012. Phases and Semantics. In Gallego, A. (ed.),
Phases: Developing the Framework, 309-342. Berlin: Mouton de
Gruyter.
Horn, George M. 2003. Idioms, metaphors and syntactic mobility.
Journal of Linguistics 39 (2), 245-273.
Howarth, Peter. 1998. Phraseology and second language proficiency.
Applied Linguistics 19, 24-44.
Ifill, Tim. 2002. Seeking the nature of idioms: a study in idiomatic
structure. Ms., Haverford College.
Isse, Yoko. 2008. On Psych-verbs in English and Japanese. Fukuoka
University Review of Literature and Humanities 39 (4), 967-984.
Iwata, Seizi. 1995. The Distinctive Character of Psych-Verbs as

Causatives. Linguistic Analysis 1-2, 95-120.
Jackendoff, Ray. 1977. X-bar Syntax: A Study of Phrase Structure.
Cambridge, MA: MIT Press.
Jackendoff, Ray. 1983. Semantics and Cognition. Cambridge, MA: MIT
Press.
Jackendoff, Ray. 1989. What is a concept, that a person may grasp it?
Mind and Language 4, 68-102.
Jackendoff, Ray. 1990. Semantic Structures. Cambridge, MA: MIT Press.
Jackendoff, Ray. 1996. The proper treatment of measuring out, telicity,
and perhaps even quantification in english. Natural Language and
Linguistic Theory 14, 305-354.
Jackendoff, Ray. 1997. The Architecture of the Language Faculty.
Jackendoff, Ray. 2002. Foundations of Language: Brain, Meaning,
Grammar, Evolution. Oxford: Oxford University Press.
Jiménez-Fernández, Ángel L. 2000. The aspectual morpheme as and
feature movement in argument small clauses. Generative Linguistics in
Poland 1, 59-69.
Jiménez-Fernández, Ángel L.; and Spyropoulos, Vassilios. 2013.
Feature Inheritance, vP Phases and the Information Structure of Small
Clauses. Studia Linguistica 67 (2), 185-224.
Jiménez Fernández, Ángel Luis; and Rozwadowska, BoĪena. 2014. An
information structure-based approach to psych constructions. Paper
presented at the Congress ‘Various Dimensions of Contrastive Studies’
in Szczyrk, Katowice.
Ji-Xin, Qiji. 2009. On trainability of idioms. Sino-Us English Teaching 6
(2), 57-60.
Johnson, Kyle. 1991. Object positions. Natural Language and
LinguisticTheory 9 (4), 577-636.
Johnson, Kyle. 1992. Scope and Binding Theory: Comments on
Zubizarreta. In Stowell, T.; and Wehrli, E. (eds.), Syntax and
Semantics, Vol. 26: Syntax and the Lexicon, 259-275. San Diego, CA:
Academic Press Inc.
Johnson, Kyle. 2001. What VP ellipsis can do, and what it can’t, but not
why. In Baltin, M. M.; and Collins, C. (eds.), The Handbook of
Contemporary Syntactic Theory, 439-479. Oxford: Blackwell
Publishers.
Johnson-Laird, Philip N.; and Oatley, Keith. 1989. The Language of
emotions: An analysis of a semantic field. Cognition and Emotion 3,
81-123.
336 References
Johnson-Laird, Philip N.; and Oatley, Keith. 1992. Basic emotions,

rationality, and folk theory. Cognition and Emotion 6 (3-4), 201-223.
Katz, Jerrold J. 1973. Compositionality, idiomaticity and lexical
substitution. In Anderson, S. R.; and Kiparsky, P. (eds.), A Festschrift
for Morris Halle, 357-376. New York: Holt, Reinhart and Winston.
Katz, Jerrold J.; and Postal, Paul. 1963. Semantic Interpretation of
Idioms and Sentences Containing Them. Quarterly Progress Report of
the MIT Research Laboratory of Electronics 70, 275-282.
Katz, Jerrold J.; and Postal, Paul. 1964. An Integrated Theory of
Linguistic Descriptions. Cambridge, MA: MIT Press.
Kayne, Richard. 1985. Principles of Particle Constructions. In Guéron, J.;
Obenauer, H.-G.; and Pollock, J.-Y. (eds.), Grammatical
Representation, 101-140. Dordrecht: Foris Publications.
Kayne, Richard S. 2004. Prepositions as Probes. In Belletti, A. (ed.),
Structures and Beyond, 192-X. Oxford: Oxford University Press.
Kavka, Stanislav. 2003. A Book on Idiomatology. Žilina: University of
Žilina.
Kellerman, Eric. 1999. A break with tradition. In Foster, S.; and Perdue,
C. (eds.), Proceedings of EUROSLA 8. AILE, 7-36. Paris: British
Institute in Paris.
Kenny, Anthony. 1963. Action, Emotion and Will. London: Routledge
and Kegan Paul.
Keysar, Boaz; and Bly, Briget. 1995. Intuitions of the transparency of
idioms: Can one keep a secret by spilling the beans? Journal of
Memory and Language 34, 89-109.
Keysar, Boaz; and Bly, Bridget. 1999. Swimming against the current:
Do idioms reflect conceptual structure? Journal of Pragmatics 31,
1559-1578.
Kim, Jaegwon. 1969. Events and their Descriptions: Some
Considerations. In Rescher, N. (ed.), Essays in Honor of Carl G.
Hempel, 198-215. Dordrecht: Reidel.
Kim, Jong-Bok. 2013. On the Existence of Small Clauses in English.
English Language and Linguistics 19 (1), 67-88.
Kim, Kyumin. 2014. Phases and idioms. WAFL 9, 209-221.
Kim, Kyumin. 2015. Phase based account of idioms and its
consequences. Linguistic Research 32, 631-670.
Kim, Young-Joo; and Larson, Richard. 1989. Scope interpretation and
the syntax of psych-verbs. Linguistic Inquiry 20 (4), 681-688.
Kiparsky, Paul. 1976. Oral poetry: Some linguistic and typological
considerations. In Stolz, B. A.; and Shannon, R. S. (eds.), Oral
Literature and the Formula, 73-106. Ann Arbor: Center for the
Coordination of Ancient and Modern Studies.
Kiparsky, Paul. 1987. Morphology and grammatical relations. Ms.,
Stanford, CA: Stanford University.
Kipper-Schuler, Karin. 2005. VerbNet: A Broad-coverage, Comprehensive
Verb Lexicon. PhD Dissertation. Philadelphia: Computer and
Information Science Dept., University of Pennsylvania.
Kitagawa, Yoshihisa. 1985. Small but clausal. Chicago Linguistic Society
21, 210-220.
Kitagawa, Yoshihisa. 1986. Subjects in Japanese and English. PhD
Dissertation. Amherst, MA: University of Massachusetts.
Klein, Katarina; and Silvia Kutscher. 2005. Lexical Economy and Case
Selection of Psych-Verbs in German. Ms., Köln/Bochum.
Knappe, Gabriele. 2004. Idioms and Fixed Expressions in English
Language Study before 1800. A Contribution to English Historical
Phraseology. Frankfurt am Main: Peter Lang.
Koopman, Hilda. 1991. The Verb Particle Construction and the Syntax of
PPs. Ms., UCLA.
Koopman, Hilda. 2010. Prepositions, postpositions, circumpositions, and
particles. In Cinque, G.; and Rizzi, L. (eds.), Mapping Spatial PPs:
The cartography of Syntactic Structures, vol. 6, 26-73. Oxford: Oxford
University Press
Koopman, Hilda; and Sportiche, Dominique. 1991. The position of
subjects. Lingua 85, 211-258.
Koster, Jan. 1975. Dutch as an SOV language. Linguistic Analysis 1.2,
111-136.
Kövecses, Zoltán. 2000. Metaphor and Emotion: Language, Culture and
the Body In Human Feeling. Cambridge: Cambridge University Press.
Kövecses, Zoltán. 2005. Metaphor in Culture: Universality and
Variation. New York: Cambridge University Press.
Kövecses, Zoltán. 2010. Metaphor: A Practical Introduction. Oxford:
Krifka, Manfred. 1989. Nominal reference, temporal constitution and
quantification in event semantics. In Bartsch, R.; van Benthem, J.; and
van Emde Boas, P., Semantics and Contextual Expression, 75-115.
Stanford, CA: CSLI Publications.
Krifka, Manfred. 1992. Thematic relations as links between nominal
reference and temporal constitution. In Sag, I.; and Szabolsci, A.
(eds.), Lexical Matters, CSLI Lecture Notes, 29-53. Chicago, IL:
University of Chicago Press.
338 References
Krifka, Manfred. 1998. The origins of telicity. In Rothstein, S. (ed.),

Events and Grammar, 197-235. Dordrecht: Kluwer.
Kroll, Judith F. 1993. Accessing conceptual representations for words in
a second language. In Schreuder, R.; and Weltens, B. (eds.), The
Bilingual Lexicon, 53-82. Amsterdam: John Benjamins Publishing
Company.
Kroll, Judith F.; and Sholl, Alexandra. 1992. Lexical and conceptual
memory in fluent and non-fluent bilinguals. In Harris, R. J. (ed.),
Cognitive Processing in Bilinguals, 191-203. Amsterdam: North-
Holland.
Lakoff, George. 1966. Stative adjectives and verbs in English. NSF-
Report 17, Harvard Computational Laboratory.
Lakoff, George. 1987. Women, Fire and Dangerous Things. What
Categories Reveal about the Mind. Chicago, London: The University
of Chicago Press.
Lakoff, George; and Johnson, Mark. 1980. Metaphors We Live By.
Chicago, IL: The University of Chicago Press.
Lambrecht, Knud. 2001. A framework for the analysis of cleft
constructions. Linguistics 39 (3), 463-516.
Landau, Idan. 2003. Introduction to Syntax: Recent Minimalism:
Chomsky (2000, 2001). Cambridge, MA: MIT Press.
Landau, Idan. 2005. The Locative Syntax of Experiencers. Beer Sheva:
Ben Gurion University.
Landau, Idan. 2010. The Locative Syntax of Experiencers (Linguistic
Inquiry Monographs 53: x+165). Cambridge, MA: MIT Press.
Langacker, Ronald. 1987. Foundations of Cognitive Grammar, Vol 1.
Theoretical Prerequisites. Stanford, CA: Stanford University.
Langlotz, Andreas. 2006. Idiomatic Creativity: A Cognitive-Linguistic
Model of Idiom-Representation and Idiom-Variation in English.
Amsterdam and Philadelphia: John Benjamins Publishing Company.
Larson, Richard K. 1988. On the double object construction. Linguistic
Inquiry 19 (3), 335-391.
Larson, Richard K. 1990. Double Objects Revisited: Reply to
Jackendoff. Linguistic Inquiry 21, 589-632.
Lasnik, Howard; and Fiengo, Robert. 1974. Complement object
deletion. Linguistic Inquiry 5 (4), 535-571.
Lasnik, Howard; and Saito, Mamoru. 1999. On the subject of
infinitives. In Lasnik, H. (ed.), Minimalist Analysis, 7-24. Malden,
MA: Blackwell.
Laufer, Batia. 1997. The lexical plight in second language reading.
Words you don’t know, words you think you know, and words you
can't guess. In Coady, J.; and Huckin, T. (eds.), Second Language

Vocabulary Acquisition: a Rationale for Pedagogy, 20-34. Cambridge:
Cambridge University Press.
Lechner, Winfried. 2006. Double object constructions (II). Paper
presented at SOSES. Stuttgart: Stuttgart University.
Legate, Julie Anne. 2002. Phases in Beyond Explanatory Adequacy. Ms.,
Legate, Julie Anne. 2012. The size of Phases. In Gallego, A. (ed.),
Phases, 233-250. Berlin: Mouton de Gruyter.
Legendre, Géraldine. 1989. Inversion with certain French Experiencer
verbs. Language 65 (4), 752-782.
Legendre, Géraldine. 1993. Antipassive with French psych-verbs. In
Dunca, E.; Farkas, D.; Spaelti, F. (eds.), Proceedings of WCCFL 12,
373-388. Stanford, CA: CSLI Publications.
Lennon, Paul. 1998. Approaches to the Teaching of Idiomatic Language.
International Review of Applied Linguistics in Language Teaching 36
(1), 11-30.
Levelt, Willem J. M. 1989. Speaking: From Intention to Articulation.
Levelt, Willem J. M.; and Meyer, Antje S. 2000. Word for word:
Multiple lexical access in speech production. European Journal of
Cognitive Psychology 12 (4), 433-452.
Levin, Beth. 1993. English Verb Classes and Alternations: A Preliminary
Investigation. Chicago, IL: University of Chicago Press.
Levin, Beth; and Rappaport Hovav, Malka. 1995. Unaccusativity: at
the Syntax-Lexical Semantics Interface. Cambridge. MA: MIT Press.
Levin, Beth; and Rappaport Hovav, Malka. 2005. Argument
realization. Cambridge: Cambridge University Press.
Levin, Lorraine. 1986. Operations on Lexical Forms: Unaccusative
Rules in Germanic Languages. PhD Dissertation. Cambridge, MA:
MIT Press.
Levinson, Stephen C.; Senft, Gunter; and Majid, Asifa. 2007. Emotion
categories in language and thought. In Majid, A. (ed.), Field Manual
Volume 10, 46-52. Nijmegen: Max Planck Institute for Psycholinguistics.
Levorato, M. Chiara. 1993. The acquisition of idioms and the development
of figurative competence. In Cacciari, C.; and Tabossi, P. (eds.)
Idioms: Processing, Structure and Interpretation, 101-128. Hillsdale,
NJ: Lawrence Erlbaum Associates.
Levorato, M. Chiara; Roch, Maja; and Nesi, Barbara. 2007. A
longitudinal study of idiom and text comprehension. Journal of Child
Language 34, 473-494.
340 References
Liontas, John. I. 2003. Killing two birds with one stone: understanding
Spanish VP idioms in and out of context. Hispania 86 (2), 289-301.
Lipka, Leonhard. 2002. English Lexicology. Lexical Structure, Word
Semantics and Word Formation. 3rd ed. of An Outline of English
Lexicology, 1990. Tübingen: Gunter Narr Verlag.
Lin, Chien-Jer Charles. 2007. Processing (In)alienable Possessions at the
Syntax- Semantics Interface. Paper presented at the Conference on
Linguistic Interfaces, University of Ulster, Northern Ireland, June 1-3,
2007.
Lin, Jimmy. 2004. Event Structure and the Encoding of Arguments: The
Syntax of the Mandarin and English Verb Phrase. PhD Dissertation.
Cambridge, MA: Massachusetts Institute of Technology.
Literary Devices Editors. 2013. Colloquialism. Retrieved from
https://literarydevices. net/colloquialism/
Liu, Dilin. 2008. Idioms: Description, Comprehension, Acquisition, and
Pedagogy. New York: Routledge.
Longobardi, Giuseppe. 1994. Reference and proper names: a theory of
N-raising insyntax and logical form. Linguistic Inquiry 25, 609-665.
Machonis, Peter. 1985. Transformations of verb phrase idioms:
Passivization, particle movement, dative shift. American Speech 60,
291-308.
MacKay, Donald G.; and Bever, Thomas G. 1967. In search of
ambiguity. Perception and Psychophysics 2, 193-200.
MacWhinney, Brian. 2008. A unified model. In Ellis, N. C.; and
Robinson, P. (eds.), Handbook of Cognitive Linguistics and Second
Language Acquisition, 341-371. New York: Routledge.
Maienborn, Claudia. 2003. Against a Davidsonian Analysis of Copula
Sentences. In Kadowaki, M.; and Kawahara, S. (eds.), NELS 33
Proceedings, 167-186. Amherst: GLSA.
Maienborn, Claudia. 2005. On the limits of the Davidsonian approach:
the case of copula sentences. Theoretical Linguistics 31 (3), 275-316.
Maienborn, Claudia. 2007. On Davidsonian and Kimian states. In
Comorovski, I.; and Heusinger, Klaus von (eds.), Existence: Semantics
and Syntax, 107-130. Dordrecht: Kluwer.
Makkai, Adam. 1972. Idiom Structure in English. Mouton: The Hague.
Marantz, Alec. 1984. On the Nature of Grammatical Relations.
Marantz, Alec. 1993. Implications of Asymmetries in Double Object
Constructions. In Mchombo, S. (ed.), Theoretical Aspects of Bantu
Grammar, 113-150. Stanford, CA: CSLI Publications.
Marantz, Alec. 1997. No escape from syntax: Don’t try morphological

analysis in the privacy ofyour own lexicon. In Dimitriadis, A.; Siegel,
L.; Surek-Clark, C.; and Williams, A. (eds.), Proceedings of the 21st
Annual Penn Linguistics Colloquium, number 4.2 in Penn Working
Papers in Linguistics, 201-225. Philadelphia: University of
Pennsylvania.
Marantz, Alec. 2001. Words. Ms., Cambridge, MA: MIT Press.
Marín, Rafael. 2001. El Componente Aspectual de la Predicación. PhD
Dissertation. Barcelona: Universitat Autònoma de Barcelona.
Marín, Rafael; and McNally, Louise. 2005. The Aktionsart of Spanish
Reflexive Psychological Verbs. Proceedings of Sinn und Bedeutung 9,
212-225. Ithaca, NY: CLC Publications.
Marín, Rafael; and McNally, Louise. 2011. Inchoativity, change of state,
and telicity: Evidence from Spanish reflexive psychological verbs.
Marschark, Marc; Katz, Albert N.; and Paivio, Allan. 1983.
Dimensions of metaphor. Journal of Psycholinguistic Research 12 (1),
17-40.
Mattissen, Johanna. 2003. Dependent-head Synthesis in Nivkh: a
Contribution to a Typology of Polysynthesis. Amsterdam: John
Mattissen, Johanna. 2006. The ontology and diachrony of polysynthesis.
In Wunderlich, D. (ed.), Advances in the Theory of the Lexicon, 287-
354. Berlin, New York: Mouton de Gruyter.
Matushansky, Ora. 2000. The Instrument of Inversion: Instrumental Case
and Verb Raising in the Russian Copula. In Billerey, R.; and
Lillehaugen, B. (eds.), WCCFL 19: Proceedings of the 19th West
Coast Conference on Formal Linguistics, 101-115. Somerville, MA:
Cascadilla Press.
Matushansky, Ora. 2005. Going through a phase. In McGinnis, M.; and
Richards, N. (eds.) Perspectives on Phases. MITWPL 49. Cambridge,
MA: MITWPL.
Mäntylä, Katja. 2004. Idioms and Language Users: the Effect of the
Characteristics of Idioms on their Recognition and Interpretation by
Native and Non-Native Speakers of English. PhD Dissertation.
Jyväskylä: University of Jyväskylä.
McArthur, Tom (ed.). 1992. The Oxford Companion to the English
Language. Oxford and New York: Oxford University Press.
McCarthy, Michael; and O'Dell, Felicity. 2002. English Idioms in Use.
342 References
McCawley, James D. 1981. The syntax and semantics of English relative

clauses. Lingua 53 (2), 99-149.
McCloskey, James; and Sells, Peter. 1988. Control and A-Chains in
Modern Irish. Natural Language and Linguistic Theory 6, 143-189.
McGinnis, Martha. 2001. Variation in the Phase Structure of
Applicatives. Linguistic Variation Yearbook 1, 105-146.
McGinnis, Martha. 2002. On the systematic aspect of idioms. Linguistic
Inquiry 33 (4), 665-672.
McGlone, Matthew S.; Glucksberg, Sam; and Cacciari, Cristina.
1994. Semantic productivity in idiom comprehension. Discourse
Processes 17, 167-190.
Meetham, Alfred R.; and Hudson, R. A. (eds.) 1969. Encyclopaedia of
Linguistics, Information, and Control. Oxford: Pergamon Press.
Mehler, Jacques; and Bever, Thomas G. 1967. Cognitive capacity of
very young children. Science 158 (3797), 141-142.
Mohanan, K. Verma; and Mohanan, Tara. 1990. Dative Subjects in
Malayalam: Semantic Information in Syntax. In Manindra, K. V.;
Mohanan, T. (eds.), Experiencer Subjects in South Asian Languages,
43-57. Stanford, CA, CSLI Publications.
Moon, Rosamund. 1992. Textual aspects of fixed expressions in learner’s
dictionaries. In Arnaud, P. J. L.; and Bejoint, H. (eds.), Vocabulary and
Applied Linguistics, 13-27. London: MacMillan.
Moon, Rosamund. 1998. Fixed Expressions and Idioms in English: A
Corpus-Based Approach. Oxford: Oxford University Press.
Moore, F. C. T. 1982. On taking metaphor literally. In Miall, D. S. (ed.),
Metaphor: Problems and Perspectives, 1-13. Sussex: The Harvester
Press Ltd.
Mueller, Rachel A. G.; and Gibbs, Raymond. 1987. Processing idioms
with multiple meanings. Journal of Psycholinguistic Research 16, 63-
81.
Müller, Gereon. 2004. Phrase Impenetrability and Wh-Intervention. In
Stepanov, A.; Fanselow,G.; and Vogel, R. (eds.), MinimalityEffects in
Syntax, 289-325. Berlin: Mouton de Gruyter.
Myers, David G. 2004. Theories of Emotion. In Psychology: Seventh
Edition, 500. New York, NY: Worth Publishers.
Nation, Paul. 2006. How large a vocabulary is needed for reading and
listening? Canadian Modern Language Review 63 (1), 59-82.
Nayak, Nandini P.; and Gibbs, Raymond W. 1990. Conceptual
knowledge in the interpretation of idioms. Journal of Experimental
Psychology: General 119 (3), 315-330.
Needham, William P. 1992. Limits on Literal Processing During Idiom

Interpretation. Journal of Psycholinguistic Research 21, 1-16.
Newmeyer, Frederick J. 1974. The Regularity of Idiom Behavior. Lingua
34 (4), 327- 342.
Nenonen, Marja. 2002. Idiomit ja leksikko: Suomen kielen lausekeidiomien
syntaktisia, semanttisia ja morfologisia piirteitä. [Idioms and the
lexicon. Syntactic, semantic, and morphological features of Finnish
phrasal idioms]. PhD dissertation. Joensuu: University of Joensuu
Publications of the Humanities.
Neufeldt, Victoria (ed.). 1991. Webster’s New World Dictionary (3rd
College ed.) New York: Prentice Hall.
Nida, Eugene A. 2001. Contexts in Translatng. Amsterdam /
Philadelphia: John Benjamins Publishing Company.
Norrick, Neil. 1985. How Proverbs Mean: Semantic Studies in English
Proverbs. Amsterdam: Mouton.
Nunberg, Geoffrey. 1978. The Pragmatics of Reference. Berkeley, CA:
Indiana University Linguistics Club.
Nunberg, Geoffrey. 1979. The non-uniqueness of semantic solutions:
Polysemy. Linguistics and Philosophy 3 (2), 143-184.
Nunberg, Geoffrey; Sag, Ivan; and Wasow, Thomas. 1994. Idioms.
Language 70 (3), 491-538.
O’Dell, Felicity; and McCarthy, Michael. 2010. English Idioms in Use
Advanced. Cambridge: Cambridge University Press.
O’Grady, William. 1998. The syntax of idioms. Natural Language and
Linguistic Theory 16, 279-312.
Ortony, Andrew; Schallert, Diane L.; Reynolds, Ralph E.; and Antos,
Stephen J. 1978. Interpreting metaphors and idioms: some effects of
context on comprehension. Journal of Verbal Learning and Verbal
Behavior 17, 465-477.
Osborne, Tim; Putnam, Michael; and Groß, Thomas. 2012. Catenae:
Introducing a novel unit of syntactic analysis. Syntax 15 (4), 354-396.
Palmer, Frank R. 1986. Semantics. Cambridge: Cambridge University
Press.
Papagno, Costanza; and Romero Lauro, Leonor J. 2010. The neural
basis of idiom processing: Neuropsychological, neurophysiological
and neuroimaging evidence. Italian Journal of Linguistics 22 (1), 21-
40.
Parodi, Claudia; and Luján, Marta. 2000. Aspect in Spanish psych
verbs. In Campos, H. (ed.), Hispanic Linguistics at the Turn of the
Century, 210-221. Somerville, MA: Cascadilla Press.
344 References
Parsons, Terence. 1990. Events in the Semantics of English: A Study in

Subatomic Semantics. Cambridge, MA: MIT Press.
Partee, Barbara. 1984. Compositionality. In Landman, F.; and Veltman,
F. (eds.), Varieties of Formal Semantics, 281-312. Dordrecht: Foris.
Pavlenko, Aneta. 1999. New approaches to concepts in bilingual
memory. Bilingualism: Language and Cognition 2, 209-230.
Peacock, George. M. 2009. Unlocking the Idioms: An LDS Perspective
on Understanding Scriptural Idioms. Springville: Cedar Fort.
Perlmutter, David. 1970. The two verbs begin. In Jacobs, R. A.; and
Rosenbaum, P. S. (eds.), Readings in English Transformational
Grammar, 107-119. Waltham, MA: Ginn and Company.
Perlmutter, David M.; and Postal, Paul. 1984. The 1-Advancement
Exclusiveness Law. In Perlmutter, D. M.; and Rosen, C. (eds.), Studies
in Relational Grammar 2, 81-125. Chicago, IL: The University of
Chicago Press.
Pesetsky, David. 1982. Complementizer-Trace Phenomena and the
Nominative Island Condition. The Linguistic Review 1, 297-343.
Pesetsky, David. 1987. Binding problems with Experiencer verbs.
Linguistic Inquiry 18, 126-140.
Pesetsky, David. 1995. Zero Syntax: Experiencers and Cascades.
Pesetsky, David; and Torrego, Esther. 2007. The syntax of valuation
and the Interpretability of Features. In Karimi, S.; Samiian, V.; and
Wilkins, W. K. (eds.), Phrasal and Clausal Architecture: Syntactic
Derivation and Interpretation. In Honor of Joseph E. Emonds, 262-
294. Amsterdam and Philadelphia: John Benjamins Publishing
Company.
Petersen, De Oliveira Almeida Maria Carolina. 2016. On Experiencers
and Minimality. PhD Dissertation. Maryland: University of Maryland,
College Park.
Pinker, Steven. 1989. Learnability and Cognition. Cambridge, MA: MIT
Press.
PiĔón, Christopher. 1997. Achievements in an event semantics. In
Lawson, A. (ed.), Proceedings of Semantics and Linguistic Theory VII,
273-296. Ithaca, NY: CLC Publications.
Plutchik, Robert. 1980. Emotion: A Psychoevolutionary Synthesis. New
York: Harper and Row Publishers.
Plutchik, Robert. 2001. Nature of emotions. American Scientist 89 (4),
344-350.
Popiel, Stephen J.; and McRae, Ken. 1988. The figurative and literal
senses of idioms, or all idioms are not used equally. Journal of
Psycholinguistic Research 17 (6), 475-487.
Postal, Paul. 1971. Crossover Phenomena. New York: Holt, Reinhart and
Winston.
Postal, Paul. 1974. On Raising: One Rule of English Grammar and Its
Theoretical Implications. Cambridge, MA: MIT Press.
Primus, Beatrice. 1999. Cases and Thematic Roles: Ergative, Accusative
and Active. Tübingen: Niemeyer.
Pulman, Stephen. G. 1993. The recognition and interpretation of idioms.
In Cacciari, C.; and Tabossi, P. (eds.) Idioms: Processing, Structure
and Interpretation, 249-270. Hillsdale, NJ: Lawrence Erlbaum
Associates.
Pustejovsky, James. 1991. The syntax of event structure. Cognition 41,
47-81.
Pylkkänen, Liina. 1997. Stage and Individual-Level Psych Verbs in
Finnish. Paper presented in the Workshop on Events as Grammatical
Objects. Ithaca, NY: Cornell University.
Pylkkänen, Liina. 2000. On Stativity and Causation. In Tenny, C. L.; and
Pustejovsky, J. (eds.), Events as Grammatical Objects: The
Converging Perspective of Lexical Semantics and Syntax, 417-444.
Pylkkänen, Liina. 2008. Introducing Arguments. Cambridge, MA: MIT
Press.
Punske, Jeffrey; and Stone, Schildmier Megan. 2015. Inner Aspect and
the Verbal Typology of Idioms. Presented at The 8th Brussels
Conference on Generative Linguistics.
Radford, Andrew. 1988a. Small Children’s Small Clauses. Transactions
of the Philological Society 86, 1-43.
Radford, Andrew. 1988b. Transformational Grammar: A First Course.
Radford, Andrew. 2000. T-contraction in a phase-based theory of
grammar. Essex Research Reports in Linguistics 33, 21-32.
Radford, Andrew. 2004. Minimalist Syntax: Exploring the Structure of
English. Cambridge: Cambridge University Press.
Ramchand, Gillian C. 1997. Aspect and predication. Oxford: Clarendon
Press.
Ramchand, Gillian. 2008. Verb Meaning and the Lexicon: A First Phase
Syntax. Cambridge: Cambridge University Press.
346 References
Rappaport Hovav, Malka; and Levin, Beth. 1988. What to Do with

Theta-Roles. In Wilkins, W. (ed.), Syntax and Semantics 21: Thematic
Relations, 7-36. New York: Academic Press.
Rappaport Hovav, Malka; and Levin, Beth. 1998. Building verb
meanings. In Butt, M.; and Geuder, W. (eds.), The Projection of
Arguments: Lexical and Compositional Factors, 97-134. Stanford, CA:
CSLI Publications.
Rayner, Keith; and Pollatsek, Alexander. 1989. The Psychology of
Reading. Englewood Cliffs: Prentice-Hall.
Reagan, Robert. T. 1987. The syntax of English idioms: can the dog be
put on? Journal of Psycholinguistic Research 16 (5), 417-441.
Reeve, Matthew. 2010. Clefts. London: University College London.
Reinhart, Tanya. 1996. Syntactic Effects of Lexical Operations:
Reflexives and Unaccusatives. OTS Working Papers in Linguistics.
Utrecht: University of Utrecht.
Reinhart, Tanya. 2002. The Theta System – An overview. Theoretical
Linguistics 28, 229-290.
Richards, Marc. 2008. On Feature Inheritance: An Argument from the
Phase Impenetrability Condition. Linguistic Inquiry 38, 563-572.
Richards, Marc. 2011. Deriving the Edge: What’s in a Phase? Syntax 14,
74-95.
Roberts, Ian. 1991. NP-movement, Crossover and Chain Formation. In
Haider, H.; and Netter, K. (eds.), Representation and Derivation in the
Theory of Grammar, 17-52. Dordrecht: Kluwer.
Roelofs, Ardi. 1992. A spreading-activation theory of lemma retrieval in
speaking. Cognition 42, 107-142.
Roelofs, Ardi. 2000. Word meanings and concepts: what do the findings
from aphasia and language specificity really say? Bilingualism:
Language and Cognition 3, 25-27.
Rosen, Carol. 1984. The interface between semantic roles and initial
grammatical relations. In Perlmutter, D.; and Rosen, C. (eds.), Studies
in Relational Grammar 2, 38-77. Chicago, IL: University of Chicago
Press.
Rosen, Sara T. 1989. Two types of noun incorporation: A lexical
analysis. Language 65 (2), 294-317.
Rosenbaum, Peter, S. 1967. The Grammar of English Predicate
Complement Constructions. Cambridge, MA: MIT Press.
Rothmayr, Antonia. 2009. Structure of Stative Verbs. Amsterdam: John
Rothstein, Susan. 2004. Structuring Events: an Essay on the Semantics of
Lexical Aspect. Oxford: Blackwell.
Rothstein, Susan. 2008. Two puzzles for a theory of lexical aspect:

semelfactives and degree achievements. In Dölling, J.; Heyde-
Zybatow, T.; and Schäfer, M. (eds.), Event Structures in Linguistic
Form and Interpretation, 175-198. Berlin, New York: Mouton de
Gruyter.
Rott, Julian A.; and Verhoeven, Elisabeth. 2016. On the typology of the
psych-alternation: Does morphological directionality influence syntax?
Paper presented at the 4th LingBaW Conference 2016, Lublin.
Rozwadowska, BoĪena. 1992. Thematic constraints on selected
constructions in English and Polish. Wrocáaw: Wydawnictwo
Uniwersytetu Wrocáawskiego.
Rozwadowska, BoĪena. 2005. A new perspective on event participants in
psychological states and events. In Broekhuis, H.; Corver, N.; Riny,
H.; Kleinhenz, U.; and Koster, J. (eds.), Organizing Grammar :
Linguistic Studies in Honor of Henk van Riemsdijk (Studies in
Generative Grammar 86), 563-571. Berlin, New York: Mouton de
Gruyter.
Rozwadowska, BoĪena. 2012. On the onset of psych eventualities. In
Cyran, E.; Kardela, H.; and Szymanek, B. Sound, Structure and Sense.
Studies in Memory of Edmund Gussmann, 533-554. Lublin:
Wydawnictwo KUL.
Rozwadowska, BoĪena; and Willim, Ewa. 2016. The puzzle of
psychological verbs: (why) are emotion verbs grammatically special?
In KuĨniak, M.; Rozwadowska B.; and Szawerna, M. (eds.), From
Motion to Emotion: Aspects of Psychical and Cultural Embodiment in
Language, 11-27. Frankfurt am Main: Peter Lang.
Runosalo, Pia. 2005. Idioms in Textbooks: A Comparative Study of Two
Upper Secondary School Textbook Series. Unpublished MA Thesis.
Tampere: The University of Tampere.
Russell, James A. 1991. Culture and the categorization of emotions.
Psychological Bulletin 110 (3), 426-450.
Ruwet, Nicolas. 1972. Théorie syntaxique et syntaxe du français. Paris:
Editions du Seuil.
Ruwet, Nicolas. 1991. Syntax and Human Experience. Chicago, IL:
University of Chicago Press.
Ruwet, Nicolas. 1993. Les verbes dits psychologiques: trois theories et
quelques questions. Recherchers linguistiques de Vincennes 22, 95-
124.
Ryle, Gilbert. 1949. The Concept of Mind. Chicago, IL: The University of
Chicago Press.
348 References
Saberian, Noorolhoda. 2011a. Idiom taxonomies and idiom

comprehension: implications for English teachers. Theory and Practice
in Language Studies 1 (9), 1231-1235.
Saberian, Noorolhoda. 2011b. The Effect of Idiom Type and Individual
Differences on Idiom Comprehension and Strategy Use of EFL
Students. Unpublished MA Thesis. Najafabad: Islamic Azad University.
Safir, Ken. 1983. On Small Clauses as Constituents. Linguistic Inquiry
14, 730-735.
Sag, Ivan A.; Baldwin, Timothy; Bond, Francis; Copestake, Ann; and
Flickinger, Dan. 2002. Multiword expressions: a pain in the neck for
NLP. In Gelbukh, A. (ed.), Computational Linguistics and Intelligent
Text Processing, Lecture Notes in Computer Science Vol. 2276, 1-15.
Berlin: Springer.
Schachter, Paul. 1973. Focus and Relativization. Language 49 (1), 19-46.
Schenk, André. 1994. Idioms and Collocations in Compositional
Grammars. PhD Dissertation. Utrecht: University of Utrecht.
Schenk, André. 1995. The Syntactic Behavior of Idioms. In Everaert, M.;
van der Linden, E.-J.; Schenk, A.; and Schreuder, R. (eds.), Idioms:
Structural and Psychological Perspectives, 253-272. Mahwah, NJ:
Lawrence.
Scherer, Klaus R. 2000. Psychological modelsofemotion. In Borod, J.
(ed.), The Neuropsychology of Emotion, 137-162. Oxford: Oxford
University Press.
Scherer, Klaus R. 2005. What are emotions? And how can they be
measured? Social Science Information 44 (4), 693-727.
Schmid, Hans-Jörg. 2010. Does frequency in text instantiate
entrenchment in the cognitive system? In Glynn, D.; and Fischer, K.
(eds.), Quantitative Methods in Cognitive Semantics: Corpus-Driven
Approaches, 101-133. Berlin, New York: Mouton de Gruyter.
Schmitt, Norbert; and Schmitt, Diane. 2014. A reassessment of
frequency and vocabulary size in L2 vocabulary teaching. Language
Teaching 47, 484-503.
Schraw, Gregory; Trathen, Woodrow; Reynolds, Ralph E.; and
Lapan, Richard T. 1988. Preferences for idioms: Restrictions due to
lexicalization and familiarity. Journal of Psycholinguistic Research 17
(5), 413-424.
Schwanenfluegel, Paula J.; and Shoben, Edward J. 1985. The influence
of sentence constraint on the scope of facilitation for upcoming words.
Schweigert, Wendy A. 1986. The comprehension of familiar and less
familiar idioms. Journal of Psycholinguistic Research 15, 33-45.
Schweigert, Wendy A.; and Moates, Danny R. 1988. Familiar idiom

comprehension. Journal of Psycholinguistic Research 17, 281-296.
Seidl, Jennifer; and McMordie, W. 1978. English Idioms and How to
Use Them (4th ed.). Oxford: Oxford University Press.
Shakespeare, William. As You Like It. Retrieved from
http://shakespeare.mit.edu/ asyoulikeit/full.html
Shaver, Phillip R.; Wu, Shelley; and Schwartz, Judith C. 1992. Cross-
cultural similarities and differences in emotion and its representation.
In Clark, M. S. (ed.), Emotion. Review of Personality and Social
Psychology (13), 175-212. Thousand Oaks, CA: Sage Publications.
Shi, Ziqiang. 1988. The Present of Past of the Particle ‘le’ in Mandarin
Chinese. PhD Dissertation. University of Pennsylvania.
Siemund, Peter. 2013. Varieties of English: A Typological Approach.
Smith, Carlota S. 1991. The Parameter of Aspect. Dordrecht: Kluwer.
Sonnenhauser, Barbara. 2010. Die Diskursfunktionen des draifachen
Artikel im Makedonischen: Perspektivität und Polyphonie. Die Welt
der Slaven 55/2, 334-359.
Sornig, Karl. 1988. Idioms in language teaching. In Hüllen, W.; and
Schulze, R. (eds.), Understanding the Lexicon, 280-290. Tübingen:
Max Niedermayer Verlag.
Speas, Margaret. 1990. Comments on Papers by James Gair, Yamura
Kachru and K.P. and Tara Mohanan. In Verma, M. K.; and Mohanan,
K. P. (eds.), Experiencer Subjects in South Asian Languages, 77-83.
Sprenger, Simone A.; Levelt, Willem J. M.; and Kempen, Gerard.
2006. Lexical access during the production of idiomatic phrases.
Stalmaszczyk, Piotr. 1996. Theta roles and the theory of theta-binding.
Papers and Studies in Contrastive Linguistics 31, 97-110.
Stanojeviü, Maja. 2009. Cognitive synonymy: a general overview. Facta
Universitatis, Linguistics and Literature series 7 (2), 193-200.
Starke, Michal. 1995. On the Format for Small Clauses. In Cardinaletti,
A.; and Guasti, M. (eds.), Syntax and Semantics 28: Small Clauses,
237-269. New York: Academic Press.
Stock, Oliviero; Slack, Jon; and Ortony, Andrew. 1993. Building
castles in the air: Some computational and theoretical issues in idiom
comprehension. In Cacciari, C.; and Tabossi, P. (eds.), Idioms:
Processing, Structure and Interpretation, 229-248. Hillsdale, NJ:
350 References
Stone, Schildmier Megan. 2008. Passivizable and nonpassivizable

idioms: A minimalist approach. Ms., Arizona: University of Arizona.
Stone, Schildmier Megan. 2009. Idioms and Domains of Interpretation.
Presentation at Arizona Linguistics Circle 3. Tucson, Arizona.
Stone, Schildmier Megan. 2013. Can idioms be passivized?: Evidence
from online processing. Coyote Papers 21, 1-22.
Stone, Schildmier Megan. 2016. The Difference Between Bucket-Kicking
and Kicking the Bucket: Understanding Idiom Flexibility. PhD
Dissertation. Arizona: The University of Arizona.
Stowell, Timothy A. 1981. Origins of Phrase Structure. PhD Dissertation.
Cambridge, MA: MIT.
Stowell, Timothy A. 1983. Subjects across categories. The Linguistic
Review 2, 285-312.
Stowell, Timothy. 1986. Psych-Movement in the Mapping Between D
Structure and LF. Paper presented at GLOW 9. Utrecht: Utrecht
University.
Stowell, Timothy A. 1991. Determiners in NP and DP. In Leffel, K.; and
Bouchard, D. (eds.), Views on Phrase Structure, 37-56. Dordrecht:
Kluwer.
Strassler, Jurg. 1982. Idioms in English. A Pragmatic Analysis.
Tübingen: Gunter Narr Verlag.
Suþková, Magda. 2010. A Comparative study of English and Czech
Idioms: Plant Idioms. MA Thesis. Brno: Masaryk University.
Svenonius, Peter. 2001. On object shift, scrambling, and the PIC. In
Guerzoni, E.; and Matushansky, O. (eds.), A Few from Building E39:
Working Papers in Syntax, Semantics, and their Interface. Working
Papers in Linguistics MITWPL, 267-289. Cambridge, MA: MIT Press.
Svenonius, Peter. 2004. On the edge. In Adger, D.; Cat C. de; and
Tsoulas, G. (eds.), Peripheries: Syntactic Edges and Their Effects,
261-287. Dordrecht: Kluwer.
Svenonius, Peter. 2005. Extending the extension condition to
discontinuous idioms. Linguistic Variation Yearbook 5, 227-263.
Svenonius, Peter. 2007. The Position of Adjectives and Other Phrasal
Modifiers in the Decompositionof DP. In McNally, L.; and Kennedy,
C. (eds.), Adjectives and Adverbs: Syntax, Semantics, and Discourse,
16-42. Oxford: Oxford University Press.
Swinney, David A.; and Cutler, Anne. 1979. The access and processing
of idiomatic expressions. Journal of Verbal Learning and Verbal
Behavior 18, 523-534.
SzymaĔska, Izabela. 2008. How pragmatics meets morphology. A

Construction Grammar analysis of a Polish idiomatic pattern. Research
in Language 6, 109-154.
Tabossi, Patrizia; Arduino, Lisa; and Fanari, Rachele. 2011.
Descriptive norms for 245 Italian idiomatic expressions. Behav Res 43,
110-123.
Tabossi, Patrizia; and Zardon, Francesco. 1995. The activation of
idiomatic meaning. In Everaert, M.; van der Linden, E.-J.; Schenk, A.;
and Schreuder, R. (eds.), Idioms: Structural and psychological
perspectives, 273-282. Hillsdale, NJ: Lawrence Erlbaum Associates.
Tanaka, Tomoyuki; and Yokogoshi, Azusa. 2010. The Rise of a
Functional Category in Small Clauses. Studia Linguistica, 1-32.
Temme, Anne. 2014. German psych-adjectives. In Machicao y Priemer,
A.; Nolda, A.; and Sioupi, A. (eds.), Zwischen Kern und Peripherie,
131-156. Berlin: Mouton de Gruyter.
Tenny, Carol L. 1987. Grammaticalization, Aspect and Affectedness.
PhD Dissertation. Cambridge, MA: MIT Press.
Tenny, Carol L. 1992. The Aspectual Interface Hypothesis. In Sag, I.A.;
and Szabolcsi, A. (eds.), Lexical Matters, 1-27. Stanford, CA: CSLI
Publications.
Tenny, Carol L. 1994. Aspectual Roles and the Syntax-Semantics
Interface. Dordrecht: Kluwer.
Tenny, Carol. 1998. Psych-verbs and Verbal Passives in Pittsburghese.
Linguistics 36, 591-597.
Titone, Debra A.; and Connine, Cynthia M. 1994a. Comprehension of
idiomatic expressions: effects of predictability, and literalness. Journal
of Experimental Psychology: Learning, Memory, and Cognition 20 (3),
671-679.
Titone, Debra A.; and Connine, Cynthia M. 1994b. Descriptive norms
for 171 idiomatic expressions: familiarity, compositionality,
predictability and literalness. Metaphor and Symbolic Activity 9, 247-
270.
Titone, Debra A.; and Conine, Cynthia M. 1999. On the compositional
and noncompositional nature of idiomatic expressions. Journal of
Pragmatics 31 (12), 1655-1674.
Tomacsek, Vivien. 2014. Approaches to the Structure of English Small
Clauses. The Odd Yearbook 9, 128-154.
Trask, Robert. 1999. Key Concepts in Language and Linguistics.
Psychology Press, 27-28.
Travis, Lisa deMena. 1984. Parameters and Effects of Word Order
Variation. PhD Dissertation. Cambridge, MA: MIT Press.
352 References
Turner, Jonathan, H. 1999. Towards a general sociological theory of

emotions. Journal of the Theory of Social Behavior 29, 133-162.
Turner, Jonathan, H. 2007. Human Emotions: A Sociological Theory.
London: Routledge.
Turner, Jonathan, H. 2009. The sociology of emotion: Basic Theoretical
arguments. Emotion Review 1 (4), 340-354.
van der Linden, Erik-Jan. 1991. A Categorical Computational Theory of
Idioms (Ms.), Tilburg: Tilburg University.
van der Linden, Erik-Jan. 1992. Incremental processing and the
hierarchical lexicon. Computational Linguistics 18, 219-238.
van Valin, Robert D. 1990. Semantic parameters of split intransitivity.
Language 66 (2), 221-260.
van Valin, Robert D.; and LaPolla, Randy J. 1997. Syntax. Structure,
Meaning and Function. Cambridge: Cambridge University Press.
van Voorst, Jan. 1992. The Aspectual Semantics of Psychological Verbs.
Linguistics and Philosophy 15 (1), 65-92.
Vega-Moreno, Rosa Elena. 2003. Relevance Theory and the construction
of idiom meaning. UCL Working Papers in Linguistics 15, 83-104.
Vega-Moreno, Rosa Elena. 2005. Idioms, transparency, and pragmatic
inference. UCL Working Papers in Linguistics 17, 389-425.
Vendler, Zeno. 1957. Verbs and times. Philosophical Review 56, 143-
160.
Vendler, Zeno. 1967. Verbs and Times. In Vendler, Z. (ed.), Linguistics
in Philosophy, 97-121. Ithaca: Cornell University Press.
Verhoeven, Elisabeth. 2008. (Non-)canonical marking of experiencer
objects: A typological comparison of Chinese, Korean, Turkish, and
Modern Greek, Language Typology and Universals 61 (1), 81-92.
Verhoeven, Elisabeth. 2010. Agentivity and stativity in experiencer
verbs: Implications for a typology of verb classes. Linguistic Typology
14, 213-251.
Verkuyl, Henk J. 1972. On the Compositional Nature of the Aspects.
Dordrecht: Reidel.
Verkuyl, Henk J. 1993. A Theory of Aspectuality: The Interaction
Between Temporal and Atemporal Structure. Cambridge: Cambridge
University Press.
Vietri, Simonetta. 2014. Idiomatic Constructions in Italian. A Lexicon
Grammar Approach. Amsterdam; Philadelphia: John Benjamins
Publishing Company.
Wadepuhl, W. 1928. German idioms. The German Quarterly 1 (2), 68-
73.
Wasow, Thomas. 1977. Transformations and the lexicon. In Culicover, P.

W.; Wasow, T.; and Akmajian, A. (eds.), Formal Syntax, 327-360.
New York: Academic Press.
Wasow, Thomas; Sag, Ivan A.; and Nunberg, Geoffrey. 1983. Idioms:
an interim report. In Hattori, S.; and Inoue, K. (eds.), Proceedings of
the 13th international congress of linguistics, 102-115. Tokyo:
Gakushuin University.
Wasow, Thomas; Sag, Ivan A.; and Nunberg, Geoffrey. 1984. Idioms:
An interim report. In Hattori, S.; and Inoue, K. (eds.), Proceedings of
the 13th International Congress of Linguistics, 87-96. Tokyo: Nippon
ToshiCenter.
Weinreich, Uriel. 1969. Problems in the analysis of idioms: Lectures
delivered before the Linguistic Institute of the Linguistic Society of
America, University of Los Angeles, June 17 - August 12, 1966. In
Puhvel, J. (ed.), Substance and Structure of Language, 23-81.
Berkeley, CA: University of California Press.
Wierzbicka, Anna. 1972. Semantic Primitives. Frankfurt am Main:
Athenäum Verlag.
Wierzbicka, Anna. 1992. Defining Emotion Concepts. Cognitive Science
16, 539-581.
Wierzbicka, Anna. 1999. Emotions Across Languages and Cultures:
Diversity and Universals. Cambridge: Cambridge University Press.
Wierzbicka, Anna. 2009. Language and metalanguage: Key issues in
emotion research. Emotion Review 1, 3-14.
Wilensky, Robert; and Arens, Yigal. 1980. PHRAN: A Knowledge-
Based Approach to Natural Language Analysis (Memorandum
UCB/ERL M80/34). Berkeley, CA: University of California.
Willey, Norman L. 1939. German idioms. The Modern Language Journal
24 (3), 221-226.
Williams, Edwin. 1980. Predication. Linguistic Inquiry 11, 203-238.
Williams, Edwin. 1981. Argument Structure and Morphology. The
Linguistic Review 1, 81-114.
Williams, Edwin. 1983. Against Small Clauses. Linguistic Inquiry 18,
287-308.
Willim, Ewa. 2012. On the feature valuation/interpretability biconditional
in Minimalist Theory. The case of (un)interpretable gender. In Cyran,
E.; Kardela, H.; and Szymanek, B. (eds.), Sound, Structure and Sense.
Studies in Memory of Edmund Gussmann, 761-806. Lublin:
Wydawnictwo KUL.
WitkoĞ, Jacek; and Dziemianko, Anna. 2006. On the syntax of idioms
and the Idiomatic Constituency Axiom. In Dziubalska-Koáaczyk, K.
354 References
(ed.), IFAtuation: A Life in IFA. A Festschrift for Professor Jacek

Fisiak on the Occasion of His 70th Birthday, 773-794. PoznaĔ:
Wydawnictwo Naukowe UAM.
Wood, Mary McGee. 1986. A Definition of Idiom. Bloomington: Indiana
University Linguistics Club.
Wundt, Wilhelm M. 1900. Die Sprache [Language] (2nd ed). Leipzig:
Engelmann.
Yadroff, Michael. 1999. Formal Properties of Functional Categories:
The Minimalist Syntax of Russian Nominal and Prepositional
Expressions. PhD Dissertation. Bloomington, IN: University of Indiana.
Yoshikawa, Hiroshi. 2008. International intelligibility in world Englishes:
focusing on idiomatic expressions. International Communication Studies
XVII (4), 219-226.
Zhu, Feng; and Fellbaum, Christiane. 2015. Quantifying Fixedness and
Compositionality in Chinese Idioms. International Journal of
Lexicography 28 (3), 338-350.
Dictionaries and Thesauri

American Heritage Dictionary of the English Language, 5th ed. 2016.
Boston: Houghton Mifflin Harcourt Publishing Company. Retrieved
from https://www.ahdictionary.com/word/search. html?q=online
American Heritage Idioms Dictionary, 3rd ed. 2002. Boston: Houghton
Mifflin Harcourt Publishing Company. Retrieved from
http://www.dictionary.com
Collins Thesaurus of the English Language – Complete and Unabridged
2nd Edition. 2002. HarperCollins Publishers 1995, 2002. Retrieved
from http://www.freethe saurus.com
Google Dictionary. Retrieved from https://docs.google.com/ document
Hill Long, Thomas. 1990. Longman Dictionary of English Idioms.
London: Longman.
Kjellmer, Goran. 1994. A Dictionary of English Collocations, Based on
the Brown Corpus. Oxford: Pergamon Press.
Longman Dictionary. Retrieved from
http://www.ldoceonline. com/dictionary
Longman Dictionary of English Idioms. 1979. Vol 2. London: Longman
Group Ltd.
Macmillan Dictionary. Retrieved from
http://www.macmillan dictionary.com
Merriam-Webster. 2004. The Merriam-Webster Collegiate Dictionary
(11th ed.) Springfield, MA: Merriam. Retrieved from
https://www.merriam-webster.com/
O’Dell, Felicity; and McCarthy, Michael. 2010. English Idioms in Use
Advanced. Cambridge: Cambridge University Press.
Online Etymology Dictionary. Retrieved from
http://www.etymonline.com
Online International Dictionary. Retrieved from http://idict.org
Online Oxford English Dictionary. Retrieved from
https://en.oxforddictionaries.com
Phrases and their origins. Retrieved from
http://www.phrases. org.uk/meanings/1184 00.html
Power Thesaurus. 1969. Retrieved from http://www.power thesaurus.org
Seidl, Jennifer; and McMordie, W. 1978. English Idioms and How to
Use Them (4th ed.). Oxford: Oxford University Press.
Simpson, John A.; and Weiner, E. S. C. (eds.) 1989. The Oxford English
Dictionary on CD-ROM (2nd ed.). Oxford: Oxford University Press.
Wielki multimedialny sáownik angielsko-polski i polsko-angielski [Great
Multimedia English-Polish and Polish-English Dictionary]. 2005.
Oxford: PWN.
WordNet 3.0, Farlex Clipart Collection. 2012. Princeton: Princeton
University, Farlex Inc. Retrieved from http://www.free thesaurus.com

A Syntactic Study

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A Syntactic Study

Uploaded by

Copyright:

Available Formats

A Syntactic Study

Psychological States in English

This book first published 2018

Cambridge Scholars Publishing

Lady Stephenson Library, Newcastle upon Tyne, NE6 2PA, UK

British Library Cataloguing in Publication Data

Copyright © 2018 by Anna Dąbrowska

ISBN (10): 1-5275-0616-9

List of Abbreviations .................................................................................. xi

List of Tables ............................................................................................ xiii

List of Charts ............................................................................................. xv

Chapter One ................................................................................................. 7

Chapter Two .............................................................................................. 73

2.4.1 Basic event categories ........................................................... 81

Chapter Three ......................................................................................... 129

Chapter Four .......................................................................................... 171

4.3.3 Co-occurrence dependencies in psychological ICEs ........... 214

Summary and Conclusions ...................................................................... 293

Appendix 1 .............................................................................................. 299

Appendix 2 .............................................................................................. 305

Appendix 3 .............................................................................................. 317

References ............................................................................................... 319

First and foremost, I would like to express my heartfelt gratitude to

Besides, my sincere thanks go both to Professor BoĪena Cetnarowska and

Finally, on a more personal note, I owe a special debt of gratitude to those

All shortcomings in this work are, of course, my own responsibility.

Table 1-1. Types of lexemic idioms (Makkai 1972: 135-169)

Table 3-4. Types of syntactic patterns for OE (class II) psychological

Figure 1-1. Sample sentences with superficial phrase-structure trees

In the field of English phraseology, linguists have shown a constant

For the purpose of the book, the recent compositional model of

(0.1) Belletti and Rizzi’s (1988) tripartite classification of psych-verbs:

As shown in (0.1), an Experiencer can be realized as either a subject

(0.2) The examples of idioms and the SE (class I) psych-verbs they

put Y through wringer

Importantly, both the psych-predicates and the psychological VP-

encyclopaedias, and put forward by linguists (e.g. Pulman 1993; Gläser

which are to be analysed in Chapter Four. To meet these objectives, the

TOWARDS DEFINING AN IDIOM

1.2 The definition of an idiom

1.2.1 Dictionaries and linguistic encyclopaedias as the sources

What is more, in his Longman Dictionary of Idioms, Hill (1990)

1.2.2 Linguists’ and scientists’ definitions of an idiom

Idiom or idiomatic phrase (...) is a phrase the meaning of which cannot be

In addition, for Weinreich (1969), “any expression in which at least

one of the core characteristics of idioms since “the more anomalies a

distinguish between substantive (lexically filled) and formal (lexically

structure is derivationally non-compositional. Moreover, it is considerably

phraseological units. Hence, some idiomatic phraseological units have

1.3 The characteristics of idioms

1.3.1 Metaphoricity / figurativeness

1.3.1.1 The notion of metaphor and figurative language

(1.1) a. Bill kicked the bucket. [Bill died]

f. They shall beat their swords into plowshares.

Referring to the examples in (1.1) and (1.2), Horn (2003: 245-247)

(1.3) a. *The bucket was kicked by Bill.1

(1.4) a. The beans were spilled by Fred.

The mobility of VP idioms is correlated by Jackendoff (1997) with a

to its NP arguments, Horn (2003) further argues that the property of

1.3.1.2 Various taxonomies of idioms

Idioms have been classified in multiple ways by different researchers

A. Makkai’s (1972) taxonomy

Types of lexemic idioms Example Meaning

Table 1-1. Types of lexemic idioms (Makkai 1972: 135-169)