Webelhuth et al: Idioms in LInguistics Theory, Part I

The Linguistic Status of Idioms
Part I: The Empirical Domain and Previous Approaches
Gert Webelhuth, Manfred Sailer, Sascha Bargmann
University of Frankfurt
Minerva Summer School, 2013
Webelhuth/Sailer/Bargmann (Ffm) Idioms 1 2013 1 / 39

Introduction and Outline
Outline of the course
Part 1
1 Characterizing the phenomenon
2 Idioms in Generative Grammar
3 Decomposable vs. non-decomposable idioms
4 Theory 1: An inference-based theory
5 Theory 2: A constructional theory
6 Theory 3: A denotational theory
7 Summary of part 1
Part 2
8 Idioms and collocations
9 The lexical identiﬁer (LID) theory
10 The LF theory
11 Summary

What is an idiom?
Idiom: phraseologism, phraseme, phraseological unit, multiword expression,
. . .
Prototypical properties:
phrasal
idiomatic: non-literal meaning; holistic meaning
ﬁxed: words cannot be exchanged; restricted syntactic ﬂexibility
lexicalized: conventionalized combination; represented as one unit

Some examples
(1) kick the bucket (‘die’)
a. idiomatic: ok
b. lexically ﬁxed: = kick the pail; = throw the bucket
c. syntactically ﬁxed: *The bucket was kicked.
d. lexicalized: ok

Some examples cont.
(2) spill the beans (‘reveal information’)
a. idiomatic: ok
b. lexically ﬁxed: = spilled the pulse; = sling down the beans
c. syntactically ﬁxed?:
The beans were spilled.
The beans appeared to be spilled.
* The beans, Pat spilled.
d. lexicalized: ok

Some examples cont.
(3) make headway (‘make progress’)
a. idiomatic: no? (cranberry word/bound word)
b. lexically ﬁxed: ??achieve headway
c. syntactically ﬁxed?
Considerable headway was made.
How much headway did they make on the job?
*That much headway I’m sure they made on the job. (Postal, 1998,
p. 31)
d. lexicalized: ok

Some examples cont.
(4) brush one’s teeth (‘clean one’s teeth’)
a. idiomatic: no? (collocation, idiom of encoding)
b. lexically ﬁxed?
I brushed my choppers.
I cleaned/polished my teeth
c. syntactically ﬁxed?
The teeth were brushed.
Those teeth he hadn’t brushed in ages.
d. lexicalized?

Idioms in Generative Grammar Historical overview
Phrasal lexical entries in Chomsky (1965)
Consider, for example, such phrases as ‘take for granted’, which abound in
English. From a semantic and distributional point of view, this phrase seems
to be a single lexical item, and it therefore must be entered in the lexicon as
such, with its unique syntactic and semantic features. On the other hand its
behavior with respect to transformations and morphological processes
obviously shows that it is some sort of Verb-with-Complement construction.
Once again we have a lexical item with a rich internal structure (Chomsky,
1965, p. 190)

Historical overview
Chafe (1968): Four problems of idioms:
◮ non-compositional
◮ transformationally defective
◮ (sometimes) syntactically ill-formed
◮ idiomatic reading of a combination is more frequent than literal meaning.
Weinreich (1969):
◮ Phrasal lexical entry lists all possible transformations.
Fraser (1970):
◮ Idioms inserted with structure in D-Structure
◮ Classification according to syntactic flexibility.
Jackendoff (1975): Phrasal lexical entries with only partial specification,
for syntactically regular idioms: structure follows from syntactic rules as
lexical redundancy rule.

Idiom arguments in Principles and Parameters
(Nunberg et al., 1994)
Idiom inserted en bloc at D-Structure
Transformations apply to DS trees, even if of idiomatic origin.
More recently: Compositional aspects of idioms used to motivate
functional projections (X gave Y the boot — Y got the boot from X,
Richards (2001))
Predictions:
◮ Idioms have a regular syntactic structure.
◮ Idioms can have only canonical form, or canonical and transformed form; but
never: only transformed form
◮ Only the idiom as a whole has a meaning, idiom parts are not assigned
meaning.

Idioms in Generative Grammar Problems for the generative approaches
Important publications to change our view on idioms
Higgins (1974): Critique of en bloc insertion, attempt for a more semantic
theory; unpublished
Ernst (1981): Modiﬁers inside idioms as argument against monolithic
semantics of idioms
McCawley (1981): Paradoxical predictions for idioms in relative clauses
Wasow et al. (1983); Nunberg et al. (1994): Two classes of idioms
distinguished by decomposability (also: Langacker (1987))
Ruwet (1991): List of arguments against the traditional en bloc insertion
view

Arguments: Regular syntactic shape
Chafe (1968); Nunberg et al. (1994):
(5) trip the light fantastic (‘dance’)
(6) kingdom come (‘paradise’)
(7) easy come easy go

Arguments: No “transformed-only” idioms
Nunberg et al. (1994):
(8) passive only: (be) cast in stone
(9) Wh-moved only: what the hell
(10) inverted only: Is the pope catholic?
(11) imperative only: Break a leg!

Arguments: Idiom parts are meaningless
Modification (Ernst, 1981)
(12) External modification:
a. Pat kicked the social bucket. (= Socially Pat kicked the bucket.)
b. Pat pulled some economic strings. (= Pat pulled some strings in
economy.)
(13) Internal modification:
a. Katz and I had by then become good friends, having long before
buried the old hatched (L. Melamed, Escape to the Future)
b. My girls should’ve buried the damn hatchet when they were in
their prime. (www; expressive modifier)
c. Pat pulled some important strings. (= Pat used some important
connections.)
The existence of internal modification readings is strong evidence that idiom
parts can be meaningful.

Arguments: Idiom parts are meaningless
Determiner variation:
(14) Pat kicked the/*a bucket.
(15) I have buried many hatchets with my parents but this still burns me up.
(www)
(16) Pat pulled the/many strings
Determiner variation supports the observations on modiﬁcation.

Additional problem: McCawley’s transformational
paradox
If the idiom pull strings must be inserted as one VP unit from the lexicon,
there is a paradox:
(17) The strings that Pat pulled got Chris the job.
bad if strings originates in the surface position
ok if strings originates inside the relative clause
(18) Pat pulled the necessary strings that got Chris the job.
ok if strings originates in the surface position
bad if strings originates inside the relative clause

Two classes of idioms
Wasow et al. (1983); Nunberg et al. (1994): decomposability
Idiomatically combining expressions (ICE): spill the beans, keep tabs on
s.o., make headway, bury the hatchet
◮ idiom parts can occur in positions/constructions that require content
◮ for example: internal modification
→ expect: syntactic flexibility
Idiomatic phrases (IPh): kick the bucket, saw logs (‘snore/sleep’), trip the
light fantastic (‘dance’)
◮ idiom parts cannot occur in positions/constructions that require content
◮ for example: no internal modification
→ less/no syntactic flexibility

Tests for ICEs
If an idiom part can occur in a position/construction that must have some
meaning, the idiom is decomposable.
Internal modiﬁcation possible
Determiner change possible

Tests for ICEs cont.
Fronting possible:
(19) The strings Pat has pulled.
(20) * The bucket Pat has kicked.
Pronominalization possible:
(21) Eventually they spilled the beans, but they didn’t spill them
deliberately.
(22) Kim’s family pulled some strings on her behalf, but they weren’t
enough to get her the job. (Nunberg et al., 1994)
(23) * Pat kicked the bucket and Chris kicked it too.
(24) * Pat tripped the light fantastic but Alex didn’t want to trip it.

Tests for ICEs cont.
Relative clause:
(25) Partially inside a RelC:
The strings that Pat pulled got Chris the job.
*The bucket that Pat kicked was unexpected.
(26) Internal modiﬁcation by a RelC:
Pat pulled the strings that got Chris the job
*Pat kicked the bucket that nobody expected.
often also considered: Passive, raising possible:
(27) The strings have been pulled.
(28) * The bucket has been kicked.

Decomposability problematic/circular?
Decomposability is taken as a purely semantic notion. Not to be confused
with:
= transparency of the expression as a whole:
saw logs (‘snore’) (transparent, non-decomposable)
spill the beans (‘divulge information’ (non-transparent, decomposable)
shoot the breeze (‘chat’) (non-transparent, non-decomposable)
= plausible paraphrasability:
kick the bucket = end one’s life (non-decomposable)

Example Analyses
Two classes
Decomposability is defined via semantic flexibility criteria.
An expression that meets some of these criteria is decomposable, all
others are non-decomposable.
Nunberg et al. (1994) see a strong connection between semantic
decomposability and syntactic flexibility. The relation might be looser
(Webelhuth and Ackermann, 1994).

Example Analyses
Aims of a formal analysis
What we want:
Varying syntactic ﬂexibility
Semantics of the well-formed strings
What we won’t talk about:
Relation between the literal and the non-literal meaning
Cognitive basis of idioms
Word play
Text-constituting potential of idioms

Example Analyses
Examples of formal analyses
Pulman (1993): Inference-based analysis
Abeillé (1995): Constructional analysis
Gazdar et al. (1985): Denotational analysis

Example Analyses Inference-based Analysis
Inference-based analysis: Sketch
Representatives: Pulman (1993), Egan (2008)
Literal parse mapped to idiomatic interpretation:
◮ Pulman (1993): sem.repr. → sem.repr. (special inference rules)
◮ example: The y [bucket’(y)](kick’(x,y)) → die’(x)
(applies if the literal reading is inconsistent in the context)
Syntax non-holistic, meaning holistic
Idiom is stored as a special inference rule, different from lexical entries.

Inference-based analysis: Strengths
no idiomatic words necessary
literal meaning available; necessary for “extended uses”
(29) If you let this cat out of the bag, a lot of people are going to get
scratched.
possibly: relation to other cases of ﬁgurative language

Inference-based account of idiom properties
Idiomaticity: mapping between lexical and idiomatic reading
Lexical fixedness: inference rule can rely on word-specific semantic
contributions
Semantic fixedness: possible, if syntactic structure correlates with
different semantic representation

Inference-based analysis: Problems
Problems (Wearing, 2012)
◮ processing: idiomatic sense sometimes faster than literal sense.
◮ vague predictions on degree of syntactic ﬂexibility:
(30) Jane had a bone to pick with Susan, and Anne had one to pick
with Ian.
(have a bone to pick with s.o. (‘X has s.th. to discuss where Y
annoyed X’)
(31) * Tony shot the breeze with Junior, and Paulie shot it with Silvio.
(shoot the breeze (‘chat’))
Other problems
◮ Idioms with bound words? (make headway, the whole (kit and) caboodle
(‘everything’))
◮ idioms with syntactic peculiarities? (trip the light fantastic)
◮ Pulman (1993): type of inference required elsewhere?
◮ Egan (2008): admits possible stronger lexicalization for many idioms

Example Analyses Constructional Analysis
Constructional analysis: Sketch
Representative: Abeillé (1995), Tree Adjoining Grammar
Idiom is represented as a syntactic tree (elementary tree)
Nodes in the tree can, but need not have semantic annotation.
IPh:
S
sem: die’(x)
NP
sem: x
VP
sem: die’(x)
V
kick
NP
D
the
N
bucket

Constructional analysis: Sketch
Representative: Abeillé (1995), Tree Adjoining Grammar
Idiom is represented as a syntactic tree (elementary tree)
Nodes in the tree can, but need not have semantic annotation.
ICE:
S
sem: The y [info’(y)](reveal’(x,y))
NP
sem: x
VP
sem: λ x.The y [info’(y)](reveal’(x,y))
V
sem: reveal’
spill
NP
sem: λP.The y [info’(y)](P(y))
D
the
N
beans

Constructional approach: Flexibility
Transformations: each elementary tree belongs to a “tree family”, where
all possible derived trees are included (such as for passive etc.)
Modification: Possibility to mark in the structure whether modifiers are
possible.
Internal modification: available if attachment node has meaning
Pronominalization: unclear

Constructional approach: Strengths
Account of syntactically ill-formed idioms (trip the light fantastic), idioms
in transformed form only (Get lost!), or idioms with bound words ((make)
headway).
All idioms are represented as units.
Parts of an idiom can have an idiomatic meaning, but only if the rest of
the idiom is present.

Constructional account of idiom properties
Idiomaticity: done via ambiguity.
Lexical ﬁxedness: lexical items and word forms are included into the
elementary trees.
Syntactic ﬁxedness: via diacritic marking

Constructional approach: Problems
Marking for applicable transformations not grounded in semantics
Analysis of pronominalization not clear

Example Analyses Denotational Analysis
Denotational Approach: Sketch
Representatives: Gazdar et al. (1985)
Hybrid approach:
◮ Idiomatic phrases: fixed tree with meaning is in the lexicon
◮ ICE: co-occurrence of idiom parts by special denotations
Words in idioms are ambiguous:
◮ spill reveal-idiom’
◮ beans secret-idiom’
◮ Pat spilled the beans: The x [secret-idiom’(x)](reveal-idiom’(pat’,x))
semantic constants as partial functions:
[[reveal-idiom’]]([[beans’]]): undefined.
[[spill’]]([[secret-idiom’]]) undefined

Denotational Approach: Sketch cont.
Passive: The beans had been spilled.
The x [secret-idiom’(x)](∃y (reveal-idiom’)(y,x))
Strengths:
◮ attempt to encode Nunberg et al. (1994)
◮ internal modiﬁcation ok
◮ syntactic ﬂexibility related to semantics

Denotational account of idiom properties
Idiomaticity: by ambiguity
Lexical fixedness: via the denotation of special, lexeme-specific
predicate-symbols.
Syntactic fixedness: fixed tree (for IPh) vs. syntactically free combination
(for ICE).

Denotational account: Problems
Phrasal lexical entry for non-decomposable idioms not well deﬁned in
Gazdar et al. (1985)
Evidence for lexical ambiguity?
Complicated underlying denotations
Difference between various types of decomposable idioms?
(32) * The beans, they didn’t spill.
(33) The strings, they didn’t pull.

Summary
(At least) 3 types of idioms
1 Idiomatic phrases: Syntactically (almost) frozen idioms, kick the bucket
2 Idiomatically combining expressions: Mobile idioms
a Syntactically connected idioms, spill the beans
b Semantically connected idioms, pull strings

Summary
Most promising analysis
Non-decomposable idiom: as completely ﬁxed tree
Decomposable idiom: normal syntactic combination; semantic constants
rather than denotations.
On Thursday: Detailed look at three idioms and outline of such a theory.

Literatur
References
Abeillé, Anne (1995). The Flexibility of French Idioms: A Representation with Lexical Tree
Adjoining Grammar. In M. Everaert, E.-J. v. d. Linden, A. Schenk, and R. Schreuder (Eds.),
Idioms. Structural and Psychological Perspectives, pp. 15–42. Lawrence Erlbaum Associates,
Hillsdale.
Chafe, Wallace (1968). Idiomaticity as an Anomaly in the Chomskyan Paradigm. Foundations of
Language 4, 109–127.
Chomsky, Noam (1965). Aspects of the Theory of Syntax. Cambridge, Massachusetts: MIT
Press.
Egan, Andy (2008). Pretense for the Complete Idiom. Noûs 42(3), 381–409.
Ernst, Thomas (1981). Grist for the Linguistic Mill: Idioms and ‘Extra’ Adjectives. Journal of
Linguistic Research 1, 51–68.
Fraser, Bruce (1970). Idioms within a Transformational Grammar. Foundations of Language 6,
22–42.
Gazdar, Gerald, Klein, Ewan, Pullum, Geoffrey, and Sag, Ivan (1985). Generalized Phrase
Structure Grammar. Cambridge, Massachusetts: Harvard University Press.
Higgins, Francis Roger (1974). On the Use of Idioms as Evidence for Movement. A Cautionary
Note. Unpublished manuscript of a talk given at LSA 1974, New York.
Jackendoff, Ray (1975). Morphological and Semantic Regularities in the Lexicon.
Language 51(3), 639–671.
Langacker, Ronald W. (1987). Foundations of Ccognitive Grammar. Stanford: Stanford University
Press.

Summary
McCawley, James D. (1981). The Syntax and Semantics of English Relative Clauses. Lingua 53,
99–149.
Nunberg, Geoffrey, Sag, Ivan A., and Wasow, Thomas (1994). Idioms. Language 70, 491–538.
Postal, Paul M. (1998). Three Investigations of Extraction. Cambridge, Mass.: MIT Press.
Pulman, Stephen G. (1993). The Recognition and Interpretation of Idioms. In C. Cacciari and
P. Tabossi (Eds.), Idioms: Processing, Structure, and Interpretation, Chapter 11, pp. 249–270.
Hillsdale, New Jersey: Lawrence Erlbaum Associates.
Richards, Norvin (2001). An Idiomatic Argument for Lexical Decomposition. Linguistic
Inquiry 32(1), 183–192.
Ruwet, Nicolas (1991). On the Use and Abuse of Idioms in Syntactic Argumentation. In Syntax
and Human Experience, pp. 171–251. Chicago, London: University of Chicago Press. Edited
and translated by John Goldsmith.
Wasow, Thomas, Sag, Ivan A., and Nunberg, Geoffrey (1983). Idioms: An Interim Report. In
S. Hattori and K. Inoue (Eds.), Proceedings of the XIIIth International Congress of Linguistics,
pp. 102–115.
Wearing, Catherine (2012). Metaphor, Idiom, and Pretense. Noûs 46(3), 499–522.
Webelhuth, Gert and Ackermann, Farrell (1994). German Idioms: An Empirical Approach.
Studies in the Linguistic Sciences 24, 455–471.
Weinreich, Uriel (1969). Problems in the Analysis of Idioms. In Weinreich (1980), S. 208–264.
Weinreich, Uriel (1980). On Semantics. University of Pennsylvania Press.

Webelhuth et al: Idioms in LInguistics Theory, Part I

Empfohlen

Empfohlen

Weitere ähnliche Inhalte

Was ist angesagt?

Was ist angesagt? (19)

Ähnlich wie Webelhuth et al: Idioms in LInguistics Theory, Part I

Ähnlich wie Webelhuth et al: Idioms in LInguistics Theory, Part I (20)

Kürzlich hochgeladen

Kürzlich hochgeladen (20)

Webelhuth et al: Idioms in LInguistics Theory, Part I