Syllabi and Abstracts


The Semantics and Pragmatics of Metaphor (Dr. Cécile Meier)

Metaphors have been studied by philosophers, but have so far been a stumbling block for formal semanticists. A few semantic studies of metaphor exist, however. (i) Simile theories propose that the meaning of a metaphor can be translated into some kind of comparison or analogy: see Bergmann (1979), Miller (1993) or van Genabith (2001). (ii) Ambiguity theories assume that the literal meaning of an expression is accompanied by a second meaning that is related in some way to the literal meaning: see L. J. Cohen (1993) or Vogel (2001). Both kinds of theories are backed up by psycholinguistic evidence. (iii) Stern (2008) amended Kaplan’s theory on demonstratives to fit metaphorical uses of linguistic expressions and proposed an indexical theory for metaphor. His theory is semantic in nature, as well, as far as issues of deixis are considered to be propositional phenomena.

One reason for the small number of semantic publications was Ted Cohen pointing out in 1975 that metaphorical expressions taken literally are not necessarily semantically anomalous. Another reason is probably Davidson’s 1978 verdict that “metaphors have no meaning beyond the literal meaning”. Both positions led to pragmatic and cognitive explanations of the phenomenon (Searle 1993; Sperber & Wilson 2008; Lakoff & Johnson 1980).

The leading question of the course The semantics and pragmatics of metaphor will be: What is the nature of metaphorical meaning, if it exists? We will give an overview of the semantic theories on metaphor and review counterexamples to them. Of special interest is the career of metaphor hypothesis. According to this hypothesis, metaphors evolve according to a scale of conventionalization. Therefore, we might distinguish different types of metaphors somewhere between novel or poetic metaphors at one end of the scale and dead metaphors at the other, where dead metaphors are usually conceived as idiomatic. We will discuss the coverage of examples by the different theories. 

Bergmann, Merrie. 1979. Metaphor and Formal Semantic theory. Poetics: International Review for the Theory of Literature 8. 213 –130. 

Cohen, Ted. 1975. Figurative Language and Figurative Acts. Journal of Philosophy 72. 669 – 690.

Cohen, L. Jonathan. 1993. The semantics of Metaphor. In: A. Ortony (ed.): Metaphor and Thought (2nd ed. pp. 58 – 70) Cambridge: Cambridge University Press.

Lakoff, George & Mark Johnson. 1980. Metaphors we live by. Chicago: University of Chicago Press.

Miller, George A. 1993. Images and models, similes and metaphors. In: A. Ortony (ed.): Metaphor and Thought (2nd ed. pp. 357 – 400) Cambridge: Cambridge University Press.

Searle, John. 1993. Metaphor.  In: A. Ortony (ed.): Metaphor and Thought (2nd ed. pp. 83 – 111) Cambridge: Cambridge University Press.

Sperber, Dan & Deirdre Wilson. 2008. A deflationary account of metaphors. In: R. Gibbs (ed.) In R. Gibbs (ed.): The Cambridge Handbook of Metaphor and Thought (pp. 84 – 108). Cambridge: Cambridge University Press.

Stern, Josef. 2008. Metaphor, semantics and context. In R. Gibbs (ed.): The Cambridge Handbook of Metaphor and Thought (pp. 262 –279). Cambridge: Cambridge University Press.

Van Genabith, Josef. 2001. Metaphors, Logic and Type Theory. In: Metaphor and symbol 16,  43-57.

Vogel, Carl. 1991. Dynamic Semantics for Metaphor, Metaphor and Symbol, 16(1-2):59-74.

The traditional common outlook on idioms has been that they are non-compositional units stored as a whole in the lexicon. Gazdar et al. 1985, Nunberg, Sag, and Wasow 1994; Sailer 2003, Horn 2003, Webelhuth 2011, among others, have shown, however, that to varying degrees different “idioms” are syntactically and morphologically flexible in that parts of the construction may undergo pronominalization, ellipsis, or movement. The degree of flexibility of idiomatic constructions is now widely held to correlate with the possibility of assigning meanings to the parts of the whole and combining those meanings compositionally. This makes at least some idioms much more compositionally transparent than is usually assumed. The course The Linguistic Status of Idioms will present an overview of the research on idioms in Generative Grammar up to the present and will then address the question of their syntactic flexibility by focusing on the compatibility of different idioms with different syntactic constructions, among others passive, raising, topicalization, restrictive relative clauses. The course will look for answers to the following questions: 1. For each idiom: is it lexically represented as one piece or more than one piece? 2. For those idioms that are (partially) compositional and hence represented by more than one independent lexical entry: which grammatical mechanism ensures, at which level of grammatical licensing, that if one part of the idiom occurs in a syntactic structure, then all other parts occur as well? 3. How is the differential lexemic selectivity of many idioms to be captured, i.e. that part of some idioms are lexemically variable, can be pronominalized, or elided? Is this best captured syntactically or semantically (e.g. with partial functions)? 4. A similar question arises for the varying compatibility of idioms with different syntactic constructions: should this be captured syntactically or can it be made to follow from differences in semantic and/or pragmatic compatibility of (parts of) the idioms and (parts of) the constructions? 5. For idioms that should be represented in one piece: Are there restrictions on the possible size of such idioms or such representations? How does it follow that such idioms typically obey the rules of syntax?

Gazdar, Gerald, Ewan Klein, Geoffrey K. Pullum, and Ivan A. Sag. 1985. Generalized Phrase Structure Grammar. Cambridge, MA: Harvard University Press and Oxford: Basil Blackwell.

Horn, George M. 2003. Idioms, metaphors and syntactic mobility. Journal of Linguistics, 39: 245-273

Nunberg, Geoffrey, Ivan A. Sag, and Thomas Wasow. 1994. Idioms. Language 70: 491– 538.

Sailer, Manfred. 2003. Combinatorial Semantics and Idiomatic Expressions in Head-Driven Phrase Structure Grammar. Arbeitspapiere des SFB 340, Universität Stuttgart and Universität Tübingen. 

Webelhuth, Gert. 2011. Capturing Collocations and Idioms in Relative Clauses without Literal Reconstruction. Talk presented at the Workshop Reconstruction Effects in Relative Clauses at the Zentrum für Allgemeine Sprachwissenschaft, Berlin.


Corpus Linguistics and Non-Compositionality (Dr. Amir Zeldes)
This course will begin with a concentrated introduction to corpus-based approaches to non-compositionality and move on to some practical work with example data. Using pre-prepared corpus resources, students will learn techniques for distributional semantic analysis, focusing on questions such as “how can we identify (non-)compositionality in texts?” and “how can meaning be acquired from usage data?”. Some pointers will also be given for tools and methods to create one’s own resources for further analysis. The evaluation will be carried with the open-source statistics software “R” using empirical distributional semantic methods such as clustering algorithms, collostructional analysis (Stefanowitsch & Gries 2003) and idiomaticity measures (Wulff 2008).

Note: students who wish to carry out the evaluation on their local machine during class should bring a laptop and install the statistics software “R” in advance from: . After installing and running the software, please make sure to download and install the package amap (by clicking on the menu “packages > install”). A link to the complete R scripts used in class will be provided before the course.

Course resources: 
You can find more resources for the course here. The password is 20130623AZ.
Please download the materials for the course from the file

Bruni, Elia, Gemma Boleda, Marco Baroni and Nam-Khanh Tran (2012), Distributional Semantics in Technicolor. In: Proceedings of ACL 2012. East Stroudsburg, PA: ACL, 136-145.
Gries, Stefan Th. (2009), Quantitative Corpus Linguistics with R: A Practical Introduction. London and New York: Routledge.
Stefanowitsch, Anatol & Stefan Th. Gries (2003), Collostructions: Investigating the Interaction of Words and Constructions. International Journal of Corpus Linguistics 8(2), 209-243.
Wulff, Stefanie (2008), Rethinking Idiomaticity: A Usage-based Approach. (Research in Corpus and Discourse.) London and New York: Continuum.

Intuitively, similes and metaphors are clearly related; this fact has led some researchers, ever since Aristotle, to reduce one to the other. Other researchers, however, have argued that the two phenomena actually differ in type. In particular, Davidson claims that there are two important differences between them: similes are always true, whereas metaphors are usually (literally) false; and similes assert the existence of a common property, whereas metaphors require the hearer to retrieve it from the context.
Hence, an investigation of the nature of metaphors must take into account their falsity and indexicality—referring to elements that are recovered from the context of the utterance.
Ideally, the best way to proceed would be to consider a language that does not allow intentional falsehoods (i.e., lies) and indexicals, and observe whether it allows metaphors; we would then add lies and indexicals to the language, and examine whether metaphors are attested in the new language. This, of course, is impossible to do with a real language; but in a science fiction novel, an alien language has been constructed along precisely these lines. The results, although based on a fictional language, are quite instructive regarding real languages. In particular, they lend support to Stern’s indexical
theory of metaphor.

Generic sentences of various types are commonly used to express proverbs. We can easily find proverbs expressed using sentences with generically interpreted indefinite singular subjects (1a), bare plurals (1b), definite singulars (1c), generically interpreted pronouns (1d,e) etc.:
  1. a. A fool's tongue runs before his wit
    b. Bad news travel fast
    c. The chain is not stronger than its weakest link
    d. You cannot teach old dogs new tricks
    e. He that is warm thinks all so

Along this diversity of generic forms, however, it seems that there are also some generic sentences which tend not to be used as proverbs, namely  unambiguously descriptive (or inductive) generics (discussed in e.g. Lawler 1973, Burton Roberts 1977, Cohen 2001,  Greenberg 2003, 2007, 2012, Krifka 2012). Such generics, exemplified in (2a-d), are typically expressed by sentences with generically interpreted bare plural subjects whose minimally contrasting indefinite singular subjects are infelicitous (as generic):

  1. a. Madrigals are popular (# A madrigal is popular)
    b.  Tall, 43 year old singers from Australia earn lots of money (A tall, 43 year old singer from Australia earns lots of money – fine as existential)
    c. Tornados are especially strong this year (# A tornado is especially strong this year)
    d. Dogs happen to have 4 legs (# / ? A dog happens to have 4 legs)

The talk will examine ways to precisely characterize what such sentences semantically have in common, and the idea that such a characterization can indirectly contribute to our understanding of the necessary semantic properties needed for the manifestation of proverbs.

Idioms at the Grammar-Cognition Interface (Prof. M. Teresa Espinal)

It is often assumed that idiomatic constructions involve non-compositional meanings (Katz & Postal 1963; Fraser 1970; Chomsky 1980, among others) or special syntax-semantics correspondences (Jackendoff 1990, 1997a,b, 2002). Reacting against this tradition, I will show that they are compositional (Marantz 1996: Nunberg et al. 1994), and more specifically I will show the relevance of syntactic argument structure in structuring the compositional meaning of idiomatic constructions, as well as the relevance of a distinction between syntactically transparent compositional meanings and syntactically non-transparent non-compositional ones (Mateu & Espinal 2007).

I will focus on the question why idioms such as John worked his guts out or John sneezed his head off are excluded from Romance languages, and how, despite generative claims to the contrary (Marantz 1997; Borer 2005; McGinnis 2002, 2005), various conceptual processes canbe claimed to overrule the aspect provided by grammar. These processes include access to image-schemas (e.g., the PATH image-schema) and metaphors (e.g., (AN EXTREME) INTENSITY IS (AN EXCESSIVE) CHANGE OF LOCATION; SCALES ARE PATHS). Furthermore, I will argue that access to these cognitive domains is what explains that idiomatic expressions can be associated with different aspectual classes at the literal and figurative interpretations (e.g., to sneeze one’s head off, which means an achievement on its literal interpretation, but an activity ‘to sneeze intensively’ on the idiomatic one). The outcome of this discussion will be that a change from telicity to atelicy is expected to arise only when metaphor constrains aspect (Espinal & Mateu 2010, Mateu & Espinal in press).

Logical Structure and Context: a Folkloristic Perspective on Proverbs (Prof. Galit Hasan-Rokem)

Proverbs are a socially acknowledged phenomenon in most known societies. They are a folk literary genre extant in speech as well as writing, and indeed a foremost example of the frequent movement between speech and writing, between orality and various forms of canonization in writing and also in speech. I shall present a folk literary and folkloristic view focusing on the topics: How have folklore scholars discussed the propensity of proverbs to create generalizations? How do proverbs relate to other forms of set speech genres?

The Language of Humor: Ambiguity (Dr. Olga Kagan)

The existence of two alternative ways in which a text can be interpreted plays an important role in the creation of the humor effect. In this class, we will consider the relation between ambiguity and humor and see how different semantic and pragmatic phenomena contribute to the creation of the humor effect. The phenomena we will consider include homonymy, polysemy, structural ambiguity, scope ambiguity, as well as the interpretation of pronouns and definite descriptions.

Idioms: The Type-Sensitive Storage model (Prof. Julia Horvath and Prof. Tal Siloni)

We will present findings from a series of theoretical, corpus-based and experimental studies we conducted on idioms, which gave rise to our Type-Sensitive Storage model (Horvath and Siloni 2012). The evidence and their implications to be discussed involve specifically the mental representation of idioms, namely their lexical storage, obtained from our investigation of phenomena of phrasal idiom distribution among diatheses, order of acquisition, and factors related to the idioms’ internal structure.  

The resulting deeper understanding of the acquisition and storage of idioms has important consequences for linguistic and psycholinguistic theories of the organization of the mental lexicon and the division of labor between the lexicon and the syntactic component of the language faculty.

Metonymy (Prof. Sebastian Löbner)

Metonymy is one of the common post-compositional processes that enable non-literal readings. For example, the sentence
Moscow is going to send blue helmets to Golan Heights
contains two metonymies: reference with the place name Moscow to the Russian government seated in Moscow, and reference with the headgear term  blue helmet to U.N. peacekeepers.
Known from antiquity as a particular kind of figure of speech, metonymy turned into an important object of study in semantics and cognitive psychology within the last decades. In semantics it is studied as a post-compositional process along with metaphor, differentiation and other kinds of conceptual shift. In cognitive psychology, metonymy became recognized as one of the basic cognitive mechanism of forming new concepts out of given ones.
In spite of the increased interest in the phenomenon there is no precise definition available yet. Traditional descriptions are based on the vague notions of ‘same domain’ and ‘contiguity’, of little explanatory value, or they operate with an open list of characteristic relations between the literal and the metonymical referents – with no apparent generalization offering itself.
The lecture on metonymy will point out properties of metonymic reference hitherto unnoticed. It will provide a connection of the phenomenon to Barsalou frames (essentially recursive attribute-value structures), pointing out that, and how, well-known examples of metonymy can be analyzed as a certain type of operations on semantic frames ( = noun and verb meanings).

Pavel Kats,
Jun 26, 2013, 7:47 AM