Linguistics Home

Syntax, Semantics and Pragmatics

The following is a sample of some of the research projects currently being directed by members of the Linguistics Department:

Case, Voice and Agreement

It is obvious that systems of case assignment, voice (active, passive, middle), and agreement interact in significant ways. Recent research, however, is deeply divided on the nature of their interaction. On the one hand, some traditional theories posit a one-to-one mapping between case and voice-determining morphemes, and allow languages simply to vary in what morphological cases are assigned by what kinds of morphemes (accounting for accusative, ergative, and split systems in a fairly mechanical, locally determinable way). On the other hand, there is a strand of investigation that analyzes the interaction as indirect, making case assignment dependent on global properties of clauses, and locating cross-linguistic variation in various global (directionality, economy) parameters. Both kinds of approaches face significant empirical challenges, and case/voice/agreement systems remain a central testing ground for theories of the human language faculty; the implications for the computational system if the second approach in particular is correct are enormous. This project aims to work out the details of a one-to-one system, exploring the consequences for our understanding of the functional structure of clauses. (Jason Merchant)

Classifiers versus Noun Class Affixes

Important typological differences have been noted cross-linguistically between numeral classifying languages and others such as English and French, which are assumed to peruse only a handful of classifiers. There is more mythology about the number of classifiers than on whether or not they are used productively. When one considers constructions such as ring of mountains and pride of lions, the claim that English has just a handful of classifiers becomes difficult to support, although the contexts in which these are used are certainly more constrained, in terms of both what the classifiers can combine with and how generally they are required in the usage of INDIVIDUATED nouns. More difficult and less studied are, on the one hand, the question of how to articulate the distinction between numeral classifiers and noun class affixes and, on the other, that of how both of these differ from uses of articles that signal individuation. ( Salikoko S. Mufwene)

The Count/Mass and the Stative/Nonstative Distinctions

There is a certain amount of cross-linguistic variation in the ways these distinctions have been articulated. Languages such as English suggest that the distinctions may not be strictly lexical. Alternations such as We ate a duck versus We ate duck suggest that the COUNT/MASS distinction may be in the proper domain of noun delimitation (having to do with whether or not a determiner is used), though many nouns display “countability preferences.” Constructions such as I am enjoying myself and Love your neighbor raise issues about the classic diagnostics that have been adduced in the literature to support the STATIVE/NONSTATIVE distinction, showing that this is not a clearcut dichotomy. It seems more adequate to assume a cline on which different verbs display different degrees of stativity. ( Salikoko S. Mufwene)

Ellipsis Cross-linguistically

The study of the nature of ellipsis phenomena across languages is in its infancy: traditional grammarians' work and typological work has entirely ignored ellipsis (try finding the entry for "sluicing" in a traditional grammar). This lacuna has started to be filled in recent years by a number of researchers examining "missing constituent" phenomena in more and more languages, in particular phenomena that bear strong resemblance to sluicing and predicate or verbal phrase ellipsis in better studied languages like English or Japanese. As part of a book compiliation, we are currently examining in detail elliptical phenomena in a wide range of languages, addressing fundamental questions concerning the nature of cross-linguistic variation in both syntactic and semantic domains. To what extent are the elliptical phenomena that can be found in a given language determined by or dependent on other features of the grammar of that language? In particular, is the presence of sluicing determined by a language having overt wh-movement? Does predicate ellipsis require a certain kind of auxiliary system? Do nominal ellipses require certain kind of agreement systems? Are all kinds of ellipses, within and across languages, subject to identical semantic and pragmatic constraints? The answers will shed light on what kind of variation needs to be allowed by our theories of grammatical competence. (Jason Merchant)

Parameters of Comparison

The ability to make and evaluate comparisons is a basic component of human cognition, yet languages vary widely in the way that comparisons are expressed syntactically. Why? Does this variability indicate a corresponding variability in the underlying semantics of comparative constructions, or is it possible to maintain a universal semantics of comparison and explain the variability in some other way? If the former, what (if any) aspects of meaning are universal in comparative constructions, and what specifically is subject to parametric variation? If the latter, what is the universal semantics of comparatives, and what are the (non-semantic) factors that are responsible for the observed typological variation? The goal of this research project is to answer to these questions. (Chris Kennedy)

Time Reference in Creoles

Is there really a time reference system specific to creoles? Do the systems of different creoles boil down to the stereotype that these vernaculars have a more developed aspectual system than tense system? Do they all have a relative tense system? What contribution can research on creoles’ time reference systems make to our understanding of time reference in their lexifiers? ( Salikoko S. Mufwene)

Unpronounced Structures

A lively debate is on concerning whether there exist unpronounced syntactic structures. Familiar battlegrounds are VP ellipsis, sluicing, and, more recently, fragment answers. This project aims to collate existing and find new evidence for such structures, especially from the presence of syntactic locality effects such as island sensitivities that cannot be explained without recourse to an unpronounced syntactic phrase marker. Our older work has examined the presence and absence of island effects in left branch structures such as attributive comparatives, and the project continues with the recent LSA symposium on ellipsis as well as work demonstrating the presence of island effects in phrasal comparatives in Greek. (Anastasia Giannakidou, Chris Kennedy, Jason Merchant)

Word Meaning

The characterization of word meaning has remained elusive in lexical semantics, raising the question of whether dictionary definitions provide meanings or simply help the user infer the denotations of the relevant words. How does traditional kinship semantics as inspired by ethnosemantics fare in the current reflection on what meaning is? Was Saul Kripke basically right in assuming that the distinction between proper names and common nouns has to do primarily with “rigid designation”? How can this distinction be characterized nowadays? Is his “baptismal theory” still supported by (some of) the issues arising today? Can one realistically provide word meanings exhaustively and in isolation? Or should we consider more serious Ludwig Wittgenstein’s “meaning-is-use theory”? ( Salikoko S. Mufwene)