Abstract not found

this paper was published in Heidi Harley (ed.), Papers from the UPenn / MIT Roundtable on Argument Structure and Aspect (MIT Working Papers in Linguistics 32, 1998), under the title Licensing in the Non-Lexicalist Lexicon. Nominalizations, Vocabulary Items and the Encyclopaedia

This paper presents data from Kannada that resist either type of analysis ( 2). The grammar of reflexivity in Kannada reveals that resultatives must be syntactically distinguished both from ECM constructions and from simple transitives. Simple transitives and ECM constructions allow verbal reflexive marking when reflexive, but resultatives do not

In this paper we analyse the interdependence of Persian nonverbal (NV) element and the light verb (LV) in determining the syntactic properties, the

: We argue that the Minimalist view of phrase structure in Chomsky 1995 should be minimally extended to allow for phrases that have more than one head, so long as the two heads have the same category features and are not attracted by a higher head. This innovation results in an elegant typology of the various kinds of syntactically distinguishable serial verb constructions (SVCs) found in Edo and related West African languages, as discovered by Stewart (1998). In particular, we claim that the different SVCs come from different choices of which phrase in the clausal structure is doubly headed: Voice, light v, or V. Moreover, details of Edo syntax allow us to make some refinements to the theory of clause structure; these include showing that Kratzer's Voice head is distinct from Chomsky's v head, and showing exactly where agents, themes and goal phrases are generated. Empirical evidence for our claims comes from a variety of syntactic and semantic sources, but especially from the positio...

this paper, we pursue a constructionalist-type explanation of the alternation we discuss in section 3 because we agree with the fundamental constructionalist hypothesis on the decomposition of meaning in the syntax. Our motivation for pursuing this kind of approach is both theoretical and empirical. Starting from the theoretical strength of a syntax-based approach, it has been shown that there are a number of syntactic phenomena that a lexicalist approach has difficulty explaining (see for example Rosen 1984 for a discussion of the phenomenon of unstable valency of unaccusative/unergative verbs or Hoekstra and Mulder (1990) on the alternating behaviour of motion verbs in Italian and Dutch). It is even more implausible that all the syntactic correlates of the unaccusative/unergative distinction (ne-cliticization, auxiliary selection, availability of resulative construction, er-nominalisation in English, and possessor datives in Hebrew etc

People often want to search a large number of articles, web pages, or other natural language documents to find the answers to specific questions. Currently, the best techniques for performing such searches treat each document as an unordered collection of words, and disregard any information about how the words relate to each other. Although this approach has been quite successful, it has a number of fundamental limitations. In order to explore the potential benefit of using more advanced, linguistically-motivated techniques, I built SQUIRE, a Semantic Question-answering Information Retrieval Engine. SQUIRE uses a number of independent transformations to process individual sentences, converting them into sets of semantic relationships. It can then search those semantic relationships in response to a user's question. SQUIRE significantly out-performs traditional IR techniques in the task of returning the individual sentences that answer a question.

Abstract not found

This paper argues for (a) the existence of both abstract Case and morphological case, the latter realizing the former according to the Elsewhere Condition; and (b) the non-existence of abstract absolutive Case. A class of ergativeabsolutive nominative Case and transitive objects receive abstract accusative Case; both nominative and accusative are realized through a morphological default, the “absolutive”. Such languages behave systematically distinctly from ergativeabsolutive languages in which intransitive subjects and transitive objects both receive abstract nominative Case. The relationship between abstract and morphological case based on the Elsewhere Condition is further supported by case mismatches in Pama-Nyungan languages. The data standardly cited against such a relationship, Icelandic quirky case, are re-examined.

This paper investigates two ways goals of motion events can be expressed in so-called ‘verb-framed’ languages (Talmy 2000), focusing on the Japanese postpositions -made and -ni. It is typically assumed that these postpositions are both goal-markers, but differ in exact goal semantics they encode, giving rise to non-overlapping distributions. Based on a range of distributional differences, I argue instead that they are more radically distinct that this:-made marks endpoints of event participants (including but not limited to paths of motion), while-ni is a dative case that marks goal arguments of motion verbs. This suggests that it is possible for two functionally distinct participant markers to converge and give the appearance of being alternate ways of realizing the ‘same ’ participant. Furthermore, adpositions such as-made, an inherently non-motion encoding resource co-opted for use in motion constructions, represent an understudied strategy for marking goals across languages, something that has ramifications for how motion typologies are constructed.