Rui P. Chaves

Enriching wordnets with new relations and with event and argument structures

This paper argues that wordnets, being concept-based computational lexica, should include information on event and argument structures. This general approach is relevant both for allowing computational grammars to cope with a number of different lexical semantics phenomena, as well as for enabling inference applications to obtain finer-grained results. We also propose new relations in order to adequately model non explicit information and cross-part-of-speech relations.

Conjunction, cumulation and respectively readings

Abstract So-called respectively readings have posed serious challenges for theories of syntax and semantics. Sentences like George and Martha respectively denounced and were denounced by the governor (McCawley 1998) show that although the conjoined verbal expressions share the same syntactic subject, they do not predicate that subject in the same way; George (not Martha) denounced the governor, andMartha (but not George) was denounced by the governor. Postal (1998, 160–163) and Gawron & Kehler (2004, 193–194) show that this phenomenon poses problems for contemporary theories of grammar and argue that it is particularly acute for theories where subcategorization and predication are linked via unification. As these authors note, the problem is even more severe in respectively readings involving filler–gap constructions. In this paper I argue that the severity of these problems has been overstated and that the data do not entail any special dissociation between predication, subcategorization, or extraction. In this paper I propose an account which is fully compatible with unification-based theories of grammar. Gawron & Kehler (2004) propose an account of respectively phenomena which covers a remarkably wide range of cases. That approach relies on a Respf operator, which is stipulated to be optionally overt. However, this analysis is arguably problematic because there are significant semantic differences between respectively readings with and without an overt realization of ‘respectively’. Rather, the data suggest that respectively readings may be special cases of more general phenomena which happen to create interpretations compatible with the semantics of the adverb ‘respectively’. This explains why respectively readings can arise without the adverb, and does not require positing a disconnect between predication and subcategorization. In fact, a sentence with a respectively reading will not differ in syntactic or semantic structure from sentences without such a reading.

An expectation-based account of subject islands and parasitism

Subject phrases impose particularly strong constraints on extraction. Most research assumes a syntactic account (e.g. Kayne 1983 , Chomsky 1986 , Rizzi 1990 , Lasnik & Saito 1992 , Takahashi 1994 , Uriagereka 1999 ), but there are also pragmatic accounts (Erteschik-Shir & Lappin 1979 ; Van Valin 1986 , 1995 ; Erteschik-Shir 2006 , 2007 ) as well as performance-based approaches (Kluender 2004 ). In this work I argue that none of these accounts captures the full range of empirical facts, and show that subject and adjunct phrases (phrasal or clausal, finite or otherwise) are by no means impermeable to non-parasitic extraction of nominal, prepositional and adverbial phrases. The present empirical reassessment indicates that the phenomena involving subject and adjunct islands defies the formulation of a general grammatical account. Drawing from insights by Engdahl ( 1983 ) and Kluender ( 2004 ), I argue that subject island effects have a functional explanation. Independently motivated pragmatic and processing limitations cause subject-internal gaps to be heavily dispreferred, and therefore, extremely infrequent. In turn, this has led to heuristic parsing expectations that preempt subject-internal gaps and therefore speed up processing by pruning the search space of filler–gap dependencies. Such expectations cause processing problems when violated, unless they are dampened by prosodic and pragmatic cues that boost the construction of the correct parse. This account predicts subject islands and their (non-)parasitic exceptions.