(Ohio State University. Department of Linguistics, 2001) Cipollone, Domenic
In this paper we take a fresh look at an old problem, the syntax and semantics of Japanese causatives. We demonstrate some seldom-noted similarities causatives bear to other Japanese morphologically complex predicates and argue why these similarities are important. Following a survey and critique of past analyses, we conclude that the principle of compositionality is at the root of the deficiencies of these analyses. We thus propose a modified, slightly non-compositional version of Manning et al.’s (1999) analysis, similar in spirit to Minimal Recursion Semantics (Copestake et al. 1995, 1999). We conclude with some discussion of possible replacements for compositionality.
(Ohio State University. Department of Linguistics, 2001) Lee, Sun-Hee
This paper deals with the formation of complex predicates and some interesting scrambling facts in Korean. First, we extend the notion of complex predicates to include various noun-verb combinations by providing syntactic and semantic evidence. Within the HPSG framework, we then propose a general schema based on argument composition, which can be used for different types of complex predicates. Furthermore, in opposition to Chung (1998)’s approach using argument composition for scrambling phenomena, we argue that linearization constraint is better to account for various permutation possibilities in Korean.
(Ohio State University. Department of Linguistics, 2001) Pappas, Panayiotis A.
This paper presents some results from an in-depth analysis of the phenomenon of variation in weak object pronoun placement in Later Medieval Greek, focusing on the language-internal parameters that affect the variation. The findings reveal a complex pattern of variation that cannot be fully understood at this stage, and pose interesting questions for further investigation.
(Ohio State University. Department of Linguistics, 2001) Stewart, Thomas W., Jr.
Accounts of Georgian morphological agreement marking on verbs have been frustrated by systematic deviations from regular morphemic behavior (co-occurrence restrictions and the so-called ‘inversion’ construction). A theory of inflection which does not assume the morpheme (e.g. Paradigm Function Morphology (PFM; Stump 1991, 1993, 2001)) permits the ready formal expression of some recalcitrant aspects of the distribution of agreement markers, but not all. By expanding the database somewhat and by capitalizing on independently motivated resources available within PFM, an approach is put forward here which shows the Georgian facts to fully respect rule ordering based wholly on proper subset exclusion (PFM’s Pāninian Determinism Hypothesis), without resorting to extrinsically imposed stipulations.
(Ohio State University. Department of Linguistics, 2001) Vaillette, Nathan
This paper explores some theoretical properties of summativity, a generalization of cumulativity. It presents an approach to plural semantics in which summativity can apply not only to lexical predicates, but also to partially saturated predicates. It is shown how this approach can be tied to an explicit type-logical syntax.
(Ohio State University. Department of Linguistics, 2001) Vasishth, Shravan
Data from Hindi center-embedding constructions (CECs) are used to evaluate three sentence processing models: Joshi’s Embedded Pushdown Automaton (EPDA), Gibson’s Syntactic Prediction Locality Theory (SPLT), and Lewis’ Interference and Confusability Theory (ICT). The SPLT and ICT (but not the EPDA) are found to correctly predict several processing facts about Hindi. However, the experimental results also reveal a problem for these two current, wide-coverage theories: neither model appears to be able to account for differences in reading time observed at noun phrases in Hindi CECs. A sentence processing model is proposed in an accompanying article (see (Vasishth & Kruijff 2001) in this volume) that can in principle be integrated with the ICT to provide a unified account of processing difficulty in the languages investigated.
(Ohio State University. Department of Linguistics, 2001) Vasishth, Shravan; Kruijff, Geert-Jan M.
A sentence processing model is presented, based on abductive and deductive inference. We show that the model makes correct predictions for an array of data involving Dutch, German, Japanese, and Hindi center-embedding constructions. It has comparable or better empirical coverage with respect to several other theories of sentence processing, and can be integrated into an existing wide-coverage model, Lewis’ Interference and Confusability Theory, to obtain an integrated theory of working memory constraints on human language processing.