An Activation-based Sentence Processing Model of English

An Activation-based Sentence Processing Model of English⋆ Kei TAKAHASHI†, Kiyoshi ISHIKAWA‡, and Kei YOSHIMOTO† †Tohoku University {kei-ta,kei}@linguist.jp ‡Hosei University and the University of Edinburgh kiyoshi@i.hosei.ac.jp Abstract. In this paper, we argue that linear order plays crucial role in an adequate account of acceptability judgment and further that the linear order effects we observe should be seen in terms of real-time processing, rather than static syntax. To capture the linear-order effects, we formulate a memory-based model that predicts fine-grained degrees of acceptability. 1 Introduction In the theoritical syntactic literature, hierarchical structures have played a central role in accounting for various phenomena. However, that does not mean that linear order has no role to play. Given that sentences are processed from left to right, it would be rather surprising if linear order played absolutely no role when native speakers make (un)acceptability judgments. In fact, some researchers (e.g. Hawkins 1994) have attempted to explain syntactic phenomena in terms of real-time processing. In this paper, we will deal with certain phenomena that could only be explained in terms of word order. Some of the linear order effects we deal with have already been pointed out in the literature, but to the best of our knowledge there exists no formalized account of the observed linear order effects. This paper reports our ongoing attempt to construct a formalized model which is based on the notion of working memory and activation value, and demonstrate how our model successfulluy accounts for the data in question. The structure of this paper is as follows: first, we summarize the kinds of linear order effects to be accounted for by our model, pointing out that they could not be accounted for in structural terms. Then we illustrate the observational generalizations in processing terms. In section 3, we propose a formalized sentence processing model and in section 4, we discuss a relation between our model and (un)acceptabilities. In section 5 we mention how our model parse sentence and in section 6, we demonstrate how our model accounts for the data in question uniformly. In section 6, we mention some remaining problems and concludes the paper. 2 Background 2.1 Problems for Grammar-based Accounts Kaplan and Bresnan (1982) pointed out the contrast in (1a-b), which is a problem for a movement- based analysis of topicalization.1 An obvious explanation for the unacceptability of (1b) would be that about cannot take a that-clause as its complement. In movement-based analyses, topicalization of the that-clause is not expected to alter the sentence’s (un)acceptable status. However, this expectation is betrayed by (1b).2 ⋆ This study is supported in part from the Tohoku University 21st Century Center of Excllence Program In Humanities (Stragetic Research and Education Center for an Integrated Approach to Language, Brain and Cognition; The LBC Research Center; http://www.lbc21.jp) and Center for Interdisciplinary Research (CIR, Tohoku University). 1 There is disagreement among native speakers about whether there is such a contrast. In this paper we follow Kaplan and Bresnan’s judgments, on the assumption that a complete grammar of …

Content truncated to 500 words