Automatic syntactic computation and use of semantic information during sentence comprehension

Summary The paper represents a contribution to the issue of the differential effects of syntactic, semantic and pragmatic information in the process of sentence comprehension. The results of four experiments are reported, in which sentences containing syntactic, semantic and pragmatic violations were presented with different tasks to see how these violations affected language comprehension, how easily they were detected and whether they affected sentence comprehension even when they were not explicitly recognized. The main results can be summarized as follows. First, the most relevant cues used to comprehend a sentence are not so much syntactic, but semantic or pragmatic. Second, with difficult or anomalous linguistic input, syntactic cues become more important. Third, even when the reader is not aware of syntactic violations, these seem to affect his processing of the linguistic input. This suggests the hypothesis of an automatic computation of syntactic information, the output of which is, however, used only when evidence coming from other sources, semantic or pragmatic, is insufficient or not clear enough.     

Individual differences in syntactic ambiguity resolution: Readers vary in their use of plausibility information

Two experiments investigated the relation between individual differences in working memory capacity and differences in the efficiency of syntactic processing. In one experiment, readers comprehended sentences containing main-verb/reduced-relative ambiguities that all resolved to the reduced-relative interpretation. High-span (but not low-span) readers processed sentences more slowly when the sentences were biased to the preferred, main-verb interpretation than when they were biased to the reduced-relative interpretation. Moreover, high-span (but not low-span) readers used information about the plausibility of the different interpretations even though low-span readers appeared to possess the requisite knowledge. In Experiment 2, readers received intensive exposure to sentences with main-verb/reduced-relative ambiguities. Exposure enhanced low-span readers' use of plausibility information. Moreover, the effect of exposure generalized to sentences that were not included in the training materials.     

The retrieval of syntactic and semantic information in the production of verbs

In Italian, there are two aspectual auxiliaries: essere (to be) and avere (to have). According to the syntactic hypothesis, aspectual auxiliary assignment in Italian is a syntactic phenomenon. Using the picture-word interference paradigm, Experiment 1 tested the predictions of this view. Results failed to support the syntactic hypothesis but could be explained within a semantic framework: A verb's thematic grid is involved in aspectual auxiliary assignment. Experiment 2 tested this hypothesis. The findings corroborated the predictions. The evidence is interpreted as suggesting the involvement of thematic information in the assignment of the aspectual auxiliary during the production of Italian verbs.     

语义和句法信息在副中央凹加工中的作用

读者能够从副中央凹中提取到什么类型的信息, 是当前阅读眼动研究领域关注的焦点问题.当前普遍认为低水平信息(如正字法信息)可以在副中央凹中得以加工, 但是高水平的信息(如语义和句法信息)能否从副中央凹中进行提取存在争议.本文总结了近年来高水平信息在副中央凹加工的研究进展, 包括拼音文字和非拼音文字(如中文)阅读过程中语义和句法预视效益的研究现状及影响因素, 当前眼动控制模型(如E-Z读者和SWIFT)对预视效益的解释和不足之处, 最后提出未来关于语义和句法预视加工的研究方向.     

Gender differences in stress reactions: Women filling the gaps

Abstract

Competing hypotheses explaining gender differences in depression were compared in a one-year longitudinal study of parents of ill children in Israel. Women were found to have more depressed mood than men when their children were ill or well, but both men and women responded with increased depression when their child was ill. Women were more likely than men to be with an ill child at the hospital, even if employed, suggesting role overload. Women did not report fewer personal or social resources than men, indicating that poverty of resources could not explain sex differences on depression. Nor did women use their resources less effectively than did men. Women were less depressed the greater their intimacy with their spouse and friends, but at all levels of intimacy were more depressed than men. Men, in contrast, isolated themselves from social support at the time of their child's illness and it was argued that men might be avoiding stress contagion. The importance of examining differential responding of men and women to stressors that have common meaning for both sexes was discussed.     

Mandarin learners use syntactic bootstrapping in verb acquisition

Mandarin Chinese allows pervasive ellipsis of noun arguments (NPs) in discourse, which casts doubt concerning child learners' use of syntax in verb learning. This study investigated whether Mandarin learning children would nonetheless extend verb meanings based on the number of NPs in sentences. Forty-one Mandarin-speaking two- and three-year-olds enacted sentences with familiar motion verbs. The presence of an extra postverbal NP led the toddlers to extend causative meanings to intransitive verbs, and the absence of a postverbal NP led them to extend noncausative meanings to transitive verbs. The effects of adding or subtracting an NP were statistically indistinguishable. Thus, the number of NPs in a sentence appears to play a role in verb learning in Mandarin Chinese.     

The consequences of reduced memory span for the comprehension of semantic versus syntactic information

Aphasic patients with restricted memory spans were assessed on their comprehension of syntactically simple sentences varying in numbers of content words and on their comprehension of sentences matched in content words but varying in syntactic complexity. Three different presentation modes were used: unlimited visual presentation, limited visual presentation, and auditory presentation. The difference between performance in the unlimited and limited modes was used as a measure of the extent to which memory decrements caused comprehension failure. These difference scores were related to memory span in order to assess whether the memory involved in sentence processing overlaps that assessed by a memory span task. For the sentences varying in number of content words, performance was worse for the limited presentation modes than the unlimited mode, and the difference between unlimited and limited presentation modes was highly related to memory span. For the sentences varying in syntactic complexity, there was no clear reduction in performance for limited presentation modes, and the difference scores did not relate to memory span. The results support the view that the processing of sentences with many content words draws on the short-term memory capacity assessed by memory span, while the processing of syntactic complexity does not.     

Redundancy and reduction: Speakers manage syntactic information density

A principle of efficient language production based on information theoretic considerations is proposed: Uniform Information Density predicts that language production is affected by a preference to distribute information uniformly across the linguistic signal. This prediction is tested against data from syntactic reduction. A single multilevel logit model analysis of naturally distributed data from a corpus of spontaneous speech is used to assess the effect of information density on complementizer that-mentioning, while simultaneously evaluating the predictions of several influential alternative accounts: availability, ambiguity avoidance, and dependency processing accounts. Information density emerges as an important predictor of speakers’ preferences during production. As information is defined in terms of probabilities, it follows that production is probability-sensitive, in that speakers’ preferences are affected by the contextual probability of syntactic structures. The merits of a corpus-based approach to the study of language production are discussed as well.     

Lexical integration: Sequential effects of syntactic and semantic information

Both semantic and syntactic context constraints can influence word processing at the level of lexical integration. In event-related brain potentials (ERPs), semantic integration is reflected by a negativity around 400 msec (N400), whereas phrase structure assignment and syntactic integration are assumed to be reflected by an early left anterior negativity and a late positivity (P600), respectively. An ERP study is presented in which participants read different types of sentences whose terminal verb was either congruent with the preceding context or incongruent due to a phrase structure violation, a semantic violation, or both. The main finding was that only the pure semantic violation condition, but not the combined semantic and syntactic violation condition, elicited a large N400. The two conditions containing phrase structure violations were predominantly characterized by a P600. Both semantic violation conditions, moreover, displayed a late negativity around 700 msec that overlapped with the P600 in the double violation condition. The absence of an N400 effect for elements that are syntactically as well as semantically incongruent with prior context suggests an early influence of phrase structure information on processes of lexical-semantic integration. The present data are discussed in comparison to previous ERP findings, and a new view of lexical integration processes is proposed.     

Immediate use of prosody and context in predicting a syntactic structure

Numerous studies have reported an effect of prosodic information on parsing but whether prosody can impact even the initial parsing decision is still not evident. In a visual world eye-tracking experiment, we investigated the influence of contrastive intonation and visual context on processing temporarily ambiguous relative clause sentences in Japanese. Our results showed that listeners used the prosodic cue to make a structural prediction before hearing disambiguating information. Importantly, the effect was limited to cases where the visual scene provided an appropriate context for the prosodic cue, thus eliminating the explanation that listeners have simply associated marked prosodic information with a less frequent structure. Furthermore, the influence of the prosodic information was also evident following disambiguating information, in a way that reflected the initial analysis. The current study demonstrates that prosody, when provided with an appropriate context, influences the initial syntactic analysis and also the subsequent cost at disambiguating information. The results also provide first evidence for pre-head structural prediction driven by prosodic and contextual information with a head-final construction.     

Untrained speakers' use of prosody in syntactic disambiguation and listeners' interpretations

We investigated how naively produced prosody affects listeners' end interpretations of ambiguous utterances. Non-professional speakers who were unaware of any ambiguity produced ambiguous sentences couched in short, unambiguous passages. In a forced-choice task, listeners could not tell which context the isolated ambiguous sentences came from (Exp. 1). However, listeners were able to correctly paraphrase the least ambiguous subset of these utterances, showing that prosody can be used to resolve ambiguity (Exp. 2). Nonetheless, in everyday language use, both prosody and context are available to interpret speech. When the least ambiguous sentences were cross-spliced into contexts biasing towards their original interpretations or into contexts biasing towards their alternative interpretations, answers to content questions about the ambiguous sentence, confidence ratings, and ratings of naturalness all indicated that prosody is ignored when context is available (Exp. 3). Although listeners can use prosody to interpret ambiguous sentences, they generally do not, and this makes sense in light of the frequent lack of reliable prosodic cues in everyday speech. Received: 3 April 1998 / Accepted: 21 October 1998     

The use of visual information in natural scenes

Despite the complexity and diversity of natural scenes, humans are very fast and accurate at identifying basic-level scene categories. In this paper we develop a new technique (based on Bubbles, Gosselin & Schyns, 2001a; Schyns, Bonnar, & Gosselin, 2002) to determine some of the information requirements of basic-level scene categorizations. Using 2400 scenes from an established scene database (Oliva & Torralba, 2001), the algorithm randomly samples the Fourier coefficients of the phase spectrum. Sampled Fourier coefficients retain their original phase while the phase of nonsampled coefficients is replaced with that of white noise. Observers categorized the stimuli into 8 basic-level categories. The location of the sampled Fourier coefficients leading to correct categorizations was recorded per trial. Statistical analyses revealed the major scales and orientations of the phase spectrum that observers used to distinguish scene categories.     

Preschoolers use phrasal prosody online to constrain syntactic analysis

Two experiments were conducted to investigate whether young children are able to take into account phrasal prosody when computing the syntactic structure of a sentence. Pairs of French noun/verb homophones were selected to create locally ambiguous sentences ([la petite ferme ] [est très jolie] ‘the small farm is very nice’ vs. [la petite] [ ferme la fenêtre] ‘the little girl closes the window’ – brackets indicate prosodic boundaries). Although these sentences start with the same three words, ferme is a noun (farm) in the former but a verb (to close) in the latter case. The only difference between these sentence beginnings is the prosodic structure, that reflects the syntactic structure (with a prosodic boundary just before the critical word when it is a verb, and just after it when it is a noun). Crucially, all words following the homophone were masked, such that prosodic cues were the only disambiguating information. Children successfully exploited prosodic information to assign the appropriate syntactic category to the target word, in both an oral completion task (4.5‐year‐olds, Experiment 1) and in a preferential looking paradigm with an eye‐tracker (3.5‐year‐olds and 4.5‐year‐olds, Experiment 2). These results show that both groups of children exploit the position of a word within the prosodic structure when computing its syntactic category. In other words, even younger children of 3.5 years old exploit phrasal prosody online to constrain their syntactic analysis. This ability to exploit phrasal prosody to compute syntactic structure may help children parse sentences containing unknown words, and facilitate the acquisition of word meanings.     

The semantics of syntactic structures

Over the past 15 years, both linguists and psycholinguists have shown a growing interest in the idea that syntactic structures can carry meanings that are independent of the meanings of specific words. This article considers how this idea relates to traditional notions of compositionality in generative grammar, and examines two modern theories that, although based on different starting assumptions, both readily allow syntactic structures to bear independent meaning. We review work from psycholinguistics suggesting that observation alone is often insufficient to support the efficient learning of word meanings, and that some of the 'slack' left by observation can be picked up by the semantics of the syntactic structures in which words appear. We argue that this convergence between linguistic theory and psycholinguistic experimentation should be no surprise, because language must be learnable.     

The use of abstract graphemic information in lexical access

Three experiments investigated the nature of the information required for the lexical access of visual words. A four-field masking procedure was used, in which the presentation of consecutive prime and target letter strings was preceded and followed by presentations of a pattern mask. This procedure prevented subjects from identifying, and thus intentionally using, prime information. Experiment I extablished the existence of a semantic priming effect on target identification, demonstrating the lexical access of primes under these conditions. It also showed a word repetition effect independent of letter case. Experiment II tested whether this repetition effect was due to the activation of graphemic or phonemic information. The graphemic and phonemic similarity of primes and targets was varied. No evidence for phonemic priming was found, although a graphemic priming effect, independent of the physical similarity of the stimuli, was obtained. Finally Experiment III demonstrated that, irrespective of whether the prime was a word or a nonword, graphemic priming was equally effective. In both Experiments II and III, however, the word repetition effect was stronger than the graphemic priming effect. It is argued that facilitation from graphemic priming was due to the prime activating a target representation coded for abstract (non-visual) graphemic features, such as letter identities. The extra facilitation from same identity priming was attributed to semantic as well as graphemic activation. The implications of these results for models of word recognition are discussed.     

Retention functions for syntactic and lexical vs semantic information in sentence recognition memory

In a continuous recognition memory design, Ss judged whether each sentence was identical in form and meaning to some previously presented sentence, then judged whether the sentence was identical in meaning irrespective of form, and, finally, rated the likelihood of recognizing the sentence ff it was presented an hour later (memorability). The Ss were given sentences that were new, identical to, or paraphrased from some previously presented sentence, at delays ranging from 0 sec to 2 h. Long-term memory for both semantic information and syntactic-lexical information decayed according to the same exponential-power retention function previously found to be characteristic of the decay of simpler verbal materials (nonsense items, letters, digits, words, and word pairs). Semantic memory primarily differed from syntactic-lexical memory in that the semantic information had a far higher degree of learning, but the decay rate for syntactic-lexical information was also approximately 5 0% greater than the decay rate for semantic information.