Rapid processing of the meaning of sentences

Abstract

It has been shown (Fischler & Bloom, 1979) that sentence contexts facilitate a lexical decision task for words that are highly likely sentence completions and inhibit the decision for words that are semantically anomalous sentence completions. In the present experiment, the sentence contexts were presented 1 word at a time, at rates from 4 to 28 words/sec. The facilitation for words that were likely sentence completions was marginal at the slower rates and absent at higher rates. In contrast, the inhibitory effects of semantic anomaly were apparent at all presentation rates. Several analyses suggested that the sentence contexts were becoming ineffective at the very highest presentation rates, but the high rates at which the sentence contexts still affected word recognition were taken as evidence that semantic information accrues at an early stage of sentence processing. Implications for Posner and Snyder's (1975) theory of attention and for models of reading were discussed.For most literate adults, the various skills involved in reading are highly overlearned. Years of practice in visual word recognition, grammatical organization, and semantic analysis enables the reader to obtain an accurate representation of a sentence quite rapidly and with little apparent effort. The recent theoretical separation of automatic and attentional processes (LaBerge, 1976; Posner & Synder, 1975; Schneider & Shiffrin, 1977) has renewed interest in various aspects of rapid reading skill. The present paper is concerned with the role of sentence context on word recognition and, more specifically, is an extension of a series of experiments reported previously (Fischler & Bloom, 1979), in which we have attempted to separate the automatic from the more effortful aspects of how context is used in reading. Effects of Sentence Contexts on Word RecognitionSeminal experiments by Morton (1964) and Tulving and Gold (1963) showed that the semantic context of a sentence could indeed influence the subsequent recognition of a word, as measured by visual duration thresholds for identification. Both studies found a reduction in threshold for words that were semantically appropriate and likely completions of a sentence context, with less effect as the Cloze probability (Taylor, 1953) of a word's being a response to the sentence context declined. For instance, the threshold for the word COLLISION was uninfluenced by the prior presentation This research was supported by National Institute of Mental Health Grant R03 MH30395 to the first author. We thank Daniel Tumarkin, Eric Nietzke, and Roger Lewis for their assistance in data collection and James Neely for his careful critique of an earlier version of the paper. Results of an earlier stage of the research were reported at the meeting

Microstates in Language-Related Brain Potential Maps Show Noun–Verb Differences

“…Curran and collaborators (1993) observed an ERP component at 136-184 msec that was different for congruent and incongruent sentence endings; the authors noted that the same component was selectively left-lateralized if semantic processing was required. Reading is comfortable at a speed of about five words per second (Fischler & Bloom, 1980) and semantic priming effects occur if only 250 msec are available to process the prime (Neely, 1977). In agreement with the present results, these observations suggest that at least some of the semantic processing of incoming information occurs within the first 200 msec.…”

Visual word recognition by bilinguals in a sentence context: Evidence for nonselective lexical access.

“…From the monolingual domain it is well known that the semantic and syntactic framework that one constructs when reading a sentence provides an important topBilingual Word Recognition in Sentences 4 down influence on lexical access of the words appearing further in the sentence. For example, there is ample evidence that more predictable words are processed faster in a variety of production and recognition tasks such as naming (e.g., McClelland & O'Regan, 1981;Stanovich & West, 1983), lexical decision (e.g., Fischler & Bloom, 1979;Fischler & Bloom, 1980;Schwanenflugel & Lacount, 1988;Schwanenflugel & Shoben, 1985) and speech monitoring (e.g., Cole & Perfetti, 1980). Similarly, eyetracking studies have consistently shown that more predictable words are skipped more often, and yield shorter fixation times (e.g., Balota, Pollatsek, & Rayner, 1985;Ehrlich & Rayner, 1981;Rayner & Well, 1996).…”

An Event-Related Brain Potential Analysis of Visual Word Priming Effects

“…This evidence connects to discussions in the RT priming literature, where it has been something of an issue whether an essentially meaningless real word (such as blank) is to be preferred over a string of nonletter symbols (e.g., asterisks, hash marks) as the neutral prime stimulus and, more in general, whether a truly neutral stimulus can be achieved at all (cf. De Groot, Thomassen, & Hudson, 1982;Fischler & Bloom, 1980;Shelton & Martin, 1992). This issue is examined under General Discussion, following which the remainder of the discussion focuses on the implications of the results obtained for the target words.…”