Academic journal article Canadian Journal of Experimental Psychology

Modeling Lexical Decision and Word Naming as a Retrieval Process

Academic journal article Canadian Journal of Experimental Psychology

Modeling Lexical Decision and Word Naming as a Retrieval Process

Article excerpt

Abstract We argue that rule-like phenomena in naming and lexical decision reflect the collapsing of information that occurs during retrieval from the lexicon, and that complex patterns in performance reflect the pattern of correlation that exists in the reader's lexicon rather than mapping rules wired into, or learned by, the processing apparatus. By using a lexicon built to scale, we show that simple retrieval operations applied to a large corpus of words correctly predict an interaction of word frequency by spelling-tosound regularity in naming, a frequency main effect in lexical decision, sensitivity to orthographically defined syllable-like structures in lexical decision, and an interaction of number of syllables with word frequency in naming.

Reading is a highly practised skill that seems almost effortless. Nevertheless, there is little consensus about the processes that a skilled reader uses when reading. The formal study of reading includes several specialized topics, each with a distinct literature: letter recognition, word identification, sentence understanding, and text comprehension.

Over the last two decades, the word-identification literature has established several benchmark phenomena in naming and lexical-decision tasks. For example, highfrequency words are named more quickly than low-frequency words, but the advantage is smaller for words whose pronunciation is consistent with spelling-to-sound rules (regular words) than for those whose spelling is not (irregular words). Second, although participants can confirm the lexical status of high-frequency words faster than lowfrequency words, the corresponding interaction of frequency with spelling-to-sound consistency is attenuated in the lexical-decision task (e.g., Seidenberg, Waters, Barnes, & Tannenhaus, 1984).1 In addition, words with many orthographic neighbours are named more quickly than words with few neighbours, and nonwords with many word neighbours are slower to reject in lexical decision (e.g., Andrews, 1989).

Over the same period, the literature has converged on a conceptual framework within which to discuss accounts of word identification. The framework is based on stimulus classification using a detector as a processing unit. A typical account might include a detector for subletter visual features, letters, groups of letters, phonemes, and words. Each detector is activated either when appropriate conditions occur in the environment or when activation is passed to it from detectors earlier in the processing chain. Although the details differ from account to account, there is wide agreement that activation of appropriate detectors is the currency of processing, in either a localist or distributed system. In McClelland and Rumelhart's (1981) words: "For every relevant unit in the system we assume there is an entity called a node" (p. 378). Likewise, Seidenberg and McClelland (1989) note that "each word processing trial begins with the presentation of a letter string, which the simulation program then encodes into a pattern of activation over the orthographic units" (p. 527). Naming a word is achieved by activating the appropriate phonological units. Early models hard wired the orthographic-to-phonological connections, whereas later models allowed the system to learn the appropriate connections.

Naming exhibits rule-like regularities: For example, words that conform to spelling-to-sound rules are named faster than those that do not follow the rules; that is, naming is slowed when the reader cannot take advantage of spellingto-sound rules. Connectionist theorists argue, however, that rule-like behaviour can be derived, without a rule-based processor, from interactions among detectors during stimulus classification (Plaut, McClelland, Seidenberg, & Patterson, 1996; Seidenberg & McClelland, 1989).2 By contrast, Coltheart (1978) staunchly defends the idea of phonological activation by a separate rule-based processor in addition to direct activation as part of the stimulus classification process (Coltheart, Curtis, Atkins, & Haller, 1993). …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.