Evidence for a parallel input serial analysis model of word processing.

Abstract

A parallel input serial analysis (PISA) model of word processing was developed and tested. The goal was to expand on the "critical processing duration" hypothesis of Johnson, Allen, and Strand (1989) so that both single-word and multiple-word presentation, letter detection data could be explained. In Experiments 1-3 four different word frequency categories on a single-presentation, letter detection task were used. These three experiments indicated that there was a curvilinear relationship between word frequency and letter detection reaction time (RT). That is, letter detection RTs for medium-high-frequency words were significantly longer than letter detection RTs for very-high-, low-, and very-low-frequency words. These results support the PISA model rather than the Healy, Oliver, and McNamara (1987) version of the unitization model. In Experiments 4-5 multiple-presentation (i.e., two words), letter detection tasks were used. The PISA model could also account for the results from these two experiments, but the unitization model could not.

The Nature of Character-Component Interaction in Chinese Character Perception

“…The results of the two experiments indicate that the medium-high frequency disadvantage that Allen and his colleagues found in the letter detection RTs for English words (Allen & Madden, 1990;Allen & Emerson, 1991) could not be established with Chinese characters. A high-frequency disadvantage was demonstrated for medium-or high-frequency components, while a medium-frequency advantage was found for lowfrequency components in Experiment 1.…”

“…

The PISA (parallel input serial access) model assumes independence and competition of word-level and letter-level processing (Allen & Madden, 1990;Allen & Emerson, 1991), while the IA (interactive activation) model proposes that wordlevel and letter-level processing interactively exchange information (McClelland & Rumelhart, 1981). This current study attempts to explore which of these two views characterizes the character-component relationship in Chinese character recognition using the component detection task.

…”

“…A more sophisticated version of the race model -the PISA (parallel input serial access) model (Allen, Benjamin, & Weber, 1995;Allen & Emerson, 1991;Allen & Madden, 1990) additionally proposed a central processor that decodes from the winning channel. Since resources of the central processor are limited, a delay in processing one of the two codes may result when the word-and the letter-level processing are approximately the same in their speeds.…”

“…It was found that an increase in word frequency results in a greater likelihood of missing a target letter during proofreading (Drewnowski & Healy, 1977;Healy & Drewnowski, 1983). Subsequent studies discovered a curvilinear relationship between letter detection RTs and word frequency whereby words of intermediate frequency took longer to process than words at the frequency extremes (Allen & Madden, 1990;Johnson, Allen, & Strand, 1989). This nonlinear word frequency effect was not paralleled by participants' performance on LDT (lexical decision task), as lexical decision times have been shown to decrease monotonically with the increase in word frequency (Allen & Emerson, 1991).…”

Processing of Kanji and Kana characters within Japanese words

“…Michiko Shimomura is now at Nara National College of Technology, Nara, Japan. Correspondencemay be addressed to Kazuhiko Yokosawa, now at NTT Basic Research Laboratories, 3-9-11 Midoricho, Musashino-shi, Tokyo 180, Japan.19 Madden, 1990;Chambers & Forster, 1975) is that wordlevel processing and letter-level processing occur simultaneously, and that the word-level processing is completed earlier than the letter-level processing. Race models thus imply that the output of word-level processing is available before the output of letter-level processing.…”