Salience of the word as a unit in the perception of language

Abstract

Seven interlocking experiments are reported in which both guessing and recognition thresholds for words are compared with those for other linguistic units both smaller than (nonword morphemes and trigrams) and larger than (nominal compounds, ordinary noun phrases, and nonsense compounds) the word. Thresholds were consistently lower for words than for morphemes or trigrams (matched or even much higher in visual usage frequency) and lower for word-like nominal compounds (e.g., stumbling block) than for ordinary noun phrases (copper block) or nonsense compounds (sympathy block). Prior exposure (through two correct recognitions) to ordinary noun phrases, nonsense compounds, and the constituent single words of nominal compounds significantly facilitated subsequent recognition of the single-word constituents, but prior exposure to nominal compounds had no effect whatsoever on subsequent recognition of their single-word constituents. These results as a whole are interpreted as supporting the following conclusions: (l) that the word has special salience in the perception of language; (2) that the reason for this salience is the unique meaningfulness of the word (or the word-like nominal compound) as a whole; and (3) that the mechanism for this salience is the convergence of feedback from central mediational processes with feed-forward from peripheral sensory processes upon the integration of word-form percepts.The first author (Osgood, 1963, and elsewhere) has argued that the word has special status as a psycholinguistic unit, being simultaneously the largest characteristic unit at the meaningless sensory integration level and the smallest characteristic unit at the meaningful representational level. For language perception, the word meets the requirements for sensory integration as a closure mechanism-redundancy of the same sensory "whole" in the Gestalt sense, relatively high frequency of occurrence, and temporal duration well within the limits of cell-assembly reverberation (cf. Hebb, 1949). For language understanding, the word is the smallest meaningful unit that accepts almost unlimited insertions at its boundaries (cf. Greenberg's essay on the linguistic definition of the word, 1957) and hence can combine with other words to form an infmite variety of sentences.Many experiments with the tachistoscope have testified empirically to an inverse relation between frequency of usage of words and their duration thresholds for recognition (e.g., Howes & Solomon, 1951; Solomon & Postman, 1952;King-Ellison & Jenkins, 1954;Postman & Rosenzweig, 1957). This relation holds when the frequencies of usage are for word forms, usually as determined from Thorndike-Lorge tables, independent of their multiple senses or meanings. At the level of sensory integration, the word is thus an empty or meaningless form, capable of multiple semantic codings depending upon linguistic context.There are other identifiable units in the surface forms of printed language, both larger and smaller than the word. Larger units include sentences, clauses...

Formulaic Language in Parkinson's Disease and Alzheimer's Disease: Complementary Effects of Subcortical and Cortical Dysfunction

“…Following James (1895), Bousfield (1953), andMiller (1956), Simon's (1974) article described memorial "chunks" as units of processing: When people are asked to remember and recall a spoken list of items, the number of recallable items-or chunks-is nearly the same for syllables, words, compound words, and idiomatic phrases. Psychological studies reveal reliable performance differences between formulaic and novel exemplars, attesting to the holistic structure of the formulaic expressions in the speaker's competence as contrasted with compositional processing of matched novel expressions (Conklin & Schmitt, 2008;Horowitz & Manelis, 1973;Jiang & Nekrasova, 2007;Libben & Titone, 2008;Osgood & Housain, 1974;Pickens & Pollio, 1979;Swinney & Cutler, 1979). These studies are supported in the visual mode by results demonstrating the significant perceptual effects of processing known holistic configurations in contrast to constituent details (Poljac, de-Wit, & Wagemans, 2012;Pomerantz, 1986;Pomerantz, Sager, & Stoever, 1977).…”

Processing times for English and Chinese words

“…In other words, the Chinese characters form two Gestalt wholes at the level of sensory integration (cf. Osgood & Hoosain, 1974), rather like English nominal compounds such as "post card." It could be these separate Gestalt wholes that prolong the pronunciation durations of the Chinese items.…”

Reenacting the route to interpretation: Enhanced perceptual identification without prior perception.

“…As a further test of this hypothesis we turned to another set of materials that were expected to produce an encoding that strongly integrated a target word with its context. In Experiment 9 word pairs consisting of a noun and a unique modifier (e.g., flying saucer and stool pigeon), like those used by Osgood and Hoosain (1974) and Schacter and McGlynn (1989), were used to encourage integrative encoding. For such items the interpretation of the modified noun is heavily influenced (usually toward an unusual sense) by the modifier.…”