Common and modality-specific processes in the mental lexicon

Abstract

Eight experiments were conducted to resolve: (1) empirical inconsistencies in repetition effects under intermodality conditions in word identification and lexical decision, and (2) an associated theoretical conflict concerning lexical organization. The results demonstrated that although more facilitation occurs under visual-prime!visual-test (VV) conditions than under auditory-prime/visual-test (AVI conditions, significant repetition facilitation also occurs under AV conditions. The results also indicated that: repetition effects observed for the VV and AV conditions apply to high-as well as to low-frequency words; they are insensitive to a variety of encoding tasks designed to emphasize different properties of words; and they are unaffected by differences in the ease of encoding of isolated auditory and visual words. The results are consistent with the existence of both modality-specific and common or modality-free processes in word recognition, in which word-frequency effects are restricted to the second and, by implication, lexical stage.Research into modality differences has focused on a variety of empirical issues over the last 2 decades. For much of this time, interest has focused on episodic questions concerning, for example, the distinctions among sensory, short-term, and long-term memory systems. At the heart of this problem, however, is the modal specificity of the system responsible for the transition from sensory to categorical representation. If categorical information emanates from a single lexicon, albeit via distinct access routines, it follows that equivalent item information will be available following auditory and visual presentation. In this case, then, modality information must be stored dependently, as an associate of the item information. But if distinct lexical systems are involved in reading and in speech recognition, the item information available following auditory and visual presentation may not be equivalent. In this case, modality information may be provided automatically by the item information.Research into the modal specificity of lexical processes has not yielded a definitive answer. One approach to the question involves observation of repetition effects in word-processing tasks. When equivalent repetition effects are observed under inter-and intramodality conditions, it is assumed that a common system is involved. But when the repetition effects are restricted to intramodality conditions, it is assumed that separate mechanisms are involved in text and speech recognition. sented in Morton's (1979) results and is reflected in his model. Using word recognition, with threshold-estimation procedures in vision and audition, he found no evidence of transfer under intermodality conditions. The model that he subsequently developed includes modality-specific word recognition units, or "logogens." According to Morton, these units are not connected except through a separate cognitive system, so intermodality transfer is absent in tasks that tap only word identification processes. Any task...

Bias in masked word identification: Unconscious influences of repetition priming

“…The naming version of the masked word identification test replicated earlier demonstrations of modality specificity in repetition priming on this task (e.g., Jacoby & Dallas, 1981;Kirsner et al, 1983). As indicated in the figure, auditory study led to only a very small, nonsignificant improvement in performance, whereas visual study produced a robust improvement.…”

“…Tests that primarily depend on conscious recollection of an item's prior occurrence show little sensitivity to changes in modality across study and test, whereas tests that involve target identificationwithout the requirement of conscious recollection are quite sensitive to changes in modality (e.g., Rajaram & Roediger, 1993). In the case of masked word identification, Clarke and Morton (1983), Jacoby and Dallas (1981), Kirsner, Milech, and Standen (1983), Levy and Kirsner (1989), and Weldon (1991) all obtained more repetition priming when words were studied visually than when they were studied auditorily. The forced-choice variant of masked word identification in question here presumably expresses the influence of prior study through the same mechanism as the standard naming version used in the experiments in which modality effects have been found (see Ratcliff & McKoon, 1997).…”

The effects on priming of word frequency, number of repetitions, and delay depend on the magnitude of priming

“…Such considerations may also be important in the evaluation of the impact on priming of manipulations of the study task. For instance, a change in modality from study to test often severely reduces, indeed sometimes completely removes, priming effects, whereas such a change has little or no effect on explicit-memory measures (see, e.g., Kirsner, Milech, & Standen, 1983). It could be argued that these modality effects on priming are in conflict with the present argument that priming variability is often too constrained for the effects of other variables to be observed.…”

On the relationship between reading, listening, and speaking: It’s different for people’s names

“…Jacoby and Dallas (1981;Experiment 6) found significant facilitation of accuracy of visual iden-tification of briefly exposed words (35 msec) from prior study of visual words but no facilitation from prior study of auditory presentation of words. Kirsner and Smith (1974) and Kirsner, Milech, and Standen (1983) found that auditory presentation of primes produced significant priming of visual lexical decisions, but the priming effect was significantly smaller than that obtained from visual presentation ofprimes. Kirsner et al interpreted their results in terms of two loci of repetition priming: a modality-specific component that reflects facilitation ofword identification and a modality-free component that is associated with access to semantic representations.…”