Interference between phonemes during monitoring: Evidence for an interactive activation model of speech perception.

Abstract

We explore the recent finding (Newman & Dell, 1978) that the time needed to detect a target phoneme in a phoneme monitoring task is increased when the preceding word contains a phoneme similar to the target. Normal adult native speakers of English monitored auditorily presented sentences and responded as quickly as possible whenever they detected a specified phoneme. We found that preceding word-initial phonemes, despite being processed more quickly, increased the response latency to the following target phoneme more than did preceding word-medial phonemes. There was also an increase in response latency even when the subject could be highly certain that the similar preceding phoneme was not an instance of the target phoneme. We argue that the interference effects are due to fundamental characteristics of perceptual processing and that more time is needed to categorize the target phoneme. We present a computer simulation using an interactive activation model of speech perception to demonstrate the plausibility of our explanation.

Meaning, sound, and syntax: Lexical priming in sentence production.

“…In both experiments there was a weak tendency for phonologically primed words to occur later in the descriptions than unprimed words. Because there are reasons to predict that hearing or producing a word results in the inhibition of phonologically related words (McClelland & Rumelhart, 1981), and because there are scattered suggestions of articulatory or phonological inhibition resulting from word repetition and phonological similarity in a variety of tasks (Crowder, 1978;Durso & Johnson, 1979;Meyer & Gordon, 1985;Nairne & Healy, 1983;Slowiaczek & Pisoni, 1985;Stemberger, Elman, & Haden, 1985), the late appearance of phonologically primed words was perhaps due to decreased accessibility relative to the unprimed words. However, firmer evidence for effects of this sort is clearly needed.…”

Spoken Word Recognition and Production

“…Newman and Dell argued that a phonological representation constructed postlcxically should allow a simple yes-no décision on match of input to target; but RT in their study showed a gradient of interference as a function of number of shared phonological features. Further, this interference effect is as strong in pre dictable as in unpredictable words (Dell fk Newman, 1980;Stemberger, Elman, & Haden, 1985).…”

Lexical effects in phonemic processing: Facilitatory or inhibitory?

“…In particular, they claimed that their race model with two response "outlets" straightforwardly allows attentional shifts caused by the specific properties of the experimental situation to control the outcome of the race. In contrast, interactive models, such as the one proposed by Stemberger et al (1985) with a single response "outlet" at the phonemic level, cannot appeal to shifts between outlets. Such connectionist models can account for the variability in lexicality effects by modifying the weightings of bottom-up and top-down connections as a function of the task demands.…”

“…Hence, the recognition of a phoneme (i.e., attaining a criterial level of activation with respect to other phonemes) depends upon the excitatory activation it receives from the feature level and from the lexical level as well as upon the inhibition from other phonemes. Stemberger, Elman, and Haden (1985) claimed that subjects responding in the phoneme-monitoring task make direct and exclusive use of activated phoneme units. McClelland and Elman (1986) explained the presence or absence of lexical effects in phoneme monitoring within the TRACE framework by assuming that lexical feedback depends upon the level of activation of the lexical units containing the target; only when there is sufficient lexical activation do lexical effects emerge in phoneme-detection latencies.…”