Perceptuomotor compatibility effects in vowels: Beyond phonemic identity

Abstract

Perceptuomotor compatibility between phonemically identical spoken and perceived syllables has been found to speed up response times (RTs) in speech production tasks. However, research on compatibility effects between perceived and produced stimuli at the subphonemic level is limited. Using a cue-distractor task, we investigated the effects of phonemic and subphonemic congruency in pairs of vowels. On each trial, a visual cue prompted individuals to produce a response vowel, and after the visual cue appeared a distractor vowel was auditorily presented while speakers were planning to produce the response vowel. The results revealed effects on RTs due to phonemic congruency (same vs. different vowels) between the response and distractor vowels, which resemble effects previously seen for consonants. Beyond phonemic congruency, we assessed how RTs are modulated as a function of the degree of subphonemic similarity between the response and distractor vowels. Higher similarity between the response and distractor in terms of phonological distance-defined by number of mismatching phonological features-resulted in faster RTs. However, the exact patterns of RTs varied across response-distractor vowel pairs. We discuss how different assumptions about phonological feature representations may account for the different patterns observed in RTs across response-distractor pairs. Our findings on the effects of perceived stimuli on produced speech at a more detailed level of representation than phonemic identity necessitate a more direct and specific formulation of the perception-production link. Additionally, these results extend previously reported perceptuomotor interactions mainly involving consonants to vowels.

Keywords: psycholinguistics; speech perception; speech production.

Fig. 1

An example of an RT measurement for a trial where the response is the vowel /u/ and the distractor is vowel /e/. The top audio channel shows the participant’s spoken response. The bottom audio channel shows the time line for the presentation of visual cue and the auditory distractor. The short vertical marker in the second channel is synchronous with the onset of the presentation of the visual cue. Stimulus onset asynchrony (SOA) is the latency between that onset and the onset of the auditory distractor (here, 150 ms). RT was calculated as the latency between (the time stamp of) the onset of the spoken response minus (the time stamp of) the onset of the visual cue

Fig. 2

Mean RTs (ms) showing phonemic congruency effects by distractor condition. The “All distractors” bar shows the mean RTs for all trials from every distractor condition (tone, congruent, and incongruent) pooled together. Error bars show 95% confidence intervals. Results are collapsed across SOA (the interactions of SOA and distractor conditions were not significant)

Fig. 3

Mean RTs (ms) showing effects of each distractor vowel for the responses /u/ (left) and /e/ (right). Error bars show 95% confidence intervals. Results are collapsed across SOAs, because the interactions of SOA and distractor conditions were not significant