Facilitated Auditory Detection for Speech Sounds

Abstract

If it is well known that knowledge facilitates higher cognitive functions, such as visual and auditory word recognition, little is known about the influence of knowledge on detection, particularly in the auditory modality. Our study tested the influence of phonological and lexical knowledge on auditory detection. Words, pseudo-words, and complex non-phonological sounds, energetically matched as closely as possible, were presented at a range of presentation levels from sub-threshold to clearly audible. The participants performed a detection task (Experiments 1 and 2) that was followed by a two alternative forced-choice recognition task in Experiment 2. The results of this second task in Experiment 2 suggest a correct recognition of words in the absence of detection with a subjective threshold approach. In the detection task of both experiments, phonological stimuli (words and pseudo-words) were better detected than non-phonological stimuli (complex sounds), presented close to the auditory threshold. This finding suggests an advantage of speech for signal detection. An additional advantage of words over pseudo-words was observed in Experiment 2, suggesting that lexical knowledge could also improve auditory detection when listeners had to recognize the stimulus in a subsequent task. Two simulations of detection performance performed on the sound signals confirmed that the advantage of speech over non-speech processing could not be attributed to energetic differences in the stimuli.

Similarities in the neural signature for the processing of behaviorally categorized and uncategorized speech sounds

“…Finally, the results of the present study showed that lexical processing could be observed for stimuli that are not behaviorally categorized, but partially detected. Although the choice of a lexical task as a measure of awareness is largely accepted in studies investigating linguistic processing without awareness (for reviews, see Kouider et al ., ; Van den Bussche et al ., ), further research would be required to assess whether lexical processing could also occur in experimental conditions involving a detection task, as recently suggested in a behavioral study showing lexical processing without subjective detection (Signoret et al ., ).…”

“…This result is in accordance with studies done in the visual modality suggesting that the lexical or semantic processing can occur without awareness (for a meta‐analysis, see Van den Bussche et al ., ). This result is also in accordance with behavioral studies done in the auditory modality showing lexical processing without awareness (Daltrozzo et al ., ; Signoret et al ., ), and suggests that lexical processing is concomitant with a variation of low‐beta (15–20 Hz) power near the left temporal regions, independently of any awareness of the category of the stimulus.…”

“…Like the words, the pseudo‐words were monosyllabic, phonotactically legal and pronounceable, but meaningless combinations of sounds in French. Overall, the pseudo‐words contained all the phonemes of the words, as done in previous studies investigating speech perception (Daltrozzo et al ., ; Signoret et al ., ). The average number of phonemes of the pseudo‐words was carefully matched to that of the words (2.3 for the words against 2.4 for the pseudo‐words; t 718 = −1.52, P > 0.125).…”

“…In the visual modality, linguistic processing of stimuli presented below categorization threshold could occur at the phonological and semantic levels (for reviews, see Kouider et al ., ; Van den Bussche et al ., ). In the auditory modality, recent studies show evidence of lexical processing for stimuli presented below categorization (Kouider & Dupoux, ), and below detection (Signoret et al ., ) thresholds. Semantic processing of auditory stimuli has also been observed below the auditory categorization threshold (Daltrozzo et al ., ).…”

“…Silences, words, pseudo‐words and complex sounds were used. Words, pseudo‐words and complex sounds were ‘energetically matched’, that is they had the same long‐term spectra and the same envelope fluctuations (see Signoret et al ., for details). These auditory stimuli were presented at two intensity levels of presentation: one at which participants could perform a categorization task at both the lexical (differentiating between word and pseudo‐word) and the phonological (differentiating between pseudo‐word and complex sound) levels; and the other at which they were unsuccessful in the categorization task at the phonological and lexical levels, similarly to Daltrozzo et al .…”