Neural correlates of attending speech and non-speech: ERPs associated with duplex perception
Introduction
The question whether a special module is responsible for speech perception or whether the transformation of acoustical information into speech segments is carried out by general auditory processing centres is a central issue in speech perception. Speech is special theorists assume that a speech-specific module transforms highly variable acoustical information into invariant articulatory sequences for speech processing (Liberman & Mattingly, 1989; Mattingly, 1991). The use of the perceiver's speech-motor system for speech recognition precludes the use of general acoustical pattern matching modules for the speech recognition task. An important piece of evidence in support of the ‘speech is special’ argument is that certain stimuli (duplex sounds, see below) simultaneously elicit a speech and a non-speech percept which is consistent with the hypothetical parallel use of two separate processing modules: a general auditory module for non-speech sounds and a specific speech recognition module that uses articulatory pattern generators as the basis for pattern matching.
Section snippets
Duplex perception
The phenomenon of duplex perception (DP) was first reported by Rand (1974), and it occurs when a single formant transition of a synthetic syllable is isolated and presented to one ear, while the remainder of the syllable (usually called the ‘base’) is presented to the opposite ear (dichotic presentation). Listeners report hearing a non-speech ‘chirp’ in the ear receiving the transition, while, at the same time, hearing an intact syllable in the other ear. It is the dual role of the isolated
Speech and non-speech mode of perception
Assuming that a phonetic module exists separately from an auditory module, it is expected that different neural activity would be found depending upon which of the modules was activated. Such differences may be found by examining event-related potentials (ERPs), which can provide insight into the underlying neural events that occur as a result of cognitive processing. Electrophysiological studies in humans have shown that ERPs can reflect the selective processing of stimuli in auditory, visual
The present study
The main purpose of this experiment was to study the possible differences in the processing of speech vs. non-speech within an ERP framework by presenting the same duplex stimulus in two different conditions: (a) listeners attending to the vowel–nasal syllable, or (b) listeners attending to the chirp. Thus, while the attentional demands are maintained in both conditions, any difference found in the ERPs will be due to differences in the endogenous, perceptual processing of the different parts
Participants
The sample consisted of 12 unpaid volunteers (5 female) between 18 and 42 years (mean=24.5 years). Two participants were excluded because excessive artefact led to more than 45% of EEG trials to be rejected. All listeners were English speakers and right handed, with no history of neurological disorder or hearing loss. All subjects were able to discriminate the syllables and identify the chirp well above chance level (acceptance criterion: accuracy ⩾75%).
Stimuli
The speech stimuli used in the study were
Behavioural data
The tasks were relatively easy with participants correctly identifying the chirp and the /em/ syllable in 87.82% of the trials in both non-speech and speech conditions (see Table 3). There was no significant difference in the performance levels for the two tasks [F(1, 78)=2.03; p>0.05].
For stimulus 1 (control /en/) the number of false alarms among conditions was 3.02% for the non-speech condition, and 12.54% for the speech condition. None of these rates differed significantly between conditions
Discussion
The two specific goals of this research were, firstly, to examine whether attending to either the speech or non-speech part of the same duplex stimulus modulated the ERPs. Secondly, to see if these different low-level ‘attentional auditory modes’ were associated with any lateralisation differences between hemispheres. Statistical analysis of the voltage distributions and dipole analysis suggest that the ERP components elicited by the speech task were different from those of the non-speech task.
Conclusion
The current study overcomes stimulus presentation (L/REA effects) by presenting diotically the same physical duplex stimulus in two different conditions: one where listeners have to process the stimulus in ‘speech mode’ and the other where a non-speech aspect of the stimulus has to be identified. The current study overcomes the problem of matched physical properties by binaurally presenting the same vowel–nasal stimuli in different tasks.
Our results show task-dependent modulation of relatively
Acknowledgements
We wish to thank Jasna Martinovic, and Harry Sumnall for critical readings of a draft of this article, and Guillaume Thierry and Horacio Barber for useful advice and methodological suggestions. Special thanks to Takako Saito, Emi Yamada and Shozo Tobimatsu for their kind and generous support. We would also like to thank two anonymous referees for their helpful comments and constructive critique. This research was supported by HOARSE grant HPRN-CT-2002-00276.
References (58)
- et al.
Is the sine-wave speech cocktail party worth attending?
Speech Communication
(1999) - et al.
Conflict monitoring and anterior cingulate cortex: An update
Trends in Cognitive Sciences
(2004) - et al.
Neural correlates of switching from auditory to speech perception
NeuroImage
(2005) - et al.
Neurological evidence in support of a specialized phonetic processing module
Brain and Language
(2001) - et al.
Event-related potentials during selective attention to speech sounds
Biological Psychology
(1983) - et al.
Changes in the perception of synthetic nasal consonants as a result of vowel formant manipulations
Speech Communication
(2003) - et al.
Towards a functional neuroanatomy of speech perception
Trends in Cognitive Sciences
(2000) - et al.
Are different kinds of acoustic features processed differently for speech and non-speech sounds?
Cognitive Brain Research
(2001) - et al.
The effect of different noise types on the speech and non-speech elicited mismatch negativity
Hearing Research
(2005) - et al.
The electrophysiological net response (‘F-complex’) to spatial fusion of speech elements forming an auditory object
Clinical Neurophysiology
(2003)