Neuro-Physiological Correlates of Voice Onset Time in Kannada and Tamil Speaking

Abstract

Voice Onset Time (VOT) refers to the time difference between two events—one articulatory and the other laryngeal. VOT is measured as the time difference between the onset of the burst (articulatory event) and the onset of vocal fold vibration (laryngeal event) Lisker [14]. The cross-linguistic variation with respect to voicing perception is more in multilingual countries like India. For example, in Tamil language, there is no good distinction between voiced and unvoiced stops; however, in Kannada, Telugu, and Hindi, there exists a difference. To verify the cross-linguistic difference, the behavioral response for VOT continuum in native Kannada and Tamil speakers, and the neuro-physiological changes for VOT continuum in native Kannada and Tamil speakers were taken as objective of the study. Two groups of subjects were participated in the study. Group I consisted of ten male Kannada speakers in the age range of 20–35 years. All the subjects were native speakers of Kannada language, and they belong to Mysore dialect and they are fluent speakers in Kannada only. All though they were exposed to English, they were not fluent speakers of English or Hindi. Group II consisted of ten male Tamil speakers in the age range of 20–35. All the subjects were native speakers of Tamil language, and they belong to Coimbatore/Chennai dialect and they are fluent speakers in Tamil only. Although they were exposed to English, they were not fluent speakers of English or Kannada or Hindi. From the naturally recorded speech sounds, /da-ta/ continuum was created using copy past synthesis method. This continuum had 10 tokens. Using this stimulus, behavioral identification curve was generated. Also electrophysiological, N100 potential was recorded using Neuro-Scan instrument (Compumedies, AUS) with five tokens, which covers the entire dynamic range of the /da-ta/ continuum. The recording was done with 16 channels. The behavioral result showed there was difference between Tamil and Kannada languages; however, electrophysiological results showed that there was a change with N100 latency with changing VOT but there was no language difference found. The author concluded that N100 may not be an effective measure to indicate to represent the voice /voiceless categorical perception. May be higher potential might give better information.