Neural coding of formant-exaggerated speech in the infant brain
Version of Record online: 23 NOV 2010
© 2010 Blackwell Publishing Ltd
Volume 14, Issue 3, pages 566–581, May 2011
How to Cite
Zhang, Y., Koerner, T., Miller, S., Grice-Patil, Z., Svec, A., Akbari, D., Tusler, L. and Carney, E. (2011), Neural coding of formant-exaggerated speech in the infant brain. Developmental Science, 14: 566–581. doi: 10.1111/j.1467-7687.2010.01004.x
- Issue online: 8 APR 2011
- Version of Record online: 23 NOV 2010
- Received: 29 September 2009 Accepted: 10 July 2010
Supplementary Figure 1 Schematic illustration of two stimulus presentation paradigms. (a) In the alternating block design, each block contains 20 stimuli of one category (A) followed by a block that contains stimuli of a different category (B). The blocks are sequentially alternated to collect sufficient trials for both stimuli. (b) In the bidirectional MMN paradigm, two long blocks are presented to represent change directions (i.e. from A to B or B to A). The MMN response to stimulus change is derived by subtracting the brain responses to the same stimuli to control for inherent acoustic differences between A and B (i.e. deviant A minus standard A and deviant B minus standard B). In theory, the bidirectional MMN paradigm allows the study of neural representations of the auditory stimuli as well as sensitivity of the neural system to change detection (Zhang et al., 2005). However, the time length that is required to collect sufficient trials in the bidirectional MMN design would exceed the limit for most infant participants at 6–12 months of age.
Supplementary Sounds 1 and 2 The sound1.wav file was the non-exaggerated /i/. The sound2.wav file was the formant-exaggerated /i/. A preliminary survey on 15 adult native English speakers (all were naïve listeners who had not been previously tested on synthetic speech in laboratory settings) indicated that both sounds were heard as /i/s. Twelve listeners reported the /i/ with expanded formants to be clearer or perceptually more exaggerated than the other /i/ sound. The other three did not give unequivocal judgment about which sound was perceptually clearer or more exaggerated, which was probably due to unfamiliarity with synthetic speech.
Supplementary Movies 1 and 2 The MNE (minimum norm estimate) movie files were generated in BESA software and converted to compressed format. The movie1.wmv file showed left-hemisphere MNE activities for the exaggerated /i/, which were scaled to the maximum MNE peak in percentages and color-coded. The movie2.wmv file showed right-hemisphere MNE activities for the exaggerated /i/ using the peak-scaled color coding scheme. A cautionary note is necessary here. As the peak MMN values for different stimuli were not the same, a direct comparison of the color-coded brain activation patterns could lead to misinterpretation.
|DESC_1004_sm_fS1.tif||3048K||Supporting info item|
|DESC_1004_sm_movie1.wmv||763K||Supporting info item|
|DESC_1004_sm_movie2.wmv||748K||Supporting info item|
|DESC_1004_sm_sound1.wav||17K||Supporting info item|
|DESC_1004_sm_ssound2.wav||17K||Supporting info item|
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.