Limits...
Visual Enhancement of Illusory Phenomenal Accents in Non-Isochronous Auditory Rhythms

View Article: PubMed Central - PubMed

ABSTRACT

Musical rhythms encompass temporal patterns that often yield regular metrical accents (e.g., a beat). There have been mixed results regarding perception as a function of metrical saliency, namely, whether sensitivity to a deviant was greater in metrically stronger or weaker positions. Besides, effects of metrical position have not been examined in non-isochronous rhythms, or with respect to multisensory influences. This study was concerned with two main issues: (1) In non-isochronous auditory rhythms with clear metrical accents, how would sensitivity to a deviant be modulated by metrical positions? (2) Would the effects be enhanced by multisensory information? Participants listened to strongly metrical rhythms with or without watching a point-light figure dance to the rhythm in the same meter, and detected a slight loudness increment. Both conditions were presented with or without an auditory interference that served to impair auditory metrical perception. Sensitivity to a deviant was found greater in weak beat than in strong beat positions, consistent with the Predictive Coding hypothesis and the idea of metrically induced illusory phenomenal accents. The visual rhythm of dance hindered auditory detection, but more so when the latter was itself less impaired. This pattern suggested that the visual and auditory rhythms were perceptually integrated to reinforce metrical accentuation, yielding more illusory phenomenal accents and thus lower sensitivity to deviants, in a manner consistent with the principle of inverse effectiveness. Results were discussed in the predictive framework for multisensory rhythms involving observed movements and possible mediation of the motor system.

No MeSH data available.


Results of the magnitude of IA.Group means of individually sorted unisensory and multisensory IA as a function of unisensory IA magnitude. 1 to 4 on the X-axis represent unisensory IA magnitude in a descending order. Error bars represent standard error of the means.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC5120798&req=5

pone.0166880.g006: Results of the magnitude of IA.Group means of individually sorted unisensory and multisensory IA as a function of unisensory IA magnitude. 1 to 4 on the X-axis represent unisensory IA magnitude in a descending order. Error bars represent standard error of the means.

Mentions: Following the d’ results supporting the PC hypothesis and the metrically induced IA, an additional index of the magnitude of IA was calculated for each individual as the difference in d’ between weak beat and strong beat positions (d’weak beat−d’strong beat), for each of the 2 (rhythm modality) × 4 (auditory interference) conditions. A greater value suggested a greater extent of IA that was specifically related to the metrical position. To examine the magnitude of multisensory compared to unisensory IA as a function of the unisensory IA magnitude, the same sorting logic as previously described was applied. Here, each individual’s IA scores in the AV conditions were sorted by the descending order of IA scores in the A conditions. The individual values of sorted IA were submitted to a 2 (rhythm modality) × 4 (auditory IA level) ANOVA, which revealed a significant interaction between the two factors, F(3, 57) = 16.35, p < 0.001, ηp2 = 0.42 (Bayes factor 8.5× 104: 1 in favor of the interaction). Follow-up partial ANOVAs conducted for each auditory IA level separately showed the following: For level 1 (greatest auditory IA), IA was lower in AV than in A, F(1, 19) = 27.28, p < 0.001, ηp2 = 0.59. For level 2, IA did not differ between AV and A, F(1, 19) = 0.40, p = 0.53, ηp2 = 0.02. For level 3 and level 4 (the two lowest auditory IA), IA was greater in AV than in A, F(1, 19) = 5.63, p = 0.03, ηp2 = 0.23, and F(1, 19) = 7.53, p = 0.01, ηp2 = 0.28 (Fig 6)


Visual Enhancement of Illusory Phenomenal Accents in Non-Isochronous Auditory Rhythms
Results of the magnitude of IA.Group means of individually sorted unisensory and multisensory IA as a function of unisensory IA magnitude. 1 to 4 on the X-axis represent unisensory IA magnitude in a descending order. Error bars represent standard error of the means.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC5120798&req=5

pone.0166880.g006: Results of the magnitude of IA.Group means of individually sorted unisensory and multisensory IA as a function of unisensory IA magnitude. 1 to 4 on the X-axis represent unisensory IA magnitude in a descending order. Error bars represent standard error of the means.
Mentions: Following the d’ results supporting the PC hypothesis and the metrically induced IA, an additional index of the magnitude of IA was calculated for each individual as the difference in d’ between weak beat and strong beat positions (d’weak beat−d’strong beat), for each of the 2 (rhythm modality) × 4 (auditory interference) conditions. A greater value suggested a greater extent of IA that was specifically related to the metrical position. To examine the magnitude of multisensory compared to unisensory IA as a function of the unisensory IA magnitude, the same sorting logic as previously described was applied. Here, each individual’s IA scores in the AV conditions were sorted by the descending order of IA scores in the A conditions. The individual values of sorted IA were submitted to a 2 (rhythm modality) × 4 (auditory IA level) ANOVA, which revealed a significant interaction between the two factors, F(3, 57) = 16.35, p < 0.001, ηp2 = 0.42 (Bayes factor 8.5× 104: 1 in favor of the interaction). Follow-up partial ANOVAs conducted for each auditory IA level separately showed the following: For level 1 (greatest auditory IA), IA was lower in AV than in A, F(1, 19) = 27.28, p < 0.001, ηp2 = 0.59. For level 2, IA did not differ between AV and A, F(1, 19) = 0.40, p = 0.53, ηp2 = 0.02. For level 3 and level 4 (the two lowest auditory IA), IA was greater in AV than in A, F(1, 19) = 5.63, p = 0.03, ηp2 = 0.23, and F(1, 19) = 7.53, p = 0.01, ηp2 = 0.28 (Fig 6)

View Article: PubMed Central - PubMed

ABSTRACT

Musical rhythms encompass temporal patterns that often yield regular metrical accents (e.g., a beat). There have been mixed results regarding perception as a function of metrical saliency, namely, whether sensitivity to a deviant was greater in metrically stronger or weaker positions. Besides, effects of metrical position have not been examined in non-isochronous rhythms, or with respect to multisensory influences. This study was concerned with two main issues: (1) In non-isochronous auditory rhythms with clear metrical accents, how would sensitivity to a deviant be modulated by metrical positions? (2) Would the effects be enhanced by multisensory information? Participants listened to strongly metrical rhythms with or without watching a point-light figure dance to the rhythm in the same meter, and detected a slight loudness increment. Both conditions were presented with or without an auditory interference that served to impair auditory metrical perception. Sensitivity to a deviant was found greater in weak beat than in strong beat positions, consistent with the Predictive Coding hypothesis and the idea of metrically induced illusory phenomenal accents. The visual rhythm of dance hindered auditory detection, but more so when the latter was itself less impaired. This pattern suggested that the visual and auditory rhythms were perceptually integrated to reinforce metrical accentuation, yielding more illusory phenomenal accents and thus lower sensitivity to deviants, in a manner consistent with the principle of inverse effectiveness. Results were discussed in the predictive framework for multisensory rhythms involving observed movements and possible mediation of the motor system.

No MeSH data available.