Are auditory representations a result of temporal predictions?

Last month an interesting review was published in the journal Trends in Cognitive Sciences arguing that ‘predictive representations of temporal regularities constitute the core of auditory objects in the brain.’ A possible consequence of this argument is that auditory sensory memory and (temporal) predictions are simply two sides of the same coin.

The authors (among which István Winkler and Sue Denham that collaborated with our Amsterdam group in the EmCAP project; see earlier blogs), review much of the recent literature using brain imaging and electrophysiological techniques. They support their hypothesis on the basis of at least five observations (and I paraphrase the authors here):

First, auditory regularity representations are temporally persistent; they have been shown to connect sounds separated by up to circa 10 seconds and persist for at least 30 seconds.

Second, auditory regularity representations encode all sound features with a resolution comparable to perception, since perceptually discriminable deviations elicit a Mismatch Negativity (MMN).

Third, when two sound streams are perceptually separated, MMN reflects the perceived sound organization, its elicitation dynamically follows perceptual fluctuations between two alternative sound organizations and the effects of priming sequences on perception.

Fourth, regularities are extracted from acoustically widely different exemplars in a sequence, including the natural variation of environmental sounds.

And finally, violations of predictive rules have been shown to elicit the MMN. For example, delivering a low tone after a short one elicited the MMN, when for most tones the rule “short tones are followed by high-pitched tones, long tones by low-pitched tones” held.

Interestingly, violations in the form of silence (i.e. no sound) - such as omissions in a natural drum-pattern - also show a MMN. And in addition, these effects are also found when attention is directed to other aspects than the sound /music or when participants are unattentive (such as in the case with sleeping neonates).

Winkler, I., Denham, S., & Nelken, I. (2009). Modeling the auditory scene: predictive regularity representations and perceptual objects Trends in Cognitive Sciences, 13 (12), 532-540 DOI: 10.1016/j.tics.2009.09.003