Review
‘Primitive intelligence’ in the auditory cortex

https://doi.org/10.1016/S0166-2236(00)01790-2Get rights and content

Abstract

The everyday auditory environment consists of multiple simultaneously active sources with overlapping temporal and spectral acoustic properties. Despite the seemingly chaotic composite signal impinging on our ears, the resulting perception is of an orderly ‘auditory scene’ that is organized according to sources and auditory events, allowing us to select messages easily, recognize familiar sound patterns, and distinguish deviant or novel ones. Recent data suggest that these perceptual achievements are mainly based on processes of a cognitive nature (‘sensory intelligence’) in the auditory cortex. Even higher cognitive processes than previously thought, such as those that organize the auditory input, extract the common invariant patterns shared by a number of acoustically varying sounds, or anticipate the auditory events of the immediate future, occur at the level of sensory cortex (even when attention is not directed towards the sensory input).

Section snippets

Central sound representation: underlying sound perception and transient sensory memory

On the basis of a large number of MMN studies, Näätänen and Winkler 6 have concluded that the neural traces involved in MMN generation in the auditory cortex encode the specific sensory information that appears in our auditory perception and sensory memory; that is, they carry the central sound representation (CSR). The authors have further proposed that the transient build-up phase of CSR underlies the perception of a (brief) sound and the slowly decaying phase of CSR the transient sensory

Auditory change detection

The MMN data (reviewed in Box 1) suggest that the auditory cortex maintains the representation of the immediate auditory past. Furthermore, each sound is automatically compared with these representations, and if some of the regularities are violated, then MMN, which represents the difference signal, is generated (see Box 1). This auditory cortex process itself is pre-perceptual but tends to trigger frontal cortex activity (reflected by the frontal MMN subcomponent 7), which probably underlies

Auditory stream formation and segregation

Typically, we receive acoustic information that originates from several simultaneous sound sources. Our central auditory system, therefore, has to segregate this mixture of the concurrent sound streams from each other and to attribute them to their original sources. The pitch, timbre, and location of the origin of the sounds, for example, provide cues used in stream segregation 1, which results in multiple perceptually separate sound streams. This process of assigning sensory inputs to the

Sound object formation

Sussman et al. have demonstrated pre-attentive auditory grouping within a single sound source 14. They presented a repetitive sequence of five tones (AAAAB) at a slow pace to subjects who were ignoring the tones (SOA=1.3 s). MMN was elicited by the frequency-deviant B tone, but it disappeared when the tone sequence was presented at a fast pace (SOA=100 ms). This suggests that the five tones were pre-attentively grouped together as a single perceptual unit, or sound object, when presented at the

Pre-attentive sound anticipation by transient extrapolatory traces

Even when subjects perform a task unrelated to the sound, MMN can be elicited by a steadily descending tone sequence that is interrupted by an ascending tone, and even by tone repetition 15 (Fig. 1). This result suggests that the auditory cortex does not only model the immediate auditory past but also forms extrapolatory traces on the basis of the regularities or trends detected in the auditory past, that is, these traces can represent anticipatory sound events 4, 17.

Extracting abstract sound patterns from the ongoing stimulus sequence

Saarinen et al. presented tone pairs to their subjects when they were reading a book 18. The standards were ascending pairs (that is, rising in frequency from the first to the second tone of the pair), whereas the deviant pairs were descending ones. Maintaining the frequency ratio between the first and second tone of the pairs, they randomly occurred on five different frequency levels, thus forming no physically constant standard pair, but rather a higher-order, ‘abstract’ standard, the

Sound categorization on the basis of relationships between sound attributes

Very recently, Paavilainen et al. found that the pre-attentive sound analysis mechanisms are even capable of extracting invariant abstract relationships between two sound features 21. Their standard stimuli continuously varied over a large range in frequency and intensity, so that there was no physically constant, repetitive standard stimulus or feature conjunction. Instead, the various examplars of the ‘standard’ stimulus obeyed a rule such as ‘the higher the frequency, the louder the

Permanent sound recognition and discrimination

MMN does not only reflect transient sensory traces but also more permanent ones. This is illustrated in Fig. 2, which shows that MMN that is not initially elicited by a change in a complex spectro–temporal pattern, develops in the course of the session, as subjects learn to discriminate these changes 22. It appears that this MMN elicitation is due to the gradual development of an accurate enough representation, or trace, for the complex spectrotemporal pattern. (Consistent with this, the

Categorical sound perception guided by recognition traces (for example, speech perception)

The existence of language-specific memory traces (in the auditory cortex) was demonstrated by Näätänen et al.27 (see also Ref. 28). They found that Finnish subjects’ MMN to an occasional replacement of the repetitive Finnish (and Estonian) vowel /e/ was larger when the deviant sound was a vowel in Finnish (/ö/) than when it was a vowel in Estonian only (/õ/). By contrast, both deviant sounds elicited enhanced MMNs of about equal amplitude in Estonian subjects, in whose mother tongue both are

Concluding remarks

In summary, this article has reviewed MMN (and MMNm) evidence to suggest the existence of different forms of pre-attentive cognitive operations in audition. It is of considerable interest that these cognitive processes occur, mostly, at the level of the auditory cortex (most probably in the secondary and association areas 27, 29, 30) where we can observe these manifestations of primitive sensory level intelligence, such as anticipation of next stimuli, speech sound perception (irrespective of

Acknowledgements

Our research is supported by the Academy of Finland, National Institutes of Health (R55 DC04263), and Hungarian National Research Fund (OTKA T022681).

References (33)

  • I. Winkler et al.

    Mismatch negativity: deviance detection or the maintenance of the ‘standard’

    NeuroReport

    (1998)
  • R. Näätänen et al.

    The concept of auditory stimulus representation in cognitive neuroscience

    Psychol. Bull.

    (1999)
  • C. Escera

    Involuntary attention and distractibility as evaluated with event-related brain potentials

    Audiol. Neuro-Otol.

    (2000)
  • M.H. Giard

    Brain generators implicated in the processing of auditory stimulus deviance; a topographic event-related potential study

    Psychophysiology

    (1990)
  • E. Sussman

    An investigation of the auditory streaming effect using event-related brain potentials

    Psychophysiology

    (1999)
  • E. Sussman

    Attention affects the organization of auditory input associated with the mismatch negativity system

    Brain Res.

    (1998)
  • Cited by (0)

    View full text