Pitch Perception: Development

The perception of pitch is fundamental to a number of aspects of music, including melody, harmony, timbre, and the perception of simultaneous musical voices in polyphonic music. Pitch is not directly given in a sound stimulus, but is derived [Page 873]in the auditory pathway through spectrotemporal processing. Sounds produced by voices and musical instruments are complex, containing multiple frequencies. Although frequency information is extracted early in the auditory pathway, pitch representations require the auditory cortex, which is immature in young infants.

Sounds that give rise to a sensation of pitch typically contain energy at a fundamental frequency, F0, and at integer multiples of that fundamental called harmonics. The cochlea in the inner ear performs a sort of Fourier analysis whereby the frequency information in an incoming complex sound wave is decomposed into its frequency components. These frequencies are analyzed in tonotopic (spectral) maps through subcortical nuclei and into primary auditory cortex. There is much evidence from functional magnetic resonance imaging (fMRI) studies, from patients with strokes causing lesions in different areas, and from animal studies that it is not until a region is just beyond the primary auditory cortex that the frequency components are integrated into a single percept (representation) of a sound with pitch at the fundamental frequency.

The auditory cortex is quite immature at birth. Not until a few months after birth do mature, fast-myelinated connections begin to develop between neurons in deeper cortical layers in the auditory cortex, and these connections are not mature until about 5 years of age, which likely explains why the perception of basic sound features takes several years to reach adult levels. Mature connections are not complete in more superficial cortical layers of auditory cortex until about 12 years of age. Connections from the auditory cortex to other cortical regions, such as the prefrontal cortex, are largely from superficial layers of auditory cortex, which likely explains why complex aspects of music processing related to memory, attention, and executive function continue to develop into the teenage years.

At birth infants are reasonably good at discriminating small frequency differences, although their discrimination thresholds continue to improve until they reach adult levels at around 10 years of age. There are two basic types of cues to frequency: temporal cues (populations of auditory nerve fibers fire at rates reflecting the frequency composition of the stimulus) and spectral or place cues (tonotopic map formation). The spectral or place mechanism matures earlier than the temporal mechanism. The improvement in frequency discrimination with age is likely due in large part to the delayed maturation of temporal compared to spectral cues.

The ability to integrate harmonics into a pitch percept takes some time to develop. Pitch perception (as opposed to simple frequency perception) can be studied by examining the ability to perceive the pitch of a complex tone in which energy at the fundamental frequency has been removed. Adults perceive the pitch of such a tone to be at the frequency of the missing fundamental. However, it is not until infants are around 3 or 4 months of age that the auditory cortex represents the pitch of the missing fundamental. Because infants cannot verbally describe what they perceive, one method for examining how sounds are encoded in the auditory cortex is to measure event-related potentials (ERPs) derived from electroencephalographic (EEG) recordings of neuronal activity. ERPs represent the brain's response to the presentation of a sound stimulus, and consist of a number of components at varying times from the onset of the sound stimulus, reflecting stages of sound processing. One ERP component, the mismatch negativity, reflects the brain's response to rare and unexpected changes in an ongoing train of stimuli. Mismatch responses can be used to examine whether infants' auditory cortex can represent different sound features. Younger infants show mismatch responses to frequency changes, but mismatch responses to the pitch of stimuli with missing fundamentals is not seen until 3 or 4 months of age.

...