auditory plasticity

The paper I’m currently trying to read is depressingly incomprehensible to me, so I’m gonna cheat and backtrack a few weeks to something with fewer undefined mathematical terms. This [pdf] is a review paper about learning-induced plasticity in the auditory cortex.

Ok, so! The auditory cortex is organized into tonotopic maps. This means that each neuron will fire most rapidly in response to a particular acoustic frequency, and that neurons are arrayed in a spectrum that corresponds to the spectrum of audible frequencies (and to the corresponding regions of the cochlea), so neurons that respond most strongly to similar frequencies are located adjacently. This is similar to the visual and somatosensory cortices, where neural maps reflect the spatial layout of the structures that provide sensory input to them (eyes, body). You can observe this directly in animals by inserting electrodes into neurons (when the animal is anesthetized), playing tones, and seeing how rapidly they fire.

Normally the tuning of these maps is very stable over time — you can test the same organism over and over again and the neurons’ optimal frequencies won’t change. However, when particular tones are made behaviorally relevant through conditioning (i.e. when they signal an impending tail shock, or sweet, sweet electrical stimulation of the dopaminergic reward circuits in the ventral tegmental area) the tonotopic maps reorganize themselves and more neurons become optimally responsive to the tone. The effect can be induced with a small number of trials and remains stable over time periods as long as two weeks; in fact, it becomes stronger after a couple nights’ sleep (which makes sense since there’s a lot of other evidence that processing that takes place during sleep aids in the consolidation of memories).

My take is that this probably points towards a developmental account of why people are really good at making fine auditory distinctions at the frequencies used in human speech, or why bats are really good at hearing their own sonar tones etc. These sound sensitivities are basically ubiquitous and consistent among the species that display them, but they are probably not “hard-wired” – the only thing that’s innate is having neural networks that are sensitive to operant conditioning, and environment does the rest. Pretty slick.

(The paper discusses possible mechanisms for this conditioning, but sadly my knowledge of the 15ish candidate brain regions and their connectivity is too half-assed to say anything substantive about them. They have a nice flowchart in Figure 5, though!)