Hypotheses relating to the function of the claustrum II: does the claustrum use frequency codes?

NMDA receptor-dependent mEC gamma rhythms are mediated by basket interneurons, but NMDA receptor-independent gamma rhythms are mediated by a novel interneuron subtype-the goblet cell (Middleton et al., 2008). [...]the hypothesis could be put forward that if a neuron (or group of neurons) receives two inputs, one oscillating at x Hz and the other at y Hz, and generates an oscillation in its output at a new frequency z, then z, whatever it is, could carry information that x has been bound to y. Translated into features of the stimuli that led to the production of x and y, this entails that z now carries information that x and y are “bound.” In any equation (a + b = c) there are many values of a and b that add to yield c. Therefore, one neuron in a feed forward chain, that receives an input oscillating at c Hz from another neuron, has no way of “telling” whether this represents a primary oscillation at c Hz, or which among the many values for a and b combinations that yield c. Thus, as an information processing system, it is useless. [...]in the case of any mechanism that “blends” the two incoming wave into a complex exit wave, the latter would have to be “interpreted” by a complicated decoding mechanism in the receiving neuron. [...]as we cannot say at this instant that we know for certain that such an operation does not take place, we suggest that, although it may be too early to abandon subhypotheses 2 and 3 entirely, we should at least put them on the back burner. [...]even if this were not so, concatenation and nesting may not be good models for the type of integration we are proposing.