Organization and Perturbation of NeuronalAssemblies

The concept of neuronal assembly as the basic building block of brain activity dates back to Sherrington (Creed, 1932) and (1949). Most of our current knowledge about the nervous system is however based on observations of single neurons, one at a time, thus making possible only tenuous inference of what could be termed “pseudo-population” properties (Georgopoulos, 1988, 1989; Lee et al., 1988; Sparks et al., 1990 among many others). Full study of neuronal assemblies and their internal and external interactions and representations requires simultaneous observation of the activity of as many neurons as possible. Technical advances over the last decade are making this increasingly possible, although never easy (reviews in Kruger, 1983; Gerstein et al., 1983, 1989; and more recently McNaughton et al., 1983; Wilson et al., 1993; Nicoleiis et al., 1997). In addition to the several multiple electrode technologies, there has been a continuous stream of developments that allow the separation of different waveforms observed on each electrode (Abeles and Goldstein, 1977; Zardoshti and Wheeler, 1992; Salganicoff et al., 1988; and many others). Each such waveform presumably is the sign of activity of a distinct individual neuron; some of the sorting methods can accommodate the changes of waveform occurring in a rapid burst of firing, or deal with the complex biological noise background of a typical recording (Wilson et al., 1993, Fee et al., 1996).