LANGUAGE IN A SOCIOBIOLOGICAL FRAME *

In an earlier paper published by The New York Academy of Sciences 1 addressed the contributions that comparative neurobehavioral research had made to delineating the structure of language.' That report distinguished two quite separate aspects of animal communication, each of which has a t its disposal a different neural mechanism. One form, the symbolic, relates the communicative act to the internal state of the animal and is in a sense arbitrary and synthetic with respect to the situation in which it occurs. The other constructs signs by differentiating the features distinctive to the situation, thus assigning attributes and characterizing it. Further, the earlier report developed the thesis originally proposed by Vowles*-3 and amplified by Beer (Ref. 4 and this volume) that three levels of neural mechanisms could be identified in the production of both sign and symbol. At the deepest level, which is shared by vertebrates and invertebrates alike, presymbolic processes are based on recurring regularities such as those that make up circadian and ultradian rhythms. Rudimentary significations, by contrast, are a t this level based on stochastic finite-state Markov-type mechanisms. In vertebrates, presymbolic and presignificant communications are constructed from these primitive operations by phrase structure hierarchies, which in man are subject to still another type of operation leading to the transformational capacities necessary to the construction of linguistic signs and s ym bob.