Stahl67 bmp

168 WALTER R. STAHL

entire field of cybernetics. Westcott (1960) emphasizes the problems of statistical fitting of “model parameters.” Clearly, there must be a firm conceptual basis for choosing what sort of parameters are to be fitted. Rashevsky (1964, 1965) notes that the large majority of reports pub-lished in the field of “mathematical biology” really deal with model Systems of one type or another. He describes a wide variety of mathematical simulation approaches, but fails to suggest quantitative criteria of similarity suitable for the objective comparison of models and prototypes.

Various types of neurophysiological and cybernetics models are con-sidered in a symposium edited by Rosenblith (1962). He stresses partic-ularly the importance of application of automata, communication, and control theory to biology, and the value of Computer simulation. Rosenblith comments: “Though there exist as yet no mathematical principles that seem valid for the whole of biology we can now test much morę effectively the generality and usefulness of certain concepts that have posed their candidacy at various levels of biophysical complexity. Computers help us in the analysis of our data and in the formulation of models.”

A number of symposia volumes on problems of biological cybernetics, artificial intelligence, and bionics are discussed below. It is elear that these disciplines are based on a deliberate design of biological artifaets. McCulloch (1962) and Muses (1962) consider some very generał ques-tions of imitating life and humans in particular. There has been a great deal of interest in cybernetic modeling in the Soviet Union. For example, Novik (1964) finds this approach to be so important as to constitute a major conceptual basis for Soviet biological phiłosophy. Guliaev (1965) finds bionics to be of great importance for futurę technology, a view-point widely held in the United States sińce approximately 1955.

The relationship of mathematics and models, as such, is considered in the section II and has not been madę entirely elear in existing works on biological simulation. A symposium on application of mathematics in biology edited by Bellman (1962) makes almost no reference to modeling theory and gives examples of the application to biology of very dissimilar mathematical approaches, including topology, differential equations, information theory, and automata theory. A volume by Kemeny and Snęli (1962) deals with mathematical models in social Sciences and offers examples based mainly on probability and operations research methods. Abstract and axiomatic models are the subject of still another volume edited by Gregg and Harris (1964) that deals with the