Valentino Braitenberg – författare

There is a tradition of theoretical brain science which started in the forties (Wiener, McCulloch, Turing, Craik, Hebb). This was continued by a small number of people without interruption up to the present. It has definitely provided main guiding lines for brain science, the devel­ opment of which has been spectacular in the last decades. However, within the bulk of experimental neuroscience, the theoreticians some­ times had a difficult stand, since it was felt that the times were not ripe yet and the methods not yet available for a development of a true theoretical speciality in this field. Thus theory remained in the hands of a fairly small club which recruited its members from theoretical physicists, mathematicians and some experimentalists with amateurish theoretical leanings. The boom of approaches which go by the name of ''computational neuroscience'', ''neuronal networks'', ''associative mem­ ory'', ''spinglass theory'', ''parallel processing'' etc. should not blind one for the fact that the group of people professionally interested in real­ istic models of brain function up to the present date remains rather small and suffers from a lack of professional organization. It was against this background that we decided to organize a meet­ ing on Theoretical Brain Science. The meeting was held April 18 - 20, 1990 and took place at Schloss Ringberg, West-Germany, a facility sponsored by the Max-Planck-Society.

There is a tradition of theoretical brain science which started in the forties (Wiener, McCulloch, Turing, Craik, Hebb). This was continued by a small number of people without interruption up to the present. It has definitely provided main guiding lines for brain science, the devel­ opment of which has been spectacular in the last decades. However, within the bulk of experimental neuroscience, the theoreticians some­ times had a difficult stand, since it was felt that the times were not ripe yet and the methods not yet available for a development of a true theoretical speciality in this field. Thus theory remained in the hands of a fairly small club which recruited its members from theoretical physicists, mathematicians and some experimentalists with amateurish theoretical leanings. The boom of approaches which go by the name of 'computational neuroscience', 'neuronal networks', 'associative mem­ ory', 'spinglass theory', 'parallel processing' etc. should not blind one for the fact that the group of people professionally interested in real­ istic models of brain function up to the present date remains rather small and suffers from a lack of professional organization. It was against this background that we decided to organize a meet­ ing on Theoretical Brain Science. The meeting was held April 18 - 20, 1990 and took place at Schloss Ringberg, West-Germany, a facility sponsored by the Max-Planck-Society.

In this essay we propose a view of the cerebral cortex from an unusual angle, determined by our search for the most general statements on the relation between cortical structure and function. Our results may be at variance with the ideas of those physiologists who have specialized subsystems of the cortex in mind, but are not in contrast with them, as we hope to show. Our view is akin to one presented by Moshe Abeles in his monograph Local Cortical Circuits ( 1982). It is related to the theory developed by Gunther Palm, for many years our partner, in Neural Assemblies (1982). It owes much to the ideas which inspire the work of George Gerstein (beginning with the classic: Gerstein 1962) and especially to the insights gained by our colleague Ad Aertsen and his group in TUbingen. Our ideas on the cortex find some resonance in a recent trend in Artificial Intelligence (Kohonen 1977; Hopfield 1982; Rumelhart et al. 1986), and indeed the diagram (Braitenberg 1974a) representing the dominant idea of that trend, the associative matrix with feedback, has served as an emblem for the editorial enterprise Studies of Brain Function since its inception in 1977. Since our earlier reports (Braitenberg 1974a,b, 1977, 1978a,b, 1986; Schilz 1976, 1978, 198la,b, 1986) enough new experimental evidence has accumulated in our own laboratory (summarized by Schilz 1989) and elsewhere to justify a more extended and more confident presen­ tation.

By means of quantitative analysis of the tissue components in the cortex of the mouse, this book presents an overall picture of the cortical network which is then related to various theories on cortical function. Centering around the idea of a diffuse network in a fairly homogeneous population of excitatory neurons, that of the pyramidal cells, it shows that the whole organisation in the cortical skeleton of pryramidal cells corresponds well with the idea of an associative memory and with the theory of cell assemblies. Provides the reader with information on quantitative neuroanatomy and also on the methods used, in particular those that vary from the norm.