The Cerebellum as a Neuronal Machine

John C. Eccles

Masao Ito

János Szentágothai

Springer, New York, 1967


Can we reverse engineer the cerebellum?

Imagine we know the complete wiring diagram of a brain structure. Let’s also say we know the sign, strength and time course of synaptic signals at every connection between the cells, the signals entering the circuit from the outside, and the way every cell’s spiking changes in response to each synaptic input.  Shouldn’t we be able to predict exactly what the circuit would do in response to any pattern of input?  And if we could do that, wouldn't we understand that structure's function within the brain?

In electronic circuit design, this process is called reverse engineering.  If we have an electronic circuit but no documentation, we still might be able to reconstruct the conceptual design – that is, the functional idea in the designer’s mind that drove the choice of components, connections and layout. This is difficult but usually possible, even for complex electronic circuits. But should we expect this to work for brain circuits? Designers of electronic circuits have a shared body of theory and a set of standard components and practices. Much of what one designer does is recognizable to another. In reverse engineering the brain, this is lacking. We aim to discover the ideas that guided the design of our own brain, to know the mind of its creator.  That is what makes brain research so exciting and mysterious. But it is an ambitious plan, and it may not succeed.

Some neuroscientists have accepted the challenge and tried to analyze brain circuits as if they were bioelectronic machines.  Many of them were inspired to do so by this classic book on the cerebellum by Eccles, Ito and Szentágothai. The book came at an optimistic moment for neuroscientists.  The preceding decade had produced amazing advances in brain science.  Microscopy, especially electron microscopy, had revealed details of the synaptic contacts between neurons for the first time. Neuronal circuits in each part of the brain were found to be composed of many copies of a small number of unique cell types each with a specific pattern of interconnections. Neurophysiologists had measured the electrical signals generated at the synaptic connections between cells, the chemical nature of synaptic transmission had been established, and the biophysical mechanism of the action potential had been revealed.  

How should we use all these advances in cellular neurophysiology to understand brain regions? The brain is structurally complicated, with many different parts, each of which seem to have a unique cellular architecture.  Maybe it would be helpful if we knew the function of the brain region ahead of time. Many neuroscientists have chosen to work on the easy parts – sensory receptors or motoneurons, whose functions are at least partly apparent.  But most of the nervous system resides in the vast gap between sensory receptor and motor output. Efforts to map function into the brain by studying behavior while stimulating or inactivating particular brain structures have produced ambiguous results. Maybe, the functions of the various parts of the brain could be understood in physiological, rather than psychological, terms. That understanding would come from knowing the operations they perform on their inputs or the signals they deliver to neurons in other brain parts.  The internal structure of the brain might be understandable as a set of logical functions or input-output operations.  By 1967, progress in computer design had been made using a modular approach.  Computer circuits were built from simple circuit modules that each performed an elementary logical function. More complex computational tasks of all types were performed by composition using these simple circuit modules. Maybe the brain might be built in a similar manner, from small subcircuits that are repeated many times.

A Neuroscience Supergroup:

In 1967, when the book was published, its authors were among the most productive and successful neuroscientists in the world.  Sir John Eccles shared the Nobel Prize in Physiology and Medicine with Hodgkin and Huxley in 1963.  Eccles had made fundamental discoveries on the nature of synaptic transmission, and especially on synaptic inhibition, while working on the neural circuits underlying spinal cord sensori-motor reflexes. Eccles had shifted his laboratory focus to the cerebellum soon afterwards. Masao Ito was the youngest member of the group, having recently set up his own laboratory at University of Tokyo after finishing three years of research in Eccles group at the Australian National University. Like Eccles, he began work on the cerebellum only a few years before the book was written, and his part of the book was about work done after his move to Tokyo.   János Szentagothai was a prominent Hungarian neuroanatomist, and chair of the Department of Anatomy at the University of Budapest.  At that time, Hungary was dominated by the Soviet Union, and scientific exchanges with western countries were limited.  Despite this impediment, Szenthagothai's work had won him international acclaim and his studies of the structure of brain circuitry were known and respected worldwide.  Before working on the cerebellum, he had made important discoveries on the structural substrates of spinal reflexes, the vestibulo-ocular system, and the hypothalamic portal system, by which the brain controls pituitary hormone secretion. The list of contributors to the book, all students or associates of the authors not listed as authors, is also impressive.  All of them would become familiar names to neuroscientists over the next 30 years – Drs. Andersen, Oscarsson, Voorheve, Llinás, Sasaki, Strata, Hámori, Tsukahara, Toyoma, Yoshida, Obata, and Uto.

According to the Preface, the book was conceived at a meeting of the International Physiological Congress in 1965. The theme of the book, according to the authors, was to “...discover the functional meaning of the patterns of neuronal connexion in the cerebellum...”

What is a neuronal machine?  

Its title is one of the book’s main attractions. Neuroscientists have always compared the brain to the most advanced technology of their own time. Of course the most exciting new technology of the 1960s ws the digital computer, and that is the kind of machine the authors were thinking about. But computers in 1963 were not the same as they are now, and "neuronal machine" probably means something different to us now than it did then. What made Eccles, Ito and Szentágothai single out the cerebellum as a candidate “neuronal machine”?  They described several machine-like features of the cerebellum.