WADD TR 60-600

GENERAL SESSION ||

FORUM DISCUSSION

Chairman: Mr. H. V. Noble

Wright Air Development Division

WHAT GOOD IS BIONICS?

Mortimer Taube
Documentation, Inc.

The question "What good is Bionics?" is like the question once addressed to a great British historian. He was asked "What good is history?" He responded, "What good is a baby?" This rhetorical response is a way of saying that it is the nature of man to study history just as it is the nature of man to make babies and that questions of the utility of either procedure are not relevant and nobody is going to stop either procedure because of an inability to answer somebody's fool question.

If it

This is not to imply that "What good is Bionics?" is a fool question. elicits the recognition that it is the nature of man to be concerned with his own anatomy and functions and to build gadgets which assist him to function better, then it is a worthwhile question. But if our question is made more specific, then it deserves and can elicit more specific answers, for example:

Is the study of history good for predicting future events?

Is the study of history good for making young people patriotic?

These questions can be answered by correlating various sets of observations.

Similarly, if we ask:

Is Bionics good for improving the defense posture of the United States? we ought to be able to present a reasonable answer.

Bionics has been defined for us as the application of the knowledge of living organisms to the solution of engineering problems. Even if positive pay-off in this direction comes slowly and hard, I would nevertheless feel that Bionics is important because its inception should mark the end of an era of wonderful nonsense.

The initial successes and great flexibility of general purpose computers led many people to the uncritical conclusion that in the operations of a computer lay the clue to the behavior and operations of biological organisms. This type of "reductionism" is endemic in the history of philosophy and science and arises concurrently with every major scientific breakthrough. In different periods of time those people who have been impressed with the latest scientific advances announce that "Now at last the riddle of the universe has been solved. We know that man is nothing but a collection of atoms," or -"We know that man "We know that man is nothing but a protein molecule," or is nothing but a digital computer." Actually, we know no such thing. Such a

view, if it is based upon anything at all, is based upon discarded Nineteenth Century metaphysics which now passes for "common sense." If anyone is interested in going beyond the specific differences between biological organisms and machines to seek a common explanation of natural phenomena, he is not going to find such an explanation in mechanical or materialistic terms but in categories which are essentially biological or organic. My authorities for this conclusion are Russell, Whitehead, Peirse, and Von Neumann, from whom I have taken these two statements:

and

"It may be that logic will have to undergo a pseudo-
morphosis to neurology to a much greater extent than
the reverse."

"It is dangerous to identify the real physical (or
biological) world with the models which are con-
structed to explain it. The problem of understand-
ing animal nervous action is far deeper than the
problem of understanding the mechanisms of a comput-
ing machine.

The issue between the Bionics approach and the "nothing-but" approach can be made sharper by referring for a moment to the question of models introduced by this statement of Von Neumann's.

In a moment of despair I walked out of the conference
on Self-Organizing Systems held a year or so ago in
Chicago and met Dr. Weyl of the Office of Naval Re-
search on the steps of the building and voiced my
inability to understand something I had just heard.
Someone had drawn a set of black boxes on the black-
board and then added: "This model may not resemble
nor explain at all the biological system it is model-
ing."
I asked Dr. Weyl what a model could be which
didn't model what it was designed to model.

Dr. Weyl soothed my fevered brow. He told me that
most of the time the models talked about were models

in the sense of this year's model of a car or a dress.

If the Bionics approach is to be taken seriously, one cannot make a model of a biological system in any significant sense if one starts from the computer or engineering end of the spectrum (or even from Venn diagrams.)

Rather one must stert as the IBM people did, as described in their paper, by studying biologic activity and trying to derive formal properties of that activity which can be modeled, i.e., described in either a set of equations or as an artificial construct.

One of the dangers of interdisciplinary conferences and the systems approach is that we all become experts in a lot of special disciplines we never studied and know very little about. I would like nothing better than