W. Ross Ashby

29 Nov 1999 17:38

British psychiatrist, one of the brighter lights of the early days of cybernetics, who is extremely little known for all the quality of his work and all the eminent people he influenced --- Herbert Simon, Norbert Wiener, Miller, Galanter & Pribram, Stuart Kauffman (see below), and so on. In fact, I can't find any references to him in any of the standard biographical reference works on science, math, technology, computers, etc. (to say nothing of Britannica, which continues its steady decline from the heights of the 11th edition), and only passing mention of him in histories like Heims's The Cybernetics Group. This is bizarre and unjust, and I'd greatly appreciate it if anyone who could tell me anything about him would write me.

And adaptation. And intelligence, esp. intelligence augmentation. The "law of requisite variety." The idea that any deterministic system will eventually evolve its "own sort of life and intelligence, possibly in the zero degree." Where did his (unfortunate) infatuation with Bourbaki come from?

Design for a Brain: The Origin of Adaptive Behavior. [I really like this book. It is, I think, wrong in places (see below), but it's clear, persuasive, even elegant in a way, and turns up in the oddest places. It is, for instance, one of the inspirations of Stuart Kauffman's The Origins of Order (though Design is much more fun to read). In one sense it's very advanced, since Ashby began incubating the ideas in the '40s, published the first version as a paper in '48, and the second edition of the book in '60: yet it's still very informative. On the other hand, his references to dynamical systems are thirty or forty years out of date, and at times I have to stop and consciously translate what he's saying into the modern language.
It remained in print through the early 1990s at least (London: Chapman and Hall, ISBN 0-412-20090-2), and perhaps is still; it is now (2013) available through archive.org. I may perhaps quote the "Summary" on p. 238:
The primary fact is that all isolated state-determined dynamic systems are selective: from whatever state they have initially, they go towards states of equilibrium. These states of equilibrium are always characterised, in their relation to the change-inducing laws of the system, by being exceptionally resistant.
(Specially resistant are those forms whose occurrence leads, by whatever method, to the occurrence of further replicates of the same form --- the so-called `reproducing' forms.)
If the systems permits the formation of local equilibria, these will take the form of dynamic subsystems, exceptionally resistant to the disruptive effects of events ocuring locally.
When such a stable dynamic subsystem is examined internally, it will be found to have parts that are co-ordinated in their defence against disturbance.
If the class of disturbance changes from generation to generation but is constant within each generation, even more resistant are those forms that are born with a mechanism such that the environment will make it act in a regulatory way against the particular evnironment --- the "learning" organisms.
This book has been largely concerned with the last stage of the process. It has shown, by consideration of specially clear and simple cases, how the gene-pattern can provide a mechanism (with both basic and ancillary parts) that, when acted on by any given environment, will inevitably tend to adapt to that particular environment.
Clearly, there are some parts of this which are over-statements (for instance, chaotic Hamiltonian systems can wander over essentially all of their phase space indifferently, i.e. they do not go to equilibria). I think it's equally clear that Ashby really was on to something, and one of my longer-term projects is to re-do the essential portions of Design for a Brain in modern terms (replacing "equilibrium" with "attractor", for instance), showing where Ashby's results break down and how they must be modified. Ideally, I'd preface this by a sketch of his life and of his influence.]
Introduction to Cybernetics [Now available on-line in PDF!]
"Principles of Self-Organizing Systems" in Heinz von Foerster and George W. Zopf, Jr., eds., Principles of Self-Organization, 1962. [Remarkably enough, for such a paper, it claims that there's really no such thing as self-organization. The argument runs as follows. By the "organization" of a system, Ashby means the rule which takes present states into future states. A self-organizing system, at the very minimum, must change its organization. One could try to represent this by making the evolution-rule depend on the current state, but obviously this just means you have a different, unchanging rule than you first thought. You could make the rule depend on some external input: then the organization would change with the input; but then it isn't really self-organizing; and if you include the input-device in the system, you're back where you started.
I don't think this is what most people have in mind when the speak about "self-organization"; rather I think it's more like what Ashby considers under the heading of "selection of states" by the organization of the system. To make some slight move towards formalizing it, a system would be self-organizing if it takes a flat, even distribution of states into a peaked, non-uniform one. If p is the distribution over states, then I think it would be enough to say that (sum) p log p was decreasing.
Cf. self-organization in these notebooks, and my talk "Is the Primordial Soup Done Yet?"]
"Principles of the Self-Organizing Dynamic System", Journal of General Psychology (1947) 37: 125--128 [First known occurance of "self-organizing" in print. Uses the same notion of organization as the above, and shows how it can apparently change if some of the variables are step-functions of the others.]
"Design for an Intelligence-Amplifier", pp. 215--234 in Claude Shannon and John McCarthy (eds.), Automata Studies (Princeton University Press, 1956)
The W. Ross Ashby Digital Archive [Complete archive of his research journal, plus bibliography, biographical information, etc.]

WRA, Mechanisms of Intelligence: Ross Ashby's Writings on Cybernetics [ed. Roger Conant]
Roberto Cordeschi, The Discovery of the Artificial: Behavior, Mind and Machines Before and Beyond Cybernetics
George J. Klir, "W. Ross Ashby" [Originally Klir's introduction to Mechanisms of Intelligence, now online]
Andrew Pickering
- "Cybernetics and the Mangle: Ashby, Beer and Pask" [PDF]
- The Cybernetic Brain