STUART KAUFFMAN: Francisco Varela is amazingly inventive, freewheeling, and creative. There's a lot of depth in what he and Humberto Maturana have said. Conversely, from the point of view of a tied-down molecular biologist, this is all airy-fairy, flaky stuff. Thus there's the mixed response. That part of me that's tough-minded and critical is questioning, but the other part of me has cottoned on to the recent stuff he's doing on self- representation in immune networks. I love it.

The work Francisco is doing on the core immune network, which is representing self, and the peripheral system, which is responding to an outside world, is very intriguing. I'm not sure whether he's correct in his thesis that the immune repertoire evolved as a means of representing self, and that an evolutionary consequence was the capacity to recognize and ward off nonself. Whether or not one agrees with that sort of ontological and evolutionary argument, the work he's doing is very nice. It's imaginative, it's tied down to facts in places where it can be tied down. He is very smart, utterly charming and graceful, and his capability in any one of a large number of languages astonishes me.

I first got to know Francisco, indirectly, in 1983, when I met Humberto Maturana in India. They'd come up with their theory of autopoiesis, which was considered gobbledygook by many tough- minded scientists if they paid any attention to it at all. After listening to Humberto, I returned to my work on autocatalytic sets, which I'd begun in 1971 and then set aside. I believe that my autocatalytic-polymer-set story is the clearest instance I know of, in terms of a formally described model, of what they mean by autopoiesis.

It's likely that 99 percent of serious biologists have never heard of Francisco. This is for two reasons. First, he's not American or English, and the bulk of serious molecular biology is done in America and England, with some being done in France, Switzerland, and Germany. Francisco, after all, comes from South America. He's not from the "right" part of the world — that is, the kind of place that usually produces biologists. Second, Francisco is a good theoretical biologist, and theory in biology is in low repute. He's done detailed simulations of immune networks and neural networks that actually function — at least on the computer — so it's good solid theoretical biology. It ties in with our work at the Santa Fe Institute on emergent collective phenomena.

I'm less florid than Francisco. Although his theoretical style may appeal to some of us theoreticians, it wouldn't appeal to tough-minded colleagues, or even to more facile experimental colleagues, who wouldn't see what the next experiment is.

This is a problem that's hard to get your mind around, if you aren't trained as a biologist. Unlike physics and chemistry, which are concept-driven and theory-driven, biology is essentially experiment-and grungy-fact-driven. Organisms are complicated, ad-hoc contraptions. That's been our view since Darwin.

Organisms are ad-hoc solutions to design problems. The standard view is that there are no deep theories of the deep meaning of ad-hoc contraptions. You take the things apart and find out how they work. Most biologists adhere to that view. Notions of underlying deep principles are not an anathema to them — they're just considered foolish.

Francisco is a philosopher, in a way. He and Humberto Maturana are right about their idea of autopoiesis. But he hasn't had a large impact in the United States. The main reason he's dismissed is that he's seen just as a philosopher. Along with Francisco, I'm among those who hold that such deep principles exist, and I'm trying to find them. I have a hard time being heard by my experimental colleagues. I would expect that Francisco has almost never been heard. In the pantheon of biological scientists, he's probably unknown.

W. DANIEL HILLIS: I used to think Francisco Varela was a mystic, because I couldn't understand his ideas. As I came to know him, I began to realize that he's actually fishing for some of the same things I am. He's trying to understand how emergent properties come from simple interactive systems. It's hard to express that question without sounding like a mystic. Cisco does not help things by genuinely being a mystic on some other issues, and hanging out with the Dalai Lama, but he's trying to get at the same issue I am. I think he's on to something, with his theories of the immune system; he's trying to look at network properties — things like attractors of the system, and so on — and trying to get above the level of looking at the chemistry of the immune system. It's yet to be seen whether that approach will actually explain anything, but I'm supportive of his quest.

Cisco clearly is a symbol for Marvin Minsky — a symbol for a set of things that Minsky is angry about. It's true that you lose perfectly good AI people when they go off into philosophy and stop doing anything useful. I think Minsky is very annoyed that one of his favorite students, Terry Winograd, started out by writing perfectly good computer programs and then went off and wrote a book on hermeneutics. That bugs Minsky, because he sees philosophy as a black hole into which his students are falling. In Marvin's mind, Cisco is a symbol of that black hole.

CHRISTOPHER G. LANGTON: Varela is one of those people who has such an engaging, articulate style of talking that when you sit and listen to him, you find yourself nodding your head and going, "Yes, yes, yes, this is all great." Then once you get out of the room, and out from under his very significant personal charm, it's hard to figure out exactly what it was he said. This is one of my problems with the field of autopoiesis. The contribution it makes is that it allows you to talk about a set of phenomena known to us from biology in a different kind of language, and sometimes just changing the language can make you look at things in a new way.

Some people who come across phenomena such as self- organization for the first time through the writings of Varela and Humberto Maturana become real advocates of autopoiesis, because it's in the context of that language that they first come across those phenomena. I came across those phenomena in the world of biology, and in the language of biology and physics, and so I'm used to thinking about them in that language, and I don't see any benefit for someone like myself in mapping them over onto the language of autopoiesis. I don't think it adds anything to our understanding of phenomenology. Once one has gone through the translation, there's no value added. It's just another way of describing the same phenomena — a way that's not particularly useful to me.

Varela would claim that he is adding something to the scientific discussion when he casts all these phenomena in his language, but whatever it is he adds always seems to slip away from me whenever I try to pin it down. I was troubled when a friend of mine pointed out that he could go through one of Varela's papers and replace the phrase "autopoietic system" with the phrase "living system" and it wouldn't change anything; in fact, several of the statements simply became tautologies. In other words, autopoiesis doesn't get me anywhere I haven't already been.

I know a lot of people, especially in Europe, who are very influenced by autopoiesis, and who are very careful in the way they describe this principle. However, I've also found that many of Varela's most ardent followers are flaming vitalists, who have found in autopoiesis a way to get beyond what they consider to be the reductionist agenda. They feel that autopoiesis allows for higher-level organizing principles in a way that what they call strict reductionist science cannot. That's epistemology, not science. The question is whether or not it's good epistemology. I don't know. Many people think it's very good, and I can't blame Varela or Maturana for the abuses wrought by their followers.

DANIEL C. DENNETT: Post hoc ergo propter hoc! "After this, therefore because of this." Francisco Varela is a very smart man who, out of a certain generosity of spirit, thinks he gets his ideas from Buddhism. I'd like him to delete the references to Buddhist epistemology in his writings. His scientific work is very important, and so are the conclusions we can draw from the work. Buddhist thinking has nothing to do with it, and bringing it in only clouds the real issues.

There are striking parallels between Francisco's "Emergent Mind" and my "Joycean Machines." Francisco and I have a lot in common. In fact, I spent three months at CREA, in Paris, with him in 1990, and during that time I wrote much of Consciousness Explained. Yet though Francisco and I are friends and colleagues, I'm in one sense his worst enemy, because he's a revolutionary and I'm a reformer. He has the standard problem of any revolutionary: the establishment is — must be — nonreformable. All its thinking has to be discarded, and everything has to start from scratch.

We're talking about the same issues, but I want to hold on to a great deal of what's gone before and Francisco wants to discard it. He strains at making the traditional ways of looking at things too wrong.