Convergence
As discussed in an earlier post, there was an article written recently that (explicitly) challenges Stephen Jay Gould’s view that the evolutionary outcome of intelligent life is highly contingent: Simon Conway Morris, “Is anybody out there?”, New Scientist, 16 November 2002. Here are my thoughts on the article, which I am writing prior to reading the debate between Conway Morris and Gould that is archived here. I’ll offer some thoughts on these essays after I read them.
Firstly, there is one sense in which we were trivially and uncontroversially determined: the very fact that we are here. But this is not, perhaps, as trivial as it might seem, and not simply because of the anthropocentric bias it immediately gives to all our theorising. Namely, it requires that we provide a proper interpretation of such talk as “if the evolutionary tape were played again, we would not be here”. But by definition, the evolutionary tape produces us, so what variations can be introduced? Gould must be read, I assume, as implying that life is the result of many chance events. Without delving into the philosophical problems of chance and necessity, such chance events by definition might have resulted in different outcomes. And so Gould must be read as saying that these different outcomes would not (necessarily) have produced us. But again, this is trivial if interpreted literally, because of course if our history was different we would not be the same, since our identity presumably includes or depends on our history. The proper interpretation must be that the different outcomes would not (necessarily) have produced creatures relevantly similar to us. And here is the crux of the debate: what here counts as relevantly similar? Presumably, we are talking about advanced intelligence, and this is the point on which Conway Morris focusses.
Conway Morris offers a number of examples that show that different species have solved similar problems with similar solutions, including the intriguing fact that camera-like eyes like our own (as opposed to the compound, multi-lense eyes of insects) have evolved independently seven times. These examples, he proposes, show that any type of intelligence that evolves must be similar to our own. The argument seems pretty compelling, but what I am unsure about is to what extent it challenges Gould’s argument. It seems (at least, from Full House; I haven’t read Wonderful Life) that Gould is emphasising that we should not see humans or human intelligence as being the unique and predestined outcome of evolution; rather, we are the particular contingent outcome that has been produced this time around. The issue of whether evolution necessitates intelligence of some sort is a partially separate issue, and it is this with which Conway Morris wants to take issue. This complicates things a little, because prima facie we can easily imagine parallel universes that do not produce human-level intelligences at all (ones in which meteor impacts are too frequent to provide the stable environmental conditions required, for example), and parallel universes that produce human-level intelligences implemented in wildly different forms (Conway Morris offers one such scenario, involving an Oxygen-rich environment that allows insect brains to grow orders of magnitude larger and develop human-level social intelligence). I don’t think Gould or Conway Morris take issue with the possibility or even probability of these alternative scenarios, and so the debate seems to be less a debate than a difference in emphasis: Conway Morris emphasising that if there was to be intelligence, it had to be like us; Gould emphasising that there did not have to be intelligence, but if there was, it would not necessarily be identical to us. I have read elsewhere a criticism of Gould that he often wants to proclaim shifts in emphasis as revolutions, and that he often has replied that the emphases are very important. My own view is that it is important to keep the Gouldian emphasis, even if we accept an argument for convergent solutions to evolutionary problems. That is, we as humans are not the unique and only possible outcome to evolution, and we should certainly not see ourselves as the raison d’etre of the universe.
(As an aside, intelligence seems to be an inherently anthropomorphic term, and so difficult to define except with reference to ourselves. But we should retain the possibility of there being intelligent life that is completely unintelligible to ourselves. The fish never dreamed it would be a human, as the proverb goes).
In closing, there are two massive editorial oversights in the Conway Morris article, as I see it, which I am extremely surprised to note given the overall quality of New Scientist. Firstly, he argues as follows (this is a reconstruction, rather than a step by step summary):
1. We can conceptualise different possibilities for forms of life as a kind of hyperspace of combinations.
2. If convergence is true, then only a minuscule set of these spaces can be filled.
3. Separate spaces converge to the same types of solutions.
4. We can conceptualise the connections between such similar solutions as tunnels between the spaces.
5. The conclusion: “And if biological hyperspace has its equivalent of “wormholes” then perhaps understanding how these work could one day help us to navigate multidimensional space-time”.
But this is patently absurd! The hyperspace nature of evolution is a conceptual apparatus we have imposed; and the notion of a tunnel cannot be supported even by the conceptual apparatus, since in evolution the different solutions are reached completely independently (wormholes are supposed to provide direct connections). To suppose this might help us to understand wormholes in space-time is completely ludicrous, the kind of speculation that shouldn’t even pass the grade of a sci-fi novel editor.
Secondly, Conway Morris blatantly presumes that advanced insect societies would be more disposed to discover the principles of quantum computation than humans; and that this would enable them to more easily develop intergalactic space travel. But the first presumption is purely speculative and the second downright ridiculous—I’ve heard many outlandish suggestions for the application of quantum computers, but their ability to help propel a human through space-time (except via better number-crunching of the physical theory involved) is not one of them.
On to the Gould exchange with Conway-Morris; more to follow.