The endless debate about extraterrestrial life has focused upon the calculation of probabilities—how many stars, how many suitable planets, the chance that life will originate on appropriate earths, the probability that life will eventually generate intelligence. I must confess that I have always viewed this literature as dreary and inconclusive, too mixed up with hope and uncertainty to reach any respectable conclusion.
Recently, several astronomers and astrophysicists have advocated a different approach—a direct search for the technological byproducts of intelligence by scanning the skies systematically with radiotelescopes, probing for signals emitted by other civilizations. This so-called SETI program (search for extraterrestrial intelligence) has been vigorously debated. Proponents claim that it would require but a minute fraction of the annual NASA budget and, whatever its chances of success, would at least move the subject from fruitless debate about probabilities to an experimental probe by the only means now available. Opponents counter that the scheme is a boondoggle, still costing millions and so virtually assured of failure that it merits not a penny of sparse public funds for science.
As an evolutionary biologist, I have no expert knowledge in most areas motivating this debate. I am moved to comment only because opponents of SETI have featured an argument from my field as one of their most powerful weapons. They state that all leading evolutionary biologists have proclaimed the existence of extraterrestrial life as nearly inconceivable. The optimism of some physical scientists therefore resides in their failure to understand the distinctive character of evolutionary reasoning. But opponents of SETI have misstated the biological argument, and I would like to explain why at least one evolutionary biologist thinks that SETI is a long shot well worth trying.
Frank J. Tipler, a mathematical physicist from Tulane University, has been the most indefatigable critic of SETI. In a long series of strongly worded articles for both technical and popular journals (New Scientist, Mercury, Physics Today, Quarterly Journal of the Royal Astronomical Society, for example), he gives “two basic reasons for my disbelief in the existence of extraterrestrial intelligent beings” (all quotes from his 1982 article in bibliography, though Tipler pursues the same themes in all his writings on SETI).
The second reason lies outside my field and I shall not dwell on it, though it must be mentioned. Tipler argues that “if ‘they’ existed, they would already be here…. Because they are not here, no such beings exist.” In short, Tipler claims that any truly intelligent creatures would search or colonize the cosmos with a device that he calls a von Neumann machine—“a computer with intelligence close to the human level, capable of self-replication and capable, indeed, of constructing anything for which it has plans, using the raw materials available in the solar system it is aimed at.” Intelligent life could therefore explore an entire galaxy “for the price of one von Neumann machine”—for this computer would mine asteroids and comets for material to build replicas of itself and its enclosing probe. These replicas would then scurry off to other suitable stars and replicate again. In a mere 300 million years, a whole galaxy could be saturated with the duplicated products of one von Neumann machine.
Such a machine could even fabricate the flesh and blood of extraterrestrials by mining the needed chemicals and then running the genetic program of its creator from stored memory:
This information could in principle be stored in the memory of a von Neumann machine, which could be instructed to synthesize an egg and place the “fertilized cell” in an artificial womb…. In nine months there would be a human baby in the stellar system, and this could be raised to adulthood by surrogate parents, constructed by the von Neumann machine.
I don’t mean to be a philistine, but I must confess that I simply don’t know how to react to such arguments. I have enough trouble predicting the plans and reactions of people closest to me. I am usually baffled by the thoughts and accomplishments of humans in different cultures. I’ll be damned if I can state with certainty what some extraterrestrial source of intelligence might do. Thus, Tipler’s second argument follows the speculative tradition that SETI, with its experimental approach, is designed to transcend.
As his first argument, however, Tipler features a different kind of claim based on the methods and data of my field. He writes:
First, all the great contemporary experts in the theory of evolution—Francisco Ayala, Theodosius Dobzhansky, Ernst Mayr, and George Simpson—are unanimous in claiming that the evolution of an intelligent species from simple one-celled organisms is so improbable that we are likely to be the only intelligent species ever to exist.
On the most mundane level, if I may play the irrelevant “expert game” for just one sentence, Tipler’s statement is empirically false. I count at least four quite respectable evolutionists in the international pro-SETI petition recently released by Carl Sagan (Tom Eisner of Cornell, Dave Raup of the University of Chicago, Ed Wilson of Harvard, and with apologies for arrogance, yours truly). Evolutionary biologists, in their usual consistency with nature’s primary theme, maintain a diversity of views on this subject.
More importantly, I think that Tipler has misunderstood what evolutionary biologists dismiss with such forcefulness by conflating two very different issues. All evolutionists who have discussed exobiology at length have clearly delineated two separate concerns—a specific claim and a general argument.
The specific issue considers detailed repeatability of any particular evolutionary sequence—in this case the evolution of creatures looking pretty much like us: bilaterally symmetrical with sense organs up front, two eyes, a nose in the middle, a mouth, and a brain. If we could start the earth’s tape anew, would intelligent creatures evolve again in this form? If other worlds share our basic chemistry and conditions, would such “humanoids” evolve on them?
The general question asks whether attributes that we would identify as intelligence might arise in creatures of any conformation—blobs, films, spheres of pulsating energy, or diffuse and unimagined forms far beyond the limited visions of most science fiction writers.
All evolutionists have vociferously denied the specific claim, and I join them in all their vigor. Many evolutionists have also gone a step further to doubt the general argument as well, but never with such certainty—and always as a personal opinion, not as a proclamation bearing the indelible imprimatur of “evolutionary theory.” I stand among those evolutionists who deny the specific claim but feel that no strong opinion can be entertained about the general argument. SETI only needs the general argument to bolster its case for support.
Gregory Bateson, the recently deceased guru of sciences that deal with complex objects and interacting systems, often emphasized that confusion of hierarchical categories may be the most common and serious fallacy of human reasoning (see his book Mind and Nature, for example). As a primary example of “category confusions,” Bateson identified the substitution of individuals for classes (or vice versa).
Casey Stengel, one of the greatest general gurus of our time, consciously committed Bateson’s fallacy of categories to avoid the heat of scrutiny in a tough moment. He was roundly criticized for blowing the Mets’s first pick in the expansion draft on a particular catcher of quite modest ability (one Hobie Landrith by name). Casey answered by invoking the class of catchers in general—“You have to have a catcher, because if you don’t, you’re likely to have a lot of passed balls.” Now Ol’ Case, as usual, knew exactly what he was saying (never let the patter known as “Stengelese” fool you). He used humor to blunt criticism because he knew that we would all recognize the fallacy of reasoning and laugh at the conflation. But we commit the same error in subtler circumstances and fail to identify our confusion.
When we use “evolutionary theory” to deny categorically the possibility of extraterrestrial intelligence, we commit the classic fallacy of substituting specifics (individual repeatability of humanoids) for classes (the probability that evolution elsewhere might produce a creature in the genera
l class of intelligent beings). I can present a good argument from “evolutionary theory” against the repetition of anything like a human body elsewhere; I cannot extend it to the general proposition that intelligence in some form might pervade the universe.
Physical scientists, following the stereotype of science as a predictable, deterministic enterprise, have often reasoned that if humans arose on earth, then we must infer (since cause leads inexorably to effect) that intelligent creatures of roughly human form would arise on any planet beginning with physical and chemical conditions similar to those that prevailed on the early earth. Perhaps this deterministic outlook is responsible for the paltry imagination of film makers and science fiction writers, with their endless creatures, all designed on a human model with two eyes, a nose, a mouth, two arms, and two legs (Close Encounters, ET, and even the more imaginative Star Wars). This tendency could be forgiven when human actors had to play the roles in our movies, but now that pieces of plastic can evoke our deepest emotions and move so subtly that ET becomes a national hero, this excuse no longer holds.
But styles of science are as diverse as their subject matter. Classical determinism and complete predictability may prevail for simple macroscopic objects subject to a few basic laws of motion (balls rolling down inclined planes in high school physics experiments), but complex historical objects do not lend themselves to such easy treatment. In the history of life, all results are products of long series of events, each so intricately dependent upon particular environments and previous histories that we cannot predict their future course with any certainty. The historical sciences try to explain unique situations—immensely complex historical accidents. Evolutionary biologists, as historical scientists, do not expect detailed repetition and cannot use the actual results of history to establish probabilities for recurrence (would a Caesar again die brutally in Rome if we could go back to Australopithecus in Africa and start anew?). Evolutionists view the origin of humans (or any particular butterfly, roach, or starfish) as a historical event of such complexity and improbability that we would never expect to see anything exactly like it again (or elsewhere)—hence our strong opposition to the specific argument about humanoids on other worlds. Consider just two of the many reasons for uniqueness of complex events in the history of life.
Mass extinction as a key influence upon the history of life on earth (see essays in section 8). Dinosaurs died some 65 million years ago in the great worldwide Cretaceous extinction that also snuffed out about half the species of shallow water marine invertebrates. They had ruled terrestrial environments for 100 million years and would probably reign today if they had survived the debacle. Mammals arose at about the same time and spent their first 100 million years as small creatures inhabiting the nooks and crannies of a dinosaur’s world. If the death of dinosaurs had not provided their great opportunity, mammals would still be small and insignificant creatures. We would not be here, and no consciously intelligent life would grace our earth. Evidence gathered since 1980 (see essay 29) indicates that the impact of an extraterrestrial body triggered this extinction. What could be more unpredictable and unexpected than comets or asteroids striking the earth literally out of the blue? Yet without such impact, our earth would lack consciously intelligent life. Many great extinctions (several larger than the Cretaceous event) have set basic patterns in the history of life, imparting an essential randomness to our evolutionary pageant.
Each species as a concatenation of improbabilities. Any animal species—human, squid, or coral—is the latest link of an evolutionary chain stretching through thousands of species back to the inception of life. If any of these species had become extinct or evolved in another direction, final results would be markedly different. Each chain of improbable events includes adaptations developed for a local environment and only fortuitously suited to support later changes. Our ancestors among fishes evolved a peculiar fin with a sturdy, central bony axis. Without a structure of this kind, landbound descendants could not have supported themselves in a nonbuoyant terrestrial environment. (Most lineages of fishes did not and could not evolve terrestrial descendants because they lacked fins of this form.) Yet these fins did not evolve in anticipation of future terrestrial needs. They developed as adaptations to a local environment in water, and were luckily suited to permit a new terrestrial direction later on. All evolutionary sequences include such a large set of sine quibus non, a fortuitous series of accidents with respect to future evolutionary success. Human brains and bodies did not evolve along a direct and inevitable ladder, but by a circuitous and tortuous route carved by adaptations evolved for different reasons, and fortunately suited to later needs.
The improbabilities of history proclaim that all species are unique and unrepeatable in detail. Evolutionary theory, as a science of history, does deny the specific argument for humanoids on other worlds. All leading evolutionists, in their writings on exobiology, have said so with gusto, and I agree. Wallace began the theme in 1903:
The ultimate development of man has, therefore roughly speaking, depended on something like a million distinct modifications, each of a special type and dependent on some precedent changes in the organic and inorganic environments, or in both. The chances against such an enormously long series of definite modifications having occurred twice over…are almost infinite.
Simpson has expressed the theme most eloquently in recent years, in his famous essay on “the nonprevalence of humanoids” (see bibliography):
This essential nonrepeatability of evolution on earth obviously has a decisive bearing on the chances that it has been repeated or closely paralleled on any other planet. The assumption, so freely made by astronomers, physicists, and some biochemists, that once life gets started anywhere, humanoids will eventually and inevitably appear is plainly false…. Let us grant the unsubstantiated claim of millions or billions of possible planetary abodes of life; the chances of such historical duplication are still vanishingly small.
But all these evolutionists have also clearly distinguished this specific proposition about humanoids from the general argument that intelligence in some other form might arise elsewhere. On the general proposition, they have maintained a diversity of opinions—leading to the empirical conclusion that “evolutionary theory” has no clear pronouncement to make. Both Wallace and Simpson extended their argument to doubt the general claim as well, but ever so much more gently, and as a personal opinion only. Simpson, for example, wrote:
Even in planetary histories different from ours might not some quite different and yet comparably intelligent beings…have evolved? Obviously these are questions that cannot be answered categorically. I can only express an opinion…I think it extremely unlikely that anything enough like us for real communication of thought exists anywhere in our accessible universe.
Other evolutionists, however, including two cited by Tipler as denying any possibility for SETTs success, also distinguish the specific from the general argument, but express far more optimism for the generality. Dobzhansky and Ayala, in a leading textbook (coauthored with G.L. Stebbins and J.W. Valentine), write (see bibliography):
Granting that the possibility of obtaining a man-like creature is vanishingly small even given an astronomical number of attempts…there is still some small possibility that another intelligent species has arisen, one that is capable of achieving a technological civilization.
I am not convinced that the possibility is so small.
Does evolutionary theory offer any insight about the general argument? We gain some sense of probabilities for repetition of a basic theme (but not of specific details) from the phenomenon known as “convergence.” Flight has evolved separately in insects, birds, pterosaurs (flying reptiles), and bats. Aerodynamic principles do not change, but morphologies differ widely (birds use feathers; bats and pterosaurs employ a membrane, but bats stretch it between several fingers, pterosaurs only from one). Marsupial “moles” and “wolves” evolved on Australia, a continent isolated from placental
mammals elsewhere. Since adaptive themes are limited and animals so diverse, convergence of different evolutionary lineages to the same general solution (but not to detailed repetition) are common. Highly adaptive forms that are easy to evolve arise again and again. More complex morphologies without such adaptive necessity offer little or no prospect for repetition. Conscious intelligence has evolved only once on earth, and presents no real prospect for reemergence should we choose to use our gift for destruction. But does intelligence lie within the class of phenomena too complex and historically conditioned for repetition? I do not think that its uniqueness on earth specifies such a conclusion. Perhaps, in another form on another world, intelligence would be as easy to evolve as flight on ours.
Tipler dismisses the issue of convergence by stating that biologist Leonard Ornstein (in an article supporting Tipler, see bibliography) has refuted the most famous of all convergences—the “camera eye” of vertebrates and cephalopods (squids and their allies)—by suggesting that this structure arose in both groups from a common ancestor, and not separately in each. Even if Ornstein were right, the dismissal of a specific case does not deny the importance of convergence as a general phenomenon. But Ornstein’s arguments are seriously flawed. He never mentions the strongest, “classical” argument for convergence—that the eyes, although so similar in design and operation, develop embryologically in fundamentally different ways (squid eyes form from skin precursors, while vertebrate eyes, the lens excepted, develop from the brain). Moreover, Ornstein’s main argument for evolution from a common ancestor relies upon a biological principle disproved more than fifty years ago. He invokes Haeckel’s discredited law that “ontogeny recapitulates phylogeny”—that an organism’s embryological development repeats the sequence of ancestral adults in its evolutionary lineage. Since the eye develops so early in embryology, Ornstein argues that it may have already existed in a very remote ancestor—early enough to predate the evolutionary split of vertebrate and squid lineages. Not only has Haeckel’s law been disproved (embryos do not repeat ancestral stages), but even Haeckel himself, in the heyday of his principle, rarely used time of appearance in embryology to specify moment of evolutionary origin—for he himself had identified and named a large class of exceptions to such a facile generality.