The Evolution of Intelligence

If for the moment we forget the details and simply use a "reasonable" working definition of human-level intelligence — that it involves a combination of factors including stereoscopic vision, symbolic language, tool use and so on — then one can ask the important question: how likely are other species to evolve a high level of intelligence?

Let us perform a thought experiment. Suppose 400 million years ago a meteorite had struck Earth, wiping out the ancestor of the vertebrate line but leaving untouched the ancestors of many other lines that are still alive today, such as the squid or the ant. Would any of those lines have given rise to intelligent species? Of course we cannot know for sure, because we live in a world in which vertebrates did not become extinct. But many evolutionary biologists think it improbable that human-level intelligence would have arisen from the squid-like ancestor or the ant-like ancestor. The reason is that evolution takes advantage of small, random mutations occurring in genetic DNA; if the change proves advantageous to an organism in the here and now, then the organism is competitively successful and the mutation propagates through the population. To repeat: evolution has no foresight. The mutation has to be of benefit now, not in the future, in order for the genes to spread. Now, there is no goal toward which evolution is working; much as we might like to think that high intelligence is the pinnacle of evolution, it simply is not. So, given this random process, the probability of producing the same complex adaptive feature from different evolutionary lines is tiny. The probability is small that the ancestor of the present-day squid could have given rise to a line that developed high intelligence.

What many SETI scientists pin their hopes on, however, is the phenomenon of evolutionary convergence. Sometimes different evolutionary lines arrive at the same solution to the only problem that matters — namely, keeping an organism alive long enough for it to pass on its genes. The classic example of evolutionary convergence is flight: birds, dinosaurs, fish, insects, mammals and reptiles all independently evolved the ability to fly. Another oft-cited example is the streamlining of marine creatures: species widely separated in evolutionary terms can nevertheless look similar. But these are convergences at low levels of complexity. It is not surprising that different creatures found that being airborne was a good way of escaping from predators, or that separate species discovered the benefits of cutting quickly through water. So the relevance of these examples of convergence to the SETI debate is minor. Enthusiasts of SETI have always argued that a more convincing example of convergent evolution is the eye.

The eye is an incredibly complex and specialized piece of machinery. That it can evolve at all is really rather wonderful. Yet it seems to have evolved independently at least 40 times, and perhaps as many as 65 times. Furthermore, eyes employ at least a dozen fundamentally different designs. For example, the compound eye of the insect is totally different in design from the camera eye of the vertebrates; it seems that the eyes of insects and vertebrates must have evolved separately. Even eyes that appear superficially to be the same — for example, that of the squid and of man — on closer examination show differences in detail. And when you consider that the last common ancestor of squid and humans was probably a sponge-like creature that lived half a billion years ago . . . well, it seems certain that the two types of eye evolved separately. That they look the same is a perfect example of convergent evolution. Or is it?

In 1993, Walter Gehring and Rebecca Quiring were studying the genetics of fruit flies.229 They found a gene — called eyeless — that seemed to act as a master control gene for the formation of an eye in fruit flies. By suitable manipulation, they could "turn the gene on" in different places and have a fly sprout an ectopic eye on its wing or its leg or its antenna. Eyeless was not the gene "for" an eye — the way genes work is much more subtle — but it seemed, among other functions, to orchestrate the action of thousands of other genes that form an eye in the early development of an embryo.

It soon became clear that the fly eyeless gene was similar to a mouse gene called small eye. A mouse with a defective small eye gene develops shrunken eyes. Furthermore, the gene is similar to a human gene responsible for the condition Aniridia, sufferers of which can have defects of the iris, lens, cornea and retina. When geneticists made a detailed comparison it was discovered that the "eye genes" in these three quite different species — fruit fly, mouse and man — were essentially identical in two crucial locations.

Georg Halder and Patrick Callaerts decided to implant the mouse small eye gene into a fruit fly. The gene worked. It caused the fly to develop ectopic eyes — fruit fly eyes, not mouse eyes. The eyes were not wired to the brain, but they looked like normal insect compound eyes and they responded to light.

All of the phyla that scientists have studied carry some form of the eyeless gene. These findings cast doubt on the received wisdom that eyes are an example of convergent evolution, because if animals really did evolve the design of their eyes independently, then one would expect them also to have evolved their own genetic signaling system. There would be no reason why a mouse gene could control the development of a fly's eye — one would expect them to "speak different languages." Perhaps, then, the last common ancestor of phyla as diverse as vertebrates, cephalopods, arthropods and nemerteans already had an eye and a version of the eyeless gene. The jury is still out, but it seems increasingly likely that the eye evolved only once — and the different visual systems we see around us are the result of evolution playing variations on an existing theme.

If the eye arose only once, then what chance is there of something even more complex — high-level intelligence — arising independently from different evolutionary lines?

0 0

Post a comment