We were meant to be...

If you think human intelligence is unique, think again. Its evolution on Earth was inevitable and it may have sprung up all over the Universe, argues Simon Conway Morris

RERUN the tape of life, as the late Stephen Jay Gould repeatedly observed, and the evolutionary outcome will be completely different. This time aardvarks, snapdragons and humans. Next time or elsewhere, who knows? The prevailing view of evolution is that life has no direction - no goals, no predictable outcomes. Hedged in by circumstances and coincidence, the course of life lurches from one point to another. It is pure chance that 3 billion years of evolution on Earth have produced a peculiarly clever ape. We may find distant echoes of our aptitude for tool making and language and our relentless curiosity in other animals, but intelligence like ours is very special. Right?

Wrong! The history of life on Earth appears impossibly complex and unpredictable, but take a closer look and you'll find a deep structure. Physics and chemistry dictate that many things simply are not possible, and these constraints extend to biology. The solution to a particular biological problem can often only be handled in one of a few ways, which is why when you examine the tapestry of evolution you see the same patterns emerging over and over again. Gould's idea of rerunning the tape of life is not hypothetical, it's happening all around us. And the result is well known to biologists - evolutionary convergence.

Organisms faced with the same challenge repeatedly arrive at the same solutions. And intelligence is one of these solutions, together with its close companions, advanced societies, communication and tool use. Of course, not every species on Earth is moving in this particular direction - life is very diverse - but intelligence does appear to be an inevitable outcome of evolution. That's why, in my view, either Homo sapiens or something very like us was bound to evolve. And if this is true on Earth, why not anywhere? If the search for extraterrestrial life succeeds, perhaps the biggest surprise will be how similar all intelligent life forms are.

The idea of convergence isn't new. Ask a biologist what the equivalent of a dolphin looked like in the Mesozoic, around 200 million years ago, and they'll say an ichthyosaur. The streamlined bodies of the mammal and the reptile cleave the ocean in much the same way, even though the ichthyosaur evolved from something like a lizard and the dolphin from something like a dog. It's a textbook example of convergence and perhaps typical of the way most people - even biologists - view the phenomenon. They often label instances of convergent evolution as "remarkable" or even "uncanny", but they don't see it as significant in the overall scheme of things. I disagree. My view is that the ubiquity of convergence makes it crucial for understanding the history of life. It allows us not only to confirm what must exist, but also to infer what can never exist. It provides a set of signposts to the highways of life.

On Earth, convergence operates at many levels, all the way from proteins and other biological molecules, to intelligence and social organisations. Consider vision. Evolution has come up with two main alternatives to the problem of seeing. The solution favoured by all insects is a compound eye with multiple lenses, which has evolved at least twice in the arthropods and three times elsewhere. Camera eyes like our own are even more interesting. Not only have they evolved independently at least seven times, but typically they are found in active, predatory and intelligent animals. All vertebrates have them, but so too do octopuses and squids - molluscs whose ancestry is very far removed from ours.

What's more, although compound and camera eyes perform the same function, they do it in markedly different ways, which helps explain why some animals have evolved one and some the other. For example, calculations by Kuno Kirschfeld from the Max-Planck Institute for Biological Cybernetics in Tübingen, Germany, show that if we had compound eyes they would need to be at least a metre across to obtain as good an image of the world as our camera eyes do.

What is true of vision is also true of the other senses. In every case, convergence is the rule. So, for example, although insects don't have noses, their method of smelling is strikingly similar to that of vertebrates. Both rely on the same configuration of nervesbut the patterns must have arisen independently because the "sniffing" proteins - and their genes - are different. And convergence even extends to senses that are far removed from our own experience. Fish, for example, feel pressure fluctuations via a sensory structure along their length known as the lateral line system. A strikingly similar system has evolved independently in octopuses, and again in the antennae of a swimming shrimp. In all cases, the structures allow them to detect water currents and nearby animals.

Even more extraordinary is the convergence between two groups of freshwater fish, one from South America and the other from Africa. Both have evolved an effectively identical mechanism to generate and receive electrical signals, a similarity that even extends to the algorithm they use to avoid jamming each other's signals. Interestingly, both groups - but especially the African fish, which are known as mormyrids - have massively enlarged brains. Presumably this is to deal with the complexities of living in an electrical world where both social communication and navigation in the murky waters depend on rapid and precise pulses of electrical information.

Mormyrid fish seem to "see" using electricity. And this isn't the only instance where evolution has co-opted one sense to stand in for another. By studying brain structures, Ken Catania from Vanderbilt University, Tennessee, has found that star-nosed moles "see" their dark subterranean world through their highly sensitive nose tentacles, while naked mole rats use their incisor teeth. These cases where different sensory inputs are integrated, together with the numerous examples of convergence, suggest that despite obvious differences in the ways animals sense their worlds, the basic processing may be much more similar than we realise. So, not only do sensory systems as different as sight and electrical reception each show multiple convergences, but all may lead to an equivalent mental map.

Minds, therefore, are not accidents of evolution, and the same is probably true of advanced intelligences. All living things must solve the problem of monitoring and responding to their external environment. Animals usually employ some sort of nervous system, so it is hardly surprising that we find the repeated emergence of complex processors in the form of brains. They take many forms, from the amazingly intricate and miniaturised insect brain, to the lobular brain of octopuses and the astonishingly complex vertebrate brain. But whatever size or shape the brain, it provides huge potential for solving problems of survival, which is why animals that have one show many important convergent behaviours.

Octopuses and bees, for example, can learn and memorise. Bees sleep. Some investigators have even gone so far as to speak of the temperaments and personalities of octopuses. And we are patently not the only conscious and intelligent vertebrates. Even so, surely there are limits to convergence: can we really identify anything that even approximates to the human mind?

Undoubtedly our minds have much in common with those of the great apes, but that's to be expected given our evolutionary proximity. More surprising is the probably even closer similarity between our mental capacities and those of dolphins. These cetaceans, of course, have remarkably large brains, which were only outstripped in size by the hominids about 1.5 million years ago as the cranial capacity of early Homo burgeoned. But the dolphin brain differs from the primate brain in many significant ways, ranging from microstructure to the emphasis and development of particular regions and lobes. As Lori Marino from Emory University, Atlanta, has shown, this makes the similarities in the intelligence and social behaviour of dolphins and apes, including humans, all the more remarkable.

Dolphins recently demonstrated beyond doubt that they are capable of recognising themselves in mirrors - joining an elite group that includes the chimp and African grey parrot who have passed this test of self-awareness. Dolphins have a sophisticated communication system. Using signature whistles they can recognise at least a hundred other individuals. They are also accomplished mimics. Dolphin social structure falls into a type known as fission-fusion where groups of individuals form fluid alliances, joining and splitting in a medley of associations, some temporary, others quite stable. The same arrangement is found in chimp societies. Despite a lack of manual dexterity, dolphins have been known to use tools. One group has learned to fit conical sponges onto their rostrum so that they can root about in the seabed without being harmed by stonefish and other venomous animals. Some experts even believe that dolphins are capable of abstract thought, based on their remarkable ability to understand and act upon communications with humans.

Inevitable intelligence

Dolphins will never be just like us, not least because their aquatic habitat poses different challenges. But by thinking about life in terms of the probability of the emergence of biological properties, rather than particular evolutionary histories, we get some interesting new perspectives. If something that looks very much like humanoid intelligence can evolve at least twice in two very different settings, then surely complex intelligences can evolve repeatedly. This also suggests that other complex end products of evolution are predictable. In which case, life is much less quirky and fortuitous than is often thought. When convergence is the rule, you can rerun the tape of life as often as you like and the outcome will be much the same.

Convergence means that life is not only predictable at a basic level, it also has direction. If our type of intelligence is inevitable, then so too are other associated evolutionary features that help define what it is to be human. Take, for example, vocalisations. The twitterings, yattering, howls and whistles of animal expression may sound like a cacophony - even when they carry a special meaning to a particular species - but they share deep-seated similarities that have evolved repeatedly. For example, the neurology and structure of bird vocalisations are strikingly similar to those of mammals.

One notable feature is babbling - the familiar sounds made by human babies as they experiment with sound production before words emerge and language crystallises. It turns out that young birds, dolphins and monkeys also babble. What's more, there's a growing suspicion that embedded in animal vocalisations are unappreciated complexities that prefigure human language. For example, Klaus Zuberbühler from the University of St Andrews has even found syntactic rules in monkey alarm calls.

And the list of convergences goes on. Chimps using stones to crack open nuts or sticks to fish for termites, are clear reminders that tool use is far from being a human prerogative. In some ways the versatility of tool-making in birds, especially New Caledonian crows, is even more impressive (New Scientist, 17 August, p 44). The ability to choose appropriate materials and manipulate them may be skills that evolve only sporadically, but their recurrent and independent emergence suggest that tool manufacture is another evolutionary inevitable. With all these examples of convergence it is difficult to avoid the conclusion that the evolution of a humanoid creature was very much on the cards since at least the time of the Cambrian explosion, more than half a billion years ago, when all the major groups of animals we see today originated.

There are two even more intriguing implications of convergence. The first concerns the hunt for extraterrestrials, especially intelligent species. No Earth-like planets have been discovered as yet, but this is simply a limitation of existing astronomical techniques. Most astronomers are confident they exist. What's more, deep space contains vast quantities of organic molecules ready, when combined with water, to seed any suitable planet that coalesces out of a whirling solar nebula. There's a general assumption that life is universal. So, if the evolution of human intelligence on Earth is unremarkable, then the history of life here might provide a reliable guide to alien biospheres.

Admittedly, a humanoid intelligence is not the only game in town. Alternatives may well exist (see "An Alien Intelligence"), but whatever its form, once hyper-intelligent life has emerged, it will begin to dramatically and irrevocably remould the planet it inhabits. What happens next is conjecture. Maybe the next step will be self-destruction. Perhaps that is also a biological probability, but surely at least a few species will make the leap to the stars. Given this, the absence of extraterrestrial signals, let alone visitations, seems to be puzzling. My own suspicion is that Earth-like planets are in very short supply. So even though the potential for the emergence of intelligence similar to our own may be universal, we might still be alone.

The second implication of evolutionary convergence is yet stranger. The complexity of biological systems at all levels, from proteins to social systems, means that the number of possible combinations is truly gigantic. Clearly, terrestrial evolution has explored only a minute fraction of these - life occupies very few biological "hyperspaces". Most biologists would argue that any alien biosphere would occupy some other hyperspaces that are remote from any terrestrial equivalent. Evolutionary convergence, however, suggests that this assumption is incorrect. Despite the theoretical immensity of possibilities, practically all of the alternatives are permanently uninhabitable. These uninhabited hyperspaces can never be occupied. Life, in the final analysis, has surprisingly few choices.

It's a bold prediction, but let me conclude with an even more fantastic suggestion. Suppose that in a particular biological hyperspace there are several isolated zones, each representing a stable node of occupation. If life has one origin, then to reach a remote point in this hyperspace somehow evolution must "learn to tunnel". This clearly happens during evolutionary convergence when very different organisms independently hit upon the same solution when faced with the same problem. 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. Evolutionary convergence may yet have some strange destinations.