Cheaper than printing it out: buy the paperback book.

The difference between wild evolution in nature and synthetic evolution in computers is that software has no body. The kind of software you load with floppy disks is straightforward. If you alter the code (for the better, you hope), you execute the program and it fulfills its orders. There is nothing between what the code is and what it does, except the wiring of the machine it runs on.

Biology is vastly different. If we take a hypothetical hunk of DNA as software code, and alter it, there is a consequential body that must be grown before the effects of the alteration can manifest itself. The development of an animal from fertilized egg, to egg producer may take years to complete; so the effect of that alteration can be judged differently depending on the stage of the growth. The same initial alteration of code can have one effect on the growing microscopic fetus and another effect on the sexually mature organism, if it survives that long. In every case, between the code alteration and the terminal effect (say, longer fingers), there is a chain of intermediate bodies governed by physics and chemistry -- the enzymes, proteins, and tissues of life -- which also must be indirectly altered by the software change. This vastly complicates mutational variation. Programming computers is no longer an adequate comparison.

You were once the size of a period. For a brief time you tumbled about as a multicellular sphere, much like pond algae. Currents swept and washed over you. Remember? Then you grew. You became sponge life, tubular, all gut. To eat was life. You grew a spinal cord to feel. You put on gill arches in preparation to breathe and burn food with intensity. You grew a tail to move, to steer, to decide. You were not a fish, but a human embryo role-playing a fish embryo. At every ghost-of-embryonic-animal you slipped into and out of, you replayed the surrender of possibilities needed for your destination. To evolve is to surrender choices. To become something new is to accumulate all the things you can no longer be.

While evolution is inventive, it is also conservative, making do with what is available. Biology rarely starts over. It begins with the past, which is distilled in the development of the organism. By the time an organism arrives at the end of its natal development, the millions of tradeoffs it has incurred forever block the chance to evolve in certain other directions. Evolution without a body is limitless. Evolution with a body, wrapped in development and prevented from retreating by its current success, is bound by endless constraints. But these constraints give it a place to stand. It may be that for artificial evolution to get anywhere, it too may need to wear a body.

When there are bodies in space, there is time. Mutations bloom in a body grown -- in time's dimension. (That's something else artificial evolution has little of so far: developmental time.) To alter development early in the embryo is to fiddle with time. The earlier a mutation expresses itself in embryonic development, the more forcefully it will resound through the organism. This also loosens the constraints against failure, so the earlier the mutation is in development, the less likely it will be workable. In other words, the more complex an organism becomes, the less likely a very early change will survive.

Early developmental change has the advantage that a small mutation can affect a suite of things in a single blow. An appropriate early tweak can invoke or erase ten million years of evolution. The famous Antennapedia mutant of the Drosophila fruitfly is an example. This single-point mutation engages the leg-making apparatus of the embryo fly to build a leg where its antenna should be. The afflicted fly is born with a fake foot sticking out of its forehead -- all triggered by one tiny alteration of code, which in turn triggers a suite of other genes. All kinds of monsters can be hatched this way. Which leads developmental biologists to wonder if the self-regulating genes of an organism might be able to tweak the genes governing these early suites into useful freaks, thus bypassing Darwin's incremental natural selection.

The curious thing about monsters, though, is that they seem to follow internal laws. While a two-headed calf may seem to us to be randomly defective, it isn't. When biologists studied freaks they found that the same type of monstrosities appeared in many species, and that their freakishness could even be categorized. For instance, a cyclops -- a relatively common freak in mammals, including humans-born with a single centrally positioned eye, will almost always have its nostrils located above its eye. This is true regardless of the species in which it appears. Similarly, two-headedness is much more common than three-headedness. Since neither mutation is a variation that offers reproductive advantage, since few of these freaks survive, natural selection cannot be selecting one over the other. This mutant order must be internally generated.

In the early and mid-19th century a French father and son team, Etienne and Isidore Geoffroy Saint Hilaire, devised a classification scheme for natural monsters. Their taxonomy of mutants paralleled the Linnean system of natural species: every monstrosity was assigned a class, order, family, genus, and even species. Their work became the foundation of the modern science of monsters -- teratology. Orderly form, the Hilaires implied, extended beyond natural selection.

Pere Alberch, at the Museum of Comparative Zoology at Harvard, is the modern spokesman for the importance of teratology in evolutionary biology. He interprets teratologies as overlooked blueprints for strong internal self-organization within living organisms. He states, "Teratologies are a superb document of the potentiality of a given developmental process. In spite of strong negative selection, teratologies are not only generated in an organized and discrete manner but they also exhibit generalized transformational rules. These properties are not exclusive to teratology; rather they are general properties of all developmental systems."

The orderly makeup of monsters -- it is after all a well -- formed foot which erupts out of a mutant Drosophila's forehead -- speaks of a deep underlying internal force which helps guide the outward shape of organisms. This "internalist" approach differs from the orthodox "externalist" approach of most adaptationists who see ubiquitous natural selection as the major shaping force. As a dissenting internalist, Alberch writes:

The internalist approach assumes, and this is a key assumption, that morphological diversity is generated by perturbations in parameter values (such as rates of diffusion, cell adhesion, etc....) while the structure of the interactions among the components remains constant. Given this assumption, even if the parameters of the system are randomly perturbed, by either genetic mutation, environmental variance or experimental manipulation during development, the system will generate a limited and discrete subset of phenotypes. Thus the realm of possible forms is a property of the internal structure of the system.

Thus we have two-headed freaks for perhaps the same reason we have bilateral arms; most likely neither is due to natural selection. Rather, internal structure, particularly the structure of the genome, and the accumulated morphogenesis of development, may be an equal or greater influence upon the variety of biological organizations possible.