A common misunderstanding of evolution is that it is all about genes. Genes are important, but organisms do a lot, too. One of the things they do is modify their environment to suit themselves, and to suit their progeny. This is called "niche construction", and I came across this rather nice essay on Jim Harrison's inanis et vacua blog on the topic, which I am posting here with his permission.
A note to add to this is the book of essays published a few years ago on this and related topics, putting the discussion in a more general context:
There is a revolution in the thinking of evolutionary theory right now. It may take, as Jim notes, a while to become widely known...
Evolutionary theory continues to develop far from the noise of irrelevant public controversies about intelligent design. Recent thinking not only departs from the so-called Neo-Darwinian synthesis of the 40s and 50s, but has already gone beyond more recent, but already long-in-the-tooth enthusiasms such as punctuated equilibrium, selfish genes, and sociobiology. Much of what's new is largely inaudible, not because it is impossibly complicated but simply because it has yet to find its Huxley, Gould, or Dawkins. Nobody has figured out how to glamorize what may turn out to be a slow revolution in how we think about evolution. I don't know if I can solve this public relations problem, but perhaps I can provide a useful oversimplification of what's going on. After all, though I was inspired to write this little piece by a recent book, Niche Construction: the Neglected Process in Evolution by F. John Odling-Smee, Kevin N. Laland, and Marcus W. Feldman, I've been thinking down these lines myself for decades and can perhaps offer, if not new ideas, at least some new ways of explaining the new ideas.
The usual account maintains that the environment of organisms guides their evolution by establishing the rules of the game. Natural selection tends to favor individuals better suited to existing conditions, and these better-adapted organisms produce more offspring. What the organisms themselves do or experience is irrelevant to the process because acquired characteristics are not passed on to the next generation. What's becoming obvious is that this very general picture is quite misleading, not because Lamarck was right and Weismann wrong, but because the activities of living things change their environment and therefore change the selection pressures that govern subsequent natural selection. To paraphrase Marx, and not accidentally, environments make organisms; but organisms make environments. Or, to quote Odling-Smee, et. al, "Rather than acting as an ‘enforcer' of natural selection through the standard physically static elements of, for example, temperature, humidity, or salinity, because of the actions of organisms, the environment [can be] viewed as changing and coevolving with the organisms on which it acts selectively."
Let's begin simply. Many organisms move purposively; even bacteria can and do choose their environments to some degree. The planktonic species that rise and sink in the water column on a daily basis are not adapted to the ocean in general. They're adapted to a narrow spatial and temporal zone inside the ocean. Many simple organisms pick their poison in more active ways. The bacteria that rot your teeth are not adapted to your mouth. They cultivate a much narrower plantation whose chemical and mechanical properties are maintained by their own collective activities. On a larger scale, innumerable species of animals burrow and have accordingly developed physiological and behavioral characteristics that make sense in a protected subterranean world but not outside of it. Or consider us. Distinctly human characters such as extended infancy and jumbo brains only make adaptive sense in a human world where vulnerable babies can be protected for many years and surpluses of food make it possible to maintain an energetically expensive nervous system.
It would be exceedingly easy to multiple examples. Indeed, once the principle is recognized, the relevant database is essentially coextensive with natural history. Charts in the Niche Construction list page after page of examples from every kingdom and phylum. What's involved is not so much a discovery as a gestalt switch, seeing vases instead of faces. That sort of mutation has happened before in the history of the sciences. Thus the equations in Einstein's 1905 relativity paper would have seemed quite familiar to many of the physicists of his time because the originality of the work did not lie in the discovery of any novel physical phenomena but in a reversal of perspective that suddenly made sense of a whole series of earlier results. Einstein didn't invent the Lorentz transformations, which already, among many other things, implied that e = mc2 to anybody who passed Intermediate Algebra. Einstein looked at something old in a new way. But the Einsteinian eureka moment wasn't just edifying: new science, including new discoveries flowed from it. Even granting the conceptual validity of the perspective, does rethinking evolution with an emphasis on the initiative of the organism and the evolution of environments change anything? In particular, will it allow biologists to solve existing puzzles in evolution and ecology; manage the environment; improve public health; or, crucially, attract grant money, the crankshaft and sine qua non of the scientific juggernaut?
Let me suggest two areas where the notion of niche construction seems to me to have immediate explanatory value: explaining the pace of evolution and explaining the evolution of complicated nervous systems:
1. Niche construction can result in positive feedback, thus drastically accelerating genotypic as well as phenotypic change. It's likely that the practice of herding cattle greatly increased the adaptive value of tolerance to milk sugars for pastoral people, for example; but as a larger percentage of the population became lactose tolerant, dairy farming became more valuable. Or consider what may be the most sumptuous niche construction project on the planet, human language--The House of Being, according to Martin Heidegger. Absent language and the ability to accumulate knowledge that it makes possible, it's hard to see how the enormous energetic cost of the human brain could have been amortized, not to mention the many physiological and behavioral alterations required by an oversized head and an interminable childhood. A smart animal without words is very like the proverbial Michelangelo without hands. The appearance of even rudimentary language surely had a significant multiplier effect on the adaptive value of intelligence.
I expect that evolution as a whole has been accelerated--if not simply made possible, period--by the niche construction mechanism. Over and beyond what specific adaptations become feasible in a partly controlled environment, the activities of living things make their surroundings generally more predictable and thus focus natural selection in much the same way a judge makes a trial manageable by narrowing down what is at issue in each case. In a safer world, the right ones are more likely to die without reproducing, the trap door spider who isn't good making a trap rather than the spider who just happened to get stepped on. There are relatively few truly random deaths in the human world, for example, so that the deaths that do occur are likely to have biological significance while in a postulated state of nature, the signal provided by differential mortality would be largely swamped by the background noise. Civilization may indeed allow some of the weak and sick to survive and even reproduce, but it also spares many healthy and strong individuals who would otherwise have died of irresistible illnesses or accidents.
Odling-Smee, et. al. point out that niche construction can also retard evolutionary change. The Galapagos woodpecker finch has adopted a life style similar to woodpeckers, for example, but without the developing a similarly sharp, pointed bill or long tongue. The culturally transmitted use of a cactus spine as a tool to peck for insects has apparently rendered these anatomical adaptations unnecessary much as knives, forks, and fire allow people to eat meat without fangs or claws.
2. The learning that organisms do in their lifetimes can have an effect on subsequent generations by providing offspring with a better environment or by the cultural transmission of useful information. The value of learning, however, depends crucially on the regularity of the environment. In a chaotic world, drawing conclusions is risky or pointless since there is literally nothing to learn. Where fine-grained knowledge of the world doesn't avail, the most adaptive strategy may be learn to slowly so that only the big, coarse lessons are absorbed after many trials. In a world that makes sense, the fastest learners have a huge advantage. One of the general consequences of niche construction, therefore, may be to make greater intelligence more adaptive by ensuring that animals grow up in a simplified, if not positively dumbed-down world.