Saturday, May 14, 2016

Natural Selection’s Reasons

If we say, “I have a good reason for being here, today,” and “The reason you slipped is because there is ice on the path,” we are using the word “reason” to describe two very different concepts.  Both deal with cause-and-effect.  My reason for being here (whatever it may have been) caused me to come, and the ice on the path caused you to slip.  But one sentence involves a purpose.  I had goals in mind, and in pursuit of those, I used my ability for mental analysis (another sense of the word “reason”), and made the decision to come.  But no sentient being (presumably) had a goal in placing the ice on the path, let alone in causing you to slip thereon.  In that case, the word “reason” refers to an explanation only, and not to a purpose.

Interestingly the dictionary does not well distinguish these two senses of “reason”.  At least, the Webster’s edition that I consulted gives as the first definition “a statement offered in explanation or justification,” which, it seems to me, specifically conflates the two concepts – a justification involves a purpose, an explanation need not.

The distinction between these two concepts is very important when considering the workings of evolution by natural selection.  There are reasons that certain traits are selected (preserved), in the sense that there are explanations for their selection, but they were not selected “for a reason”.  Evolution has no purpose in selecting them.  Evolution is not, in fact, an agent that can conceive a purpose.   There is no end goal for the process of evolution.  In Aristotelean terms, evolution has efficient causes; it does not have final ones.  Evolution is a mechanical process  - as mechanical as the process of fusion in a star producing heat and light.  Everything accomplished by natural selection has a reason why it happens (explanation);  it does not happen for a reason (purpose).

It is hard to keep your head wrapped around this distinction.  We are so used to conflating the two concepts.  I do not think our use of the same word for both is the reason (explanation) for the conflation.  I think the conceptual conflation is the reason for the multiple meanings of “reason”.  For us, in our daily lives, so many things are explained, at least partially, by the reasons that people have for doing the things they do.  It is natural for us to think of chains of events in terms of purposes.  It is natural for us, also, to project agency onto things that do not actually have it (onto events that do not actually have a sentient agent behind them).  This is what the philosopher Daniel Dennett has called our “hyperactive agent detector”.  Being a Darwinian, he of course posits explanations – reasons – for how we might have evolved such a characteristic by natural selection.

Now waitaminnit.  Suppose we evolved a hyperactive agent detector, as Dennett suggests, because it was safer to err on the side of imagining a non-existent agent, for instance imagining a tiger when we hear the wind in the grass, than to possibly ignore a real agent (by assuming it is the wind when it is really a tiger).  The consequences (cost) of one class of error are greater than those of the other.  But don’t we have a purpose in evading the tiger?  Isn’t evading tigers a goal of ours?  So doesn’t purpose enter into the process of natural selection?  Don’t we have a reason for evolving that trait, as well as there being a reason why we did?

The answer is that, while yes, we may have a purpose in evading that tiger, no, that purpose does not influence the process of natural selection.  If I think I hear a tiger, I may well decide I want to get away; I conceive a purpose – to escape the tiger.  This may cause me to climb a tree. (Well, a really skinny tree.  Tigers can climb trees, I think, but they’re heavier than I am.)  My purpose may be intimately involved in my actually escaping the tiger.  But it had nothing to do with my possessing that nagging anxiety that asked, “What is that sound?  Was that a tiger?”  If Dennett’s hypothesis is correct (I suspect it is), then proto-people who possessed a hyperactive agent detector were more likely to survive, and have offspring, than proto-people who did not.  The fact that they may have DESIRED to survive and have offspring was not causally efficacious in the natural selection process.  In fact, the desires to survive and to have offspring (or at least, to have sex) THEMSELVES evolved, by natural selection, by exactly the same purposeless but explicable process:  proto-creatures that didn’t have such desires (or didn’t act as if they did, anyway) tended not to have descendants.

It is not necessarily true that purpose – the ordinary purposes of people like us – can NEVER have a role in natural selection.  Darwin, after all, modeled the idea of natural selection after the kind of selection practiced by (for instance) human stock breeders, who allow only animals with certain desirable traits to reproduce.  Humans are in nature, and stock breeders are an important element in the environment of their animals, just as are predators the environments we collectively call “the wild”.  Calling one form of selection “natural” and the other not is at least a little bit an artificial distinction that comes from a definition of “nature” as “everything but humans”.  But we humans and our cultures also constitute a very, very important part of our OWN environment.  Artificial selection can produce significant phenotypical changes in a very small number of generations – geological time scales are not required.  Mightn’t humans also have evolved to conform to our own cultural expectations, not just culturally (by learned/taught behavior), but also biologically, by natural selection, i.e., by genetic change?  Jonathan Haidt thinks so (and cites other researchers).  To some extent, don’t theories of group selection and the evolution of morals depend on something like this?  At the very least, it seems that such a hypothesis, if true, would strongly support certain group selection theories.

But the idea that humans and their purposes form an important part of the environment in which natural selection of humans takes place is very different than saying that evolution itself has a purpose, or that any specific trait arises “for a purpose”.  As easy as it is to slip into that kind of language, that just ain’t the way it works.  From time to time, new traits arise.  The ones that work well stick around, simply because they work well.  The ones that don’t work so well, go extinct.  If everything seems to fit in the end, it’s because the things that fit are the ones that are still around.  For now.