Genes, evolution, science education, and science

"One well-known path to change is a heavily favorable mutation in a single gene. But it may be well known only because it is easy to study. Another path is exploitation of mildly favorable differences that already exist in many genes ... Theorists have argued over this point for years, but researchers have been able to address it only recently." ... Natural selection cuts broad swath through fruit fly genome, New York Times, 20 September, 2010

"Many scientists ... expected that surveys of our genomes would reveal considerable evidence of novel genetic mutations that have recently spread quickly throughout different populations by natural selection ... But it turns out that  ... most of the detectable natural selection appears to have occurred at a far slower pace than researchers had envisioned." ... How we are evolving, Scientific American, October, 2010

"The Human Genome Project has failed so far to produce the medical miracles that scientists promised.  Biologists are now divided over what, if anything went wrong, and what needs to happen next ... For years, many scientists held the hypothesis that certain common variants would be especially prevalent in people with particular diseases ... From the very beginning, the scheme was met with some dissent ... susceptibility to biologically complex diseases probably derives from inheritance of many rare disease-promoting variants that probably number in the hundreds - perhaps thousands - in any given individual."  ...  Revolution postponed, Scientific American, October, 2010

In three quite distinct realms of biology - genetics, human evolution, and biomedicine, a similar new story is emerging.  The convergence seems to me interesting for thinking not only about biology but also about science education and science in general.

Gregor Mendel and his pea plants made a highly significant contribution to biological understanding: the idea that there were discrete units of heredity, later called genes, that can be transmitted undisturbed from generation to generation and tend to assort independently and randomly.  But Mendel's way of establishing this also laid the seeds of a persistent misunderstanding of inheritance.  Mendel picked for analysis observable traits of his pea plants that bore fairly direct and simple relationships to genes. Many generations of genetic studies, biology laboratories, and textbooks later, what has stuck in most peoples' minds is not Mendel's most important and general insights but rather the notion that there is a fairy direct and simple relationship between genes and traits.  What people tend to remember is not what Mendel actually discovered but rather the simplification he made use of to discover it. 

The commonly emerging new story begins to correct this simplification.   The majority of recently evolved human traits seem now to have resulted from selection pressures that affect lots of genes, each making a small contribution to trait change, rather than a small number each making a large contribution.  Laboratory studies of artificial selection in Drosophila are showing the same pattern.  And its beginning to look like human disease susceptibility similarly is in general a trait influenced by many genes rather than a few.

In one sense, all this is to the good, indicative of a self-corrective force in science: older understandings are modified as new observations are made.  One hopes that biology courses, laboratories, and textbooks will quickly catch up, that enough emphasis will be placed on polygenic inheritance to offset the tendency to misread Mendel's findings, and subsequently the ones of many other investigators who similarly simplified the relation between genes and traits in order to shed light on other important issues.

There is, though, more to the story than this.   Yes, the notion of a close relationship between traits and individual genes is supported by observations, as scientific understandings should be, but the fact that the observations reflect an investigator predilection to seek out and report situations where such a close relationship existed was largely ignored, not only by consumers of science but by scientists themselves.  There was never the sort of broad collection of observations that would allow a neutral observer to make a reasoned estimate of the relative likelihood of polygenic as opposed to single gene influences on traits.  Despite this, many scientists were not only comfortable commiting their own careers to a presumption of simple relationships between genes and traits  but to spend and advise the spending of large amounts of of federal research money based on this presumption (cf Thoughts about science and medicine).  And to do so even after related lines of research made the polgenic relationship increasingly likely. We're down to thirty thousand genes or so in the human genome, a number that seems dwarfed by the number of traits one needs to account for. 

What happened?  Why is it only now that significant numbers of researchers are publishing findings in support of polygenic trait influences?  Do scientists develop collective mindsets about problems that persist for reasons relatively unrelated to a careful and critical analysis of relevant observations?  Are scientists, like other human beings, prone to betting on things, and recommending that others bet of things, that are consistent with common ways of making sense of things, no matter how ill-founded those may actually be?  Are scientists content to study what is "easy to study," whether or not it actually speaks to important deeper questions? 

The answer to all of these questions is yes.  Science is always done by human beings in a social context and factors of the kind mentioned are always involved.  This makes science less "objective" than it is often portrayed as being but it also gives it a a social cohesiveness and inertia than contributes significantly to its progress while sometimes also retarding it.   Rather than trying to disguise this and other features of the inherent subjectivity of science as something vaguely shameful, we ought to acknowledge it and its significance (cf Revisiting science in culture and The subjectivity/objectivity spectrum).  And teach about it, so that both we and non-scientists are fully aware of it, and can take it into account in evaluating the significance of scientific understandings. 

I think though that the question of why biologists were reluctant to embrace polygenic inheritance has a still deeper and more interesting answer and implication.  My guess is that if you seriously pressed scientists on why they preferred to base research on a presumption that traits reflect small numbers of genes rather than large numbers, the response would be an Occam's razor argument in one form or another: it is "simpler" to account for traits in terms of small numbers of genes and one ought to try out "simpler" explanations before moving on to more complex explanations.

Occam's razor has some value as a heuristic for scientific research (or inquiry of any kind): it usefully encourages one to entertain the possibility that what might at first appear to be quite complicated phenomena can be accounted for more simply.  But Occam's razor needs to be more clearly recognized as just a useful heuristic, and not  either an operating mode for what is being investigated nor an unchallengeable principle of scientific inquiry.  Biology is full of phenomena that at first glance might have seemed amenable to simple explanations but have proven in fact to require more complicated ones (cf From the head to the heart), a list to which one can now add more common patterns of gene/trait relationships.  

The Occam's razor argument is another example of the involvement of subjective factors in science, not only in that it may or may not be a good guide to what is being studied at any given time but also in that what is more or less "simple" depends to a significant extent on the current understandings of the investigator.  Unpredictable behavior in a system, for example, may seem more complicated, something to be made sense of in simpler terms, if one starts from the premise of a well-ordered universe with simple cause and effect relations.  If one starts instead from a premise that everything is constantly in random motion, then unpredictability is the simpler thing to account for and orderly, predictable behavior becomes the more complicated problem.  Similarly, if one starts with the premise that there is a single, observer-independent reality, then differences in the perception of reality by different people is the more complicated problem.  If one instead starts with the premise that reality is always observer-dependent, then it is similarities among people in perceptions of reality that is the more complicated phenomenon requiring explanation.   If one starts with the premise that the universe consists of independent elements that interact in simple uni-directional ways, lots of things seem more complicated than they do if one starts with a presumption that things are always interacting in multiple, bidirectional ways.

All this may seem a bit far afield from genes and evolution, but I think there is actually a closer relation than it might at first seem between the emerging story about genes and evolution, and thinking about the nature of science and science education.  As both scientists and science educators, we are prone to looking for, and telling, compelling stories, stories that give us simpler ways of making sense of a world that often seems inchoate beyond our understanding and control.  Such stories can be and often are useful within particular contexts, but ought never to be presumed to be universally valid.  It is not only that there are necessarily based on a finite number of observations and so always subject to revision give a future observation.  It is also that our practical preference for "simpler" stories is itself a significant bias on the stories we create.       

Evolution is a good simpler story, helping to make sense of, among other things, the enormous diversity of life forms we find around ourselves.  What though seems particularly interesting to me about the story of evolution is that the story itself contains a cautionary note about potential limits of simpler stories.   The story of evolution, of descent with variability coupled with differential survival,  can account for what is at the present, including, probably, the existence of consciousness and human intention,  but may be fundamentally incapable of predicting with any certainty what will come next.  The limitation has to do not only with a contribution of random events to the process but also with the fact that the process of evolution brings into existence things that in turn alter the process.  The appearance of photosynthetic procaryotes pushed evolution in directions it might otherwise not have taken.   The same is almost certainly true of the appearance of eucaryotic cells, of multicellularity, and so forth.  Consciousness and human intention are similarly altering the direction of evolution in the present in important but not fully predictable ways.  

What this suggests is that we need to be more aware of a distinction between accounting for the present and predicting the future.  The story of evolution not only involves random processes but also has no goal or objective.   For all these reasons, it serves to make sense of things in the present without also serving to fully predict the future. Given a tendency in many peoples' minds to equate "understanding" with prediction, it might make sense to more strongly emphasize this somewhat unusual feature of the story of evolution in teaching it.   The story of evolution is not the way to account for the present which in turn makes it possible to predict the future; it is instead a way to make sense of the present which opens new possibilities for conceiving futures. Maybe that's a good way to think not only about evolution but science (and inquiry) in general as well?  Science/inquiry is not simply an observer of a process but also a contributor to it,  in ways that open new possibilities but will always have somewhat unpredictable future consequences as well?   Maybe we should should make this point clearer not only in teaching about evolution but science and inquiry in general.