Monday, April 30, 2012

Metaphysics in science, Part II: Life in a cave

'Metaphysics' is an area of philosophy that goes back to the ancients.  The word has many meanings, but for our purposes it is a contrast between the hard-nosed reality of the world we can see, touch, and feel, and the world of ideas in our heads.

The archetype of metaphysics, that illustrates the basic idea, goes back to Plato and that's what we'll use in this series of posts.  In his Republic, he likened the world we live in to a cave, in which we sit facing the wall, able only to know of what we see there, which are but shadows of reality.  That's because there is a 'real' world outside the mouth of the cave.  Light shining into the cave produces shadows of the objects in that world onto the wall of the cave.  All the encaved people can see are these shadows, ephemeral indications of the True things but that the people cannot actually see.

So, in this metaphysical view we see chairs, dogs, good and evil only as imperfect shadows of their true ('ideal') nature.  We assume these ideals, but have to infer them via their imperfect representations.

Other philosophers objected to this idealism, saying that the idea of the actual existence of these metaphysical entities like 'dog' or 'good' were mistaken, and that what we actually see is all there is.  For our purposes we can refer to this purely empirical view as Aristotelian.  There may be chairs and dogs, but there is no 'the chair' or 'the dog'!

In the early part of the modern Age of Science, René Descartes suggested a fundamental dualism between mind and matter, that can be seen as an extension of these differences.  There was matter--the real stuff we can grasp and measure and poke electrodes into, and then there was something resident, but immaterial in our heads (call it mind, soul, or whatever).

Science has become hard-nosed about reality: it is only what we can touch, feel, or measure physically.  We know that consciousness is elusive to our present understanding, but believe (or, at least, operationally act as if) mind is just the 'emergent' result of our neurons working by the billions in the confined physical space of our heads.

More importantly, in the Age of Science it has become routine to sneer at metaphysics.  Science, perhaps arrogantly, perhaps accurately, views the testable empirical world as the only world.  It wants to disabuse us of abstract, undemonstrable soft-headed metaphysical notions.

But let's take a look at some of the issues from a biological viewpoint 
After Mendel, the idea of 'genes' developed.  Genes (at the time inferred from parent-offspring resemblance, as in Mendel's peas) were assumed to exist, but we didn't know what they were.  Darwin's idea of evolution invoked these entities (he didn't use the term itself, which was coined much later).  The idea was of things transferred in Mendelian patterns; we had the audacity to name them, though we had no idea what they actually were!

But other schools of thought, Marxism in particular, strongly resisted such notions as being metaphysical:  we say there are genes, but we never actually see one, so they really are like Plato's shadows in the cave.  And as materialists we had long ago abandoned such infantile thought-centered views of reality.  Instead, they argued, genes were metaphysical entities that are invoked by the capitalist world to justify ruthless Darwinian competition and cruel inequality, which thus justified social inequity.

Marxists (in the Soviet Union, and in the particular incarnation of the agricultural policy under a  plant biologist named Trofem Lysenko) argued that individuals are not condemned by their metaphysical 'genes', and can improve themselves, and incorporate (that is, literally build into their material nature) their improvements. These improvements can then be transmitted to their progeny: real gains are thus made without invoking laws of competition and the like, which they argued were just convenient metaphysical entities.

It turned out, of course, that the hypothesized 'genes' really do exist as material, molecular entities.  Good-bye metaphysical notions!

But is this really as clear-cut as all that?  And does it matter?  Do we really live, or give the lie to, our materialism?  Are we true to our sneering at metaphysics?  The answer, even for hard-nosed science, is at least not totally clear.  Let's look at genes first, and then evolution.

Genes:  Platonic shadows of nonexistent ideals?
There are countless references to genes these days.  Genes have names.  So, for example, there is the 'beta globin' gene.  'The' normal variant of 'the' globin gene is called HbA.  But this gene is the one that, in one of its variant forms (called HbS), confers sickle-cell anemia (a red blood cell disorder) on those who carry the HbS  form.  We can talk, seemingly sensibly, about 'the globin gene.'  But what is it?  Indeed, is there a globin gene?

In fact, the globin gene is a Platonic ideal if there ever was one.  What we each have are Aristotelian manifestations of what we call 'the' globin gene.  Like chairs or dogs, they can be different in each person (as in the HbS), but there is no 'the' globin gene.

The same goes for the driving force of so much of today's biological science:  'the human genome': Does it exist?  We think the answer is, manifestly not!

The human genome is a phrase that today refers to a single reference sequence of human DNA, available for all to see online (well, you can't actually see it, but you can scroll your screen along its sequence of A,C,G, and T nucleotides, 3+ billion of them).  We have agreed to accept this as 'the' human genome.  But it in reality is neither 'the' human genome sequence nor even 'a' human genome sequence!  In fact, it's the sequence of bits of DNA from several people (each of whom had two copies, one inherited from each of their parents).  Nobody has, and nobody ever has had, 'the human genome' sequence.  And there is debate about whether this is an appropriate abstraction.  For example, the donor humans were 'normal' when their DNA was sampled, but undoubtedly will die of something, that may be affected by their genes.  So it's not a 'normal' genome in any serious sense; it was not randomly sampled from our species.  So some have suggested that we should collect a set of sequences as references, or use some other kind of abstraction that incorporates known sequence variation in humans.  That, however, would still be a judgment about what to consider as our Platonic abstraction representing the DNA that we each carry around.

It is no secret (and no problem, either) that the idea of a reference sequence is a baseline by which we compare other human DNA sequences to identify and examine differences, variation, and the like.  That's a convention, such as that a red light means 'Stop'.  If all we see are actual instances of human DNA, then there is no 'the human genome sequence' any more than there is 'the chair'.  So we formally accept the notion of a Platonic abstraction, and we think every scientist would argue that this is just a pragmatic abstraction that makes our daily work possible, anchoring it so we can be objective about human genetics.  There isn't any particular problem about its metaphysical nature.

It's something like pi, the ratio of a radius of a circle to its circumference, or the square root of 2.  These numbers don't actually exist  in a sense (their decimals go on forever), but they are immensely useful and their 'results' are seen all over the place.  For that matter, there isn't a perfect 'circle', either: only its Platonic ideal.

The Marxist allegation that our use of 'gene' was metaphysics was right in a sense, but it was not 'just' metaphysics, and has proved to be correct in the sense that real genes do exist as empirical, material, molecules we can study, that are transmitted, that change, and that affect our traits.  The guess that inherited elements exist was based on Mendel's work and much else, and the idea of a 'gene' was a practical guide used very effectively to find the actual manifestations of the idea.  We were able systematically to do this because research consistently homed in on manifestations of this guiding, or one can say, reference ideal.

Evolutionary theory: what sort of reality is it?
Well, if we can see the issues in terms of material things like genes, what about less clear things like 'good'?  Does 'good' exist in any ideal sense, other than in various manifestation that, indeed, must be judged individually by individuals, and hence may not really have actual existence as entities?  Of course, philosophers make their living dealing with such questions, but we can turn to another manifestation (so to speak) of this kind of question: scientific theory.  Let's take Darwin and evolutionary theory.

Here we are dealing with processes, not things.  Species exist, again if we recognize that we give labels to Platonic ideals (e.g., 'type specimens'), and that the idealism is just a path towards understanding the real world that does exist.

What is evolution, or for example, natural selection?  Are these abstract concepts that exist on their own, or are they instances?  If the latter, what are they instances of ? What is the standing of a theory?  In particular, if the chair you're sitting on is, after all, a chair, one would have no objection to our speaking about chairs generally as if the Platonic ideal existed.  That is, the assumption that chairness exists is not particularly worrisome.

But what if a suggested ideal in this abstract sense is a theory?  One might say that the evolution of antibiotic resistance by bacteria is an instance of natural selection.  But what we want to know is whether 'natural selection' is somehow 'out there' as an ideal, overarching truth, or instead is just a term we give when we observe something happening.  This is not a trivial question.  That's because we know not everything is a chair, or dog, so we have no difficulty dealing with symbolic abstractions like 'chair' and 'dog': it doesn't makes us imagine that there is a true, ideal 'dog'.

But a theory is supposed to be universal, ubiquitous.  For example there is a big difference between saying 'some times natural selection produces adaptations'  and asserting that 'adaptations are due to natural selection.'

Here a way to see that there's a problem.  If a theory represents a 'law of nature' for example, one can ask whether that very idea invokes Platonic realities.  We don't just want to explain an apple falling on Newton's head as 'gravity was involved in this instance', as if other apples need not fall on heads.  We want gravity to be unexceptioned, universal, out there in the distant galaxies beyond our telescopes, and inside our cells: 'out there' in some real not just imagined sense.  And what about the evidence for a theory when it exists.....or doesn't?

These are important issues of practical value, and were at the core of modern ideas about science, including the idea of induction and replication: repeat observations show the general nature of something, and the idea of deduction: with a theory we can reliably predict consequences if we observe causes.

We will continue to explore this in the next posts.

No comments: