Friday, September 20, 2013

Is there truth? Does science find it?

I listened yesterday to the first installment of this seasons's BBC Radio 4 radio program, In Our Time. Happy it's back!  This episode discussed the life and work of the French Enlightenment-era genius and polymath, Blaise Pascal.

The Enlightenment period was an intellectual fervor that turned from authority to empiricism in the belief that understanding the real world comes from observation and empiricism, not just consulting authority (like the Bible or Aristotle).  Main names of that era include Galileo, Descartes, Newton, Bacon, and so on.  They all were part of the push to understand theories or laws of nature based on careful experiment and observation.
Blaise Pascal, 1623-1662

These thinkers, and others, rejected the notion that perfect truth could be seen in the classical or sacred texts, but that we could at least better understand the world through careful scientific methods.  But how 'better'?  Is there truth about the  world?  Does it follow universal regularities--laws of nature, as Newton and his intellectual descendants to the present day say--or not?  If so, the job of science was to identify and understand those laws.  But there is a further implication of that view: there must be a truth!

If truth exists, does science find it?
The popular idea about science is that it seeks the truth about physical nature and, for those not holding out for non-physical truth such as religion, even about the nature of 'mind'.  Perhaps the popular belief is that in many areas science has done its job already.  For most, however, science is seen as approaching better and better understanding of truth in various areas like physics, chemistry, geology, biology and others.  The idea is that truth exists and we approach it asymptotically, that is, gradually getting ever closer--even if we, as imperfect creatures, may never get 100% there.

Pascal calculator

Philosophers of science had a longstanding view that science works by gradually refining its insights with very focused methods that get closer and closer.  Yes, there may still be major or even transforming insights yet to come, and yes, we can't now know what they'll be.  Only when the next Newton, Einstein, Maxwell, or Darwin come along will some jump rather than gradual approach to truth occur.  Often we hear the phrase 'paradigm' to refer to the current theory, and 'paradigm shift' to such huge transformations of our view--such as when evolution replaced creationism: you can't look back, once you see the new view and realize that it is.....what?  'true'?  'better'?  more convincing or convenient?  more useful for engineering and business?

Blaise Pascal and others, apparently even Newton himself, did recognize that science wasn't necessarily approaching Truth.  Only God could know the truth and (according to many of them) so had we until the fall in the Garden of Eden.  Still, they felt we could approach truth, and many believed that mathematics was some ultimately true and consistent language for doing that.

In the mid-20th century Thomas Kuhn and others looked at science not as a purely objective path of a train moving to the station of Truth, but instead as a social phenomenon, a society of individuals who accept a given explanation (that's the 'paradigm') and press towards making that explanation fit ever more facts.  This might be seen as just a sociological analysis of the way science works. But Kuhn and others said that that is not correct!  Science is a social mechanism for what we accept as truth, but there is no way to say that we are approaching truth.  A paradigm can explain facts we care about today, but not things we're not even asking about.  Nor, since we don't actually know what the truth is (or if there is such a thing!), we can't say science is 'progressing' towards understanding it.  Science leads to better explanation or more accurate prediction or account of aspects we currently care about, but usually omits consideration of things not convenient or interesting--or issues that haven't yet arisen.

Many of our posts on MT are about the struggle to understand the world and the various sociological aspects about the claims, activities, and so on that characterize modern life and health sciences.   An important aspect of this is the notion of causation and whether it can ever be purely deterministic or, perhaps more perplexingly, can be truly probabilistic.  We use  probability and statistics routinely in science, for various reasons, but often implicitly explaining that this is because there are always measurement and other sources of error, or things too miniscule for us to identify perfectly, or that we have to collect incomplete samples and try to generalize from them to the whole in some way.

One truth?  Many?  None?
Still, do we believe that Nature has universal--truly universal, everywhere on earth and everywhere else--laws or truth that apply without exception?  Such truth would be deterministic--if you know a situation perfectly you can predict the future perfectly.  If so, we might expect that scientific methods will be truly be able to reveal that truth, at least asymptotically.

Or, if truth is that causation is probabilistic (as it seems to be in quantum mechanics as it's done today), there isn't one state of Nature but an array each with some probability of being observed.  Even there, we assume a fixed or deterministic distribution, or set, of those probabilities.  If that's a correct view, then science--if we do it right--should be able to asymptotically approach the true values of those probabilities.

Could it be that we're more fundamentally wrong?  Could Nature not have just one truth in a given area, nor a universal truth?  If that's so, how would we know?  How would we know how many truths there were, or when and where they applied?  Are they discrete--that is, does one truth stand alone relative to other truths about the same thing?  If so, where are they relative to each other: do truths overlap, or is one truth more dominant than another?   If these questions had specific answers, we might expect that science could asymptotically approach this set or distribution of truths.  It's weird, but theories such as multiverse or many-worlds cosmologies entertain such ideas and treat them quite seriously.  Can we, or should we, try to build such concepts into our daily scientific life?

Today,  scientists can excuse ourselves as still being in an infantile state of understanding Nature, and plead for further support by saying that complete truth has eluded us so far but, because it exists, it's important to continue to refine our asymptotic approach to it.  That is, the more work we do, the closer we get to complete understanding.

Pascal's answer was that in frustration we'll be driven ultimately to theological religion (Christianity, for him) as the only, or at least the safest, way to go. It that's what happens to science, it will be quite a surprise....even a paradigm shift!


Cathy Sander said...

I think that the universe is beyond all human concepts, so frankly I'm not entirely concerned about whether the Truth exists. Our knowledge is a human creation and cannot possibly be isomorphic to how the world is. What matters to us, it seems, is to have a sense of intimacy with the world, by having a capacity to "fall in love", so to speak, with the world beyond our own ideas. No wonder I feel so strongly about science as a spiritual discipline: for me, it allows us to practice humility and patience at what we study, tending with care and concern about what we would like to understand. "Truth" as a concept has a theological ring to it, as I see it. It may matter, but it's not the only driving force behind us doing research in the first place (I'm currently researching in THz metamaterials). I think that we do science for non-science reasons, after all. The initial sense of wonder is non-scientific in origin, but it drives us in this quest for understanding.

Ken Weiss said...

Too often we pursue a current belief about truth even when we know it's not accurate. So we fall in love, as you say, with what we already know is at least to some known extent fictional. Whether or not ultimate truth is discernible, or even exists as one might think it must, is a different order of question. In that sense, it may be viewed as a spiritual issue.

Cathy Sander said...

I agree on that point of ultimate truth as a spiritual issue. Fortunately, though, we don't live life that way most of the time. We're more attuned to what's immediately around us, which is good enough. However, are people aware of that reality that they live in? For the most part, I don't think so. So it's ultimately a trust/faith issue, which cannot be resolved scientifically. It's just what I said: "we do science for non-science reasons". Or in simple terms: we're humans who so happen to do research.