Thursday, November 3, 2011

Faster than a speeding ..... neutrino? What d'ya know?

Comfort Facts--the scientific analog of comfort food, are principles we think we really, truly know are at the heart of how the universe runs.  Even scientists (Professors, no less!) may not know everything, or as much as they (we?) want you to think we know.  But if we at least know some of the deepest fundamentals, if Nature is orderly then sooner or later we can build on the bedrock facts to get an ever-closer understanding of How Things Work.  But what if even the bedrock is shaky?  What can a scientist, much less an ordinary poor mortal, think?

Among the more secure Comfort Facts is that nothing travels faster than the speed of light.  The ultimate e=mc^2, where c is the speed of light.  But what if it's Eeecc! not true?  Where is the ground we stand on?  The issues are technical, largely beyond anything we know specifically, and there are various contexts in which one can read that c is not the Ultimate Limit.  To suggest that this Law of Laws can be broken borders on, well, heresy.  But a recent report claims that tiny particles called neutrinos in fact, do break the Law of Laws.

Can neutrinos travel faster than the speed of light or can't they?  Last month's report of just this observation has riled up the physics community, with most physicists saying it can't be true (it is, after all, against the rules!).  No one yet is sure how the researchers erred, or even that they did, but the general idea seems to be that Einstein's law has been tested so many times it can't be false now.  As reported last week in the NYT,
The neutrino news came from a group of physicists based at the Gran Sasso underground laboratory in Italy and doing business under the apt acronym Opera. The neutrinos, they reported on Sept. 23 in a paper and at a special symposium at CERN, the European Center for Nuclear Research, had beaten a metaphorical light beam from CERN to Gran Sasso, a distance of 457 miles, by 60 nanoseconds.
The initial response of physicists assembled at CERN and around the world was that there was probably a mistake somewhere in the experiment. Einstein’s theory is the basis of all modern physics, and has been tested a zillion times.
Technically, relativity does allow some particles, known as tachyons, to go faster than light — in fact it forbids them to slow down to light speed. The hitch is that they would have imaginary masses, whatever that means. And there is also the possibility, in some versions of string theory, of particles’ taking a shortcut through another dimension. But allowing anything to travel faster than light would open up the possibility of all kinds of problems with cause and effect and even time travel.
We aren't physicists and certainly can't comment on what might have gone wrong with the experiment, or indeed whether something necessarily did go wrong.  It would be pretty exciting, we have to say, if it were true, if something totally paradigm-toppling were to happen in our lifetimes.  But, be that as it may, this whole brouhaha touches on questions that philosophers of science have been grappling with for decades, if not centuries, and that we can comment on. 

Black swan, Wikimedia Commons
Now a fact can be replicable, such as estimates of c, without being universally true.  A new observation can always come along, such as the neutrino experiment that is different from or more sensitive than, or tests a previously untested condition than has been tested before.  The argument by induction, that repeated observation increases one's belief in something, can obviously be false.  The classic philosophy of science classroom example of the problem is the assertion "All swans are white," based on many observations.  But next along the river, one may see a black swan.

One might argue that this simply overturns an approximate, but faulty hypothesis about the true nature of swans.  But one may counter to defend orthodoxy, that the definition of 'swan', or perhaps some conditions of swandom, were not proper, and the idea is still true under a revised definition.  Others would say this is the critical kind of falsiification result that helps us tinker with our scientific hypotheses til we get to the right one.

In that view, a theory in any science is always, and only, a best-guess-at-the-time about Nature.  We'll never get to the absolute truth or even if we did, how would we know we can't come upon something better later?

The 1600s and 1700s ushered in the modern age of science, framed largely in terms of the idea that there are true, unexceptioned, universal Laws of Nature, and that science's job is to find out what those laws are.  Gravitation, relativity, properties of electricity and chemistry and so on, are examples from the physical sciences.  Darwin thought natural selection was such a law--indeed he repeatedly referred to patterns he observed or theorized, as 'laws'; so did his contemporaries in life and even social sciences.

But the history of all sciences is one of revision, so a reason, or an excuse, for accepting imperfect descriptions of Nature is often that we'll never know Truth and should accept that, and that our models today are something we should accept.  It is often now said, in such a context, that nobody believes in Laws of Nature any more.  But is that true?

We think not.  We think that whether people state it or not, most people in and out of science essentially if tacitly believe that there is one Truth and only one.  Whether we call it a 'law' or by some other name, it must be that way.  If Nature is not law-like in that way, or if in some part of the universe things are fundamentally different from what they are here, we would have no way to understand the essence of the universe--not even in principle!  But at the same time, if there is no Truth, how can we evaluate scientific evidence, make predictions, and the like?  Saying there are no material universals verges on mysticism, the ultimate NotSo for science.

If there is ultimate truth, we may never find it, being but fallible mortals evolved to detect what our ancestors needed to detect about Nature.  But the idea that there is such a Truth is at least an anchor for our worldview and certainly of our work in science.  We essentially believe that our approximations are approximations of something real--indeed that is what the word 'approximation' means.

Depending on one's view of what we know, or what there is to know, the speedy neutrino may turn out to fly past everything we thought we knew. But if it turns out that the measurements were faulty, and the little guy's just poking along at below-c speed, we cannot assume we had it right after all.  There will always come along another claimant to audition for the role of the black swan.

-Ken and Anne


John R. Vokey said...

Your beautiful image of a black swan notwithstanding, I always thought that the more interesting problem of strict induction was revealed by Hempel's paradox: non-black non-swan things count as much inductively in favour of the hypothesis as does the next white swan. Brown cows, purple flowers, red wine all logically are equally a confirmation of the all swans are white claim as is the next white swan. As they don't feel as confimatory is, of course, the paradox. The resolutions (there are many) of the paradox all point to the different ways we interpret both claims and the supporting evidence.

But, in this case, surely, no one feels that yet, another slow-moving non-neutrino adds anything to the debate. But, as Hempel notes, why not? I think abduction may play a role here, but now I am really outside my comfort zone.

Ken Weiss said...

These are good points, and they come up in philosophy classes but rarely in science; at least, in genetics and evolution classes, students often simply doze off when questions are raised. 'Who cares?,' is often the attitude. 'What's that got to do with biology?'

Most are purely empirical: we'll do our experiment, do our statistical test and get a p-value, and be happy.

To me, the important question is how we build in principles that we take, often without stating or even perhaps being aware, as true by assumption, and evaluate data relative to the assumption being held as given.

John R. Vokey said...

That should have read "...non-WHITE non-swan things count as much inductively in favour of the hypothesis as does the next white swan."

I raised abduction (reasoning to the best explanation a la J. S. Peirce) as it seems to capture that non-mechanistic reasoning that seems to undergird much of how we actually do science, as I think was Ken's point.

Ken Weiss said...

I'll have to look at Pierce, whose work I don't know.