14 May 2015

The Role Of Philosophy In Physics


One of the evergreen topics in online discussions of physics is what role, if any, philosophy plays in the discipline. I’ve been blogging since 2002 (which is after the dinosaurs, but before the giant armored sloths), and some variant of the philosophy argument comes around every year or so. Sometimes several variants at once, if somebody prominent has a book to sell.

The most recent outburst has been better than most, with this post by Tim Maudlin at PBS, and this Scientific American essay by Victor Stenger, James Lindsay, and Peter Boghossian. The Stenger et al. piece includes a historical summary of a lot of philosophical thinking, and Maudlin nicely articulates the key advantages of a philosophical approach:

Philosophers strive for conceptual clarity. Their training instills certain habits of thought—sensitivity to ambiguity, precision of expression, attention to theoretical detail—that are essential for understanding what a mathematical formalism might suggest about the actual world. Philosophers also learn to spot the gaps and elisions in everyday arguments. These gaps provide entry points for conceptual wedges: nooks where overlooked alternatives can take root and grow. The “shut up and calculate” ethos does not promote this critical attitude toward arguments; philosophy does.

(Of course, you might quite reasonably argue that mathematical rigor is at least as strong a standard as philosophical rigor, with the same emphasis on clearly stating assumptions, and so on. I’ve certainly been struck by that attention to detail when I’ve taught the physics portion of our team-taught Integrated Math and Physics sequence that combines the first two physics courses with the intro calculus sequence.)

As I said, the current round is better than most, because it avoids the lazy fall-back position of most of these arguments, which is pointing at the philosophical inclinations of Niels Bohr, possibly even quoting Werner Heisenberg’s description of him as “primarily a philosopher, not a physicist” as if that were something more positive than a backhanded compliment. As I’ve written before, Bohr is a pretty bad example of philosophy in physics, as he was maddeningly vague and a horribly unclear writer.

Maudlin does cite Einstein, another of the standard fall-back arguments, and Einstein’s early career is a good example of a philosophical approach paying off. Not so much because of the magnet stuff that Maudlin mentions, but more in his approach to space and time. Einstein’s 1905 paper introducing special relativity is a great example of reforming physics by challenging the philosophical assumptions. Einstein points out, correctly, that a proper treatment of moving objects requires a clear definition of what it means to talk about separated events happening at the same time. Any sensible definition of simultaneity requires a clear statement of the procedure by which the timing of events will be determined, and Einstein correctly points out that any such scheme necessarily leads to relativity when the speed of light is a constant.

But Einstein is an interesting case, because his later career points out a weakness of the philosophical approach to physics. While Einstein played an essential role in kicking off quantum QTM -3.88% mechanics, he later rejected the theory because he had philosophical issues with it. Unlike relativity, where his approach was directly and cleanly translated to mathematics, though, his anti-quantum arguments were mostly as fuzzy as Bohr’s pro-quantum ones. Einstein’s reaction to quantum theory was mostly an argument of dissatisfaction– quantum predictions offended a hard-to-define intuition about how the world “ought” to work, but he never managed a solid alternative.

There’s a bit of irony to this, of course, because quantum pioneers like Bohr and Heisenberg felt (with some justification) that they were following the same track Einstein did with his work on relativity. Their insistence that one could not talk sensibly about a reality beyond the results of measurements is, in some sense, an extension of Einstein’s observation that speaking of simultaneous events makes no sense without first explaining how one would go about determining simultaneity. (I’m not entirely sure how to sort these two approaches into the “positivist” and “instrumentalist” camps of the Stenger et al. article.) Heisenberg and Bohr never quite understood Einstein’s recoil from their position, and to be honest, it’s not entirely clear that he had a well-thought-out philosophical justification for that.

Einstein’s most important philosophical contribution to the debate over quantum mechanics was, in the end, an assumption that was not stated clearly enough. This was the famous Einstein, Podolsky, and Rosen paper of 1935 (the “EPR” paper), which is the clearest statement Einstein managed of his problems with the idea we now call “entanglement.” Einstein, Podolsky, and Rosen point out that it’s possible to prepare a pair of particles in a joint state, and then separate them by a large distance before measuring their properties. Quantum mechanics says that the state of these particles is indeterminate until they are measured, but also predicts that the measurements will be correlated in a way that allows the person measuring the state of particle A to predict with certainty the result of a measurement of particle B, an arbitrary distance away. Einstein, Podolsky, and Rosen argued that this points to a fundamental incompleteness in quantum theory– that there must be some deeper reality in which those measurement outcomes are defined in advance.

This was a bombshell in the philosophical debate Einstein had by then been carrying on for nearly a decade with Bohr and others. Bohr’s response was characteristically muddled, but thirty years later the situation was given clarity by the Irish physicist John Stewart Bell, who worked out the concrete implication of the hidden assumption in the EPR paper, namely that measurements of particle A are complete independent of those of particle B, by virtue of the distance between then and the finite speed of light. Bell proved mathematically that this requirement places very tight constraints on the sort of correlations you can observe between measurements of A and B, and that these differ from the predictions of quantum mechanics in a concrete and testable way. And starting with John Clauser and colleagues in the mid-70′s, then Alain Aspect in in the early 80′s, physicists have done these tests and shown that Einstein was wrong. Here’s a brief cartoon I wrote for TED-Ed explaining the idea:

No comments: