Quantum mechanics is strange, but not that strange

There’s a piece in Discover describing an experimental result that, according to the article, “shows that measurements performed in the future can influence the present.”  This sort of thing should always set off your extreme-skepticism response.  Sure enough, although the experimental result is pretty cool, it absolutely in no way implies backward-in-time causation.  Even by the standards of other silly woo-woo articles on quantum mechanics, this one is pretty bad.

If you read the article carefully, or better yet if you read this more technical and more accurate description, you’ll find that what’s actually been shown is that later measurement results are correlated with earlier ones.  To be specific, suppose you prepare a bunch of particles at one time,  then make a measurement on them at an intermediate time, then finally make another measurement at  a later time.  You then retroactively split the particles into two groups based on the result of that later measurement.  What’s been discovered is that those two groups yield different outcomes on the earlier measurements.

That’s not at all surprising, and it in no way implies backwards-in-time causation.  Suppose we both know that Professor Bertlmann‘s socks never match.  Suppose I look at one of his socks now, and you look at his other sock an hour from now.  If you “measure” that his sock is green, then you can conclude that my measurement must have yielded non-green.  Do you think that your measurement somehow sent influences back in time, causing my measurement to be non-green?  If so, then you should think the same thing about these quantum mechanics results.  If not, not.

As far as I can tell (this isn’t my field), the precise nature of the correlations that result in this experiment are interesting and potentially quite useful in allowing more accurate measurements of various things that people are interested in.  So it’s a good experiment.  It just doesn’t have the amazing philosophical implications imputed to it.