Sunday, November 01, 2009 ... /////

David Mermin on reality and abstractions

In May 2009, N. David Mermin wrote a provoking two-page essay in Physics Today:

Mermin is the author of the "Shut up and calculate" dictum about the philosophy of quantum mechanics that is sometimes attributed to Feynman although this attribution is not based on any actual quotes of Feynman. He is also the author of a great pedagogical example of Bell's inequalities - that was used by Mulder and Scully in The Fabric of the Cosmos by Brian Greene: it's also reviewed in my lecture on entanglement and interpretation of QM.

While you may see that he's quite a serious guy, the most important source of credibility of his article is obviously the headline, "reference frame". ;-) (No, I couldn't win a copyright lawsuit against Physics Today: they used the headline back in the 20th century but be sure that I didn't know about that until recently.) Unlike other blogs, I can give you a link to the whole article:

In the essay, Mermin criticizes the bad habit of the physicists to identify the reality with the abstractions.

He begins with two examples of bad habits:
1. belief that the wave function in quantum mechanics must be a real wave
2. idea that the reality "is" the set of quantum fields at spacetime points.
Mermin and interpretation of QM

Now, it's obvious that I completely agree with Mermin on his first point - the interpretation of quantum mechanics. He is one of the people who have explained the ontological character of the objects in quantum mechanics most beautifully.

Needless to say, he declares the objects, equations, and methods in quantum mechanics to be just tools designed to make predictions about the reality. Mermin concludes that the reality we care about may always be organized in terms of events in a classical space and time.

But the procedures to organize and relate these events to each other in order to predict the correct outcomes may be very complex - and they often require all the "unreal", mathematically rich structures of quantum mechanics - structures that inherently follow a very different logic but whose final results may be organized in terms of the ordinary logic of "classical events".

Mermin also admits that when he was older, he understood that Niels Bohr (whom Mermin liked to dismiss, when Mermin was younger) was very wise and modern when it came to these philosophical aspects of quantum mechanics.

Mermin and the structure of spacetime

In fact, I also agree with the specific content of Mermin's second example. His point was to emphasize that the quantum fields don't have to be the final word. They describe "almost" everything we know about the real world. But at the very end, quantum field theory is almost certainly not a theory of "quite" everything.

The theory of everything includes quantum gravity and quantum gravity is almost certainly not equivalent to a strict local quantum field theory - obeying the conventional definitions - in the bulk. String theory has taught us many insights indicating that things are not as simple. More generally, even outside the toolkit of string theory, we know that dynamics must be "slightly" non-local because the information may (and must) get out of black holes before it evaporates.

We know that dimensions of space may emerge, disappear, transform into new types of charges, change their topology, admit equivalent i.e. dual descriptions with completely different topologies of spacetime whose physical consequences are nevertheless indistinguishable. Mermin doesn't enumerate the huge diversity of new ways how to "dissolve" and "transform" the spacetime geometry. All of them cast doubts on the opinion that "everything is quantum fields".

But for his general purpose, it's enough that he reminds the readers that the relationships between the space and time were profoundly changed by relativity and that quantum gravity brings us, at least morally, some kind of "quantum foam" near the Planck scale which shouldn't be instantly dismissed.

Mermin and philosophical generalizations

I have talked about his two examples but his message was supposed to be more far-reaching. Whether Mermin's general philosophical point makes any sense depends on his definitions of "reality" and "abstraction".

Well, if his "abstraction" is meant to represent the current theoretical frameworks and the "reality" is the exact answer - the answer that may perhaps be fully described and understood by the perfect theory in the future, then he's obviously right. His statement simply says "be careful, our current theories and concepts may still undergo transformations in the future".

If we consider, for a while, the hypothetical case that we already possess the right theory of something - or everything - then the distinction between "reality" and "abstraction" may still be made but it doesn't teach us much. Clearly, atoms are not just the state vectors in the Hilbert space controlled by the right Hamiltonian (and they surely differ from the piece of paper with Schrödinger's equation written on it). But the states with the right dynamics do describe the reality - and they may describe everything about it. So from some point of view, you may also identify them (the description and the real objects).

Or if you're irritated by such an identification, you may insist that the reality is always something different than the description, even when you're know that the description is exact. If the description is correct, the reality is isomorphic to the description - so the distinction you have made is kind of physically redundant.

Fine, I wouldn't really argue about it. You're allowed to distinguish them.

But there is one more point I would like to stress - and it is a point where I would be slightly negative about Mermin's approach. One must always be careful about excessive philosophical generalizations. They may always be wrong. Yes, Mermin is saying nearly the same thing but he doesn't quite seem to realize that his thesis is an abstraction, too. ;-)

His thesis is that "abstractions must always be distinguished from reality". It is an abstract thesis that always leads you to question the current theoretical framework used to describe the world. While it's good to question things, one should always spend an appropriate amount of time with positive thinking as well as doubts. The right ratio is determined by the evidence.

And some statements about the world - i.e. some features of our description - can actually be quite accurate and quite universal. They will be a part of all improved descriptions of reality that will be found in the future, too. And in this sense, they are effectively a part of reality.

I am convinced that the Lorentz symmetry (or local Lorentz symmetry, to reduce neverending, unnecessarily confusing debates about the role played by the Lorentz invariance in GR) and especially the general postulates of quantum mechanics are among those universal features of the descriptions in modern physics. In some sense, they are parts of reality (or something isomorphic to the reality, if you wish) - and it's much better to "reify" them than to try to "trash" them all the time.

Why? Simply because they're almost certainly true and profound. So "reification" may sound as an insult and Mermin surely wanted this insult to be used as often as possible: Mermin wanted to spread this meta-abstraction or a meme. ;-) But in some contexts, the physicists' approach that some people could call "reification" is simply the right one because Nature may work in this way and She has no obligation to protect the warriors against reification in every single battle.

She is constrained by no philosophical dogmas that we have artificially invented. She is only constrained by the real dogmas, the laws of physics, and we must be careful, bright, educated, creative, and industrious if we want to find out which of these dogmas are real and which of them are just an excess baggage invented by our ancestors that should be jettisoned.

And that's the memo.

Hat tip: Sabine Hossenfelder (I don't think that her criticism makes any sense - she doesn't understand any physics, after all)