Disagreeing with Roger Penrose


… which, of course, takes a large amount of chutzpah, and a much larger amount of over-self-confidence, given that leading-edge physics is really not my field at all. (Penrose on Wikipedia.) Which of course will not keep me from doing it anyway.

So I recently chewed my way through his classic “The Emperor’s New Mind“, which is an excellent 450-page overview over not just physics but also some computer science (the science aspect of it), some neuroscience and much else besides. It’s supposedly written for the lay person — the kind of lay person who has no trouble imagining infinite-dimensional complex-numbered vector spaces. Which limits the market for his ideas just a tad :-) Of course, he left out all the “real” math, simply abbreviating entire fields of mathematical physics with a single letter, which was a wise choice. In my case, I remembered just enough of good old German EE-style linear algebra plus vector calculus to sort of totter along with him on the simplified level.

What he’s really interested in is an explanation of consciousness, from the perspective of hard science and particular quantum mechanics, one of his specialties. He brilliantly describes the entire field of quantum mechanics as U plus R, where:

  • U: the set of quantum mechanical processes by which complex amplitude fields evolve
  • R: the processes (“observation by a conscious mind”), by which complex amplitude fields collapse into observed probabilities.

I had far lesser minds as teachers back in the days, who seemed to rather confused about all of that themselves. He correctly points out that U — which is most of what happens — is totally predictable, and there are no probabilities involved at all. Only when somebody starts “observing”, the wave functions collapse and we get something that has probabilities attached to it.

Like everybody ever since quantum mechanics emerged, he’s puzzled by R, and why in the world it should have anything to do with a conscious mind observing. He’s going out on a limb in the last part of this book, and proposes an explanation, which is a great idea: much better to have a strawman that can be tested, explored, changed, torn etc., that simply be puzzled.

Well, I really dislike this strawman. Not that it matters, but no physicist reads this blog anyway, so why not jot down my dislike. In essence, he says “whenever things get too big, they start collapsing”. So R kicks in as soon as we look at macroscopic objects i.e. objects that are bigger than some threshold, and he has some ideas how big that threshold should be. In his own classification of SUPERB vs lesser categories of theories, I hardly can see it be in the running. He gives no good reasoning why that should be so, and does not address the obvious objections against the strawman either (well, perhaps he does that elsewhere, but not in this book). Like “doesn’t that seem to fly in the face of demonstrated entanglement across long distances?” Or “the screen in the two-slit experiment is obviously macroscopic, and it is clearly part of the experiment, why doesn’t it immediately trigger the threshold?”

But I really dislike it because I thought throughout the book he was building up towards something else, which would have been much more intriguing, which I’d put as some version of this:

“We know that the quantum mechanical equations must not be 100% correct because they are incompatible with general relativity. So let’s postulate they are as correct as Newton’s equations compared to Einstein’s equations: Newton’s equations are perfect for all situations which we had investigated so far (that was certainly true for Newton in his day), and only increasingly inaccurate for situations that we hadn’t looked at (like experiments close to the speed of light). For quantum mechanics, we don’t know yet what the equivalent of “close to the speed of light” is — the set of circumstances in which they become increasingly inaccurate. (And there might be more than one.) We know for sure they don’t explain R.

“What about if the human brain somehow had evolved in a way that it created those special circumstances in which the U is really increasingly inaccurate, and has morphed into R? If DNA and self-replicating organic machinery can evolve, why not that?” (We could imagine some special constellation of particles that the brain sets up in which, say, resonance occurs that does not occur anywhere else, which creates those special conditions. Or something like that.)

This would mesh with his idea that the brain depends on quantum effects to do its magic, and also resolve why there is an R in addition to U and it only happens with “conscious minds”. It would also explain his conjecture — which I’m not convinced of, by the way — that the human mind can resolve questions that aren’t computable in the computer science sense.

Personally, my gut feel is that this (revised) strawman is almost certainly false. But then, I have no reasoning to offer why that should be so. So ignore my gut feel. But if it were true, it would be a SUPERBly intriguing idea. I’ll credit Penrose, because I thought through most of the book, that is what he was going to spring on us towards the end, even if he didn’t.

I wonder whether anybody is investigating anything along these lines. That’d be a cool project.