Monday, January 20, 2020

Carroll rejects panpsychism

Jerry Coyne agrees with Sean M. Carroll about panpsychism failing to explain consciousness:
Goff himself, in the podcast below, repeatedly states that he’s “heartened” by the increasing (but still minority) view among philosophers that panpsychism is the way to go in explaining consciousness. And others, like physicist Lee Smolin, authors Annaka Harris and Philip Pullman, and philosopher Stephen Law, have endorsed Goff’s new trade book, though this doesn’t mean they all endorse panpsychism. ...

Sean, of course, is a physicist, cosmologist, and author, who knows a lot about philosophy. Debating him is Philip Goff, a philosopher at Durham University and perhaps the most vociferous advocate of panpsychism (he has a new book about it). Sean states from the outset that he doesn’t accept panpsychism, and that materialism (his view of the world) is perfectly capable of explaining consciousness, though it’s a hard problem and will take a long time to understand. ...

A brief view of the controversy. Goff avers that materialism won’t help us understand consciousness because all it produces are correlations between brain activity and conscious experience. That, he says, is useless because it doesn’t enable us to get at the heart of consciousness: subjective experience or “qualia”. As he says, “How can you capture in an equation the spiciness of paprika?” ...

In response to Goff’s statement that we need to know what consciousness really is,  Carroll answers he doesn’t really care about the “intrinsic nature of subjective experience”. If you have consciousness and know how it’s produced from neurons and the brain, that’s all there is to know.

As I said, the heart of the disagreement starts about 71 minutes in, when Goff argues that yes, everything is conscious: even the mass, spin, and charge of physical particles like electrons are forms of consciousness: a “limited form of conscious experience.” But that’s about as far as he goes in defining consciousness of inanimate objects. When Carroll asks him if he means that everything is conscious, because everything has physical properties, and whether the the Universe’s wave function is also conscious (Sean talks about that wave function his latest book Something Deeply Hidden), Goff gives a reluctant “yes”. Goff also declares that panpsychism is completely congruent with what physics tell us about the Universe. 
Before the discovery of DNA, people wondered whether life could be reduced to chemistry. We do not know whether consciousness can be similarly reduced.

SciAm also has an article on panpsychism.

It is funny to see these guys argue about who is being more materialist.

Coyne and Carroll do not even believe in free will, so they must have a very limited view of consciousness. What is consciousness, if not the ability to come to your own conclusions about yourself and the world? If they don't believe in that, they I say they don't believe in consciousness.

I am not sure it makes sense to say that electrons have a tiny bit of consciousness, but I am not sure it is any crazier than someone thinking a century ago that a single DNA molecule could contain the mystery of life, in coded form.

I hate to see mainstream science popularizers promoting their personal speculations as if they were scientific facts.

Jerry Coyne argues:
Like Massimo, I’m perplexed, because when you ask Goff to tell us in what sense electrons are “conscious”, he just redefines their properties — spin, mass, charge, and so on — as consciousness. But if you pull that trick, then explaining human consciousness just becomes a purely physical problem, given Goff’s addendum that when you bundle enough conscious atoms and molecules and neurons together, you get a human brain. In other words, why isn’t consciousness then an epiphenomenon of the collection of molecules that make up the brain?

Further, Goff seems to think there is some “intrinsic nature” of matter that isn’t given by its behavior and observable properties. But to a physicist, the described properties of an electron completely characterize an electron for any purpose that we want. And if you call those properties “consciousness” and say that when there are enough conscious particles in a lump you get “higher” humanlike consciousness, then you’re saying nothing beyond describing neuroscientists are already trying to do. There are no essences beyond what we can observe. Or, if there are, Goff can’t tell us what they are, though he strains mightily to do so.
No, the electron is not completely characterized by mass, charge, and spin.

To see why, just read this new SciAm article on double-slit and triple-slit experiments. As it explains, photons and electrons give interference patterns that cannot be understood from just the quantum numbers of the particles. Indeed, the patterns disappear if you put detectors in the slits.

Quantum mechanics uses wave functions to predict these interference patterns, and the wave function have info that goes far beyond the quantum numbers. Because the theory is so successful, many are tempted to say that the wave function completely characterizes the electron. But we don't know that. The wave function is not observable, and we are almost never sure that we even have the right wave function.

Textbooks sometimes say that the electron is completely characterized by mass, charge, and spin, because there is a quantum theory of identical particles. All electrons are identical in the sense that if you have a lot of them in a system, there is a symmetry that applies to them. The symmetry has experimental consequences, so we are very confident about it. I am not disputing that.

But the wave function encodes other info, such as position, momentum, and entanglement with other particles. The wave function successfully predicts experiments, but physicists commonly complain that it is not telling us the whole story.

So based on current physics, it is possible that an electron has some infinitesimal consciousness.

A philosopher law professor suggests that panpsychism is the most preposterous of currently fashionable philosophical views. He links to a poll comparing it to other supposedly-preposterous views, like free will. I am sure that philosophers have much more preposterous views than these!

Update: Here are the poll results:
Poll description

Rank order the philosophical views, below, from the MOST to the LEAST preposterous.
1. External world skepticism
2. Realism about possible worlds
3. Panpsychism
4. Libertarianism about free will
5. Grounding is a real and unitary relation
6. Non-naturalist moral realism

1 comment:

  1. Dear Roger,

    Interesting write-up.

    1. True, a conscious living being is an integration of the soul/mind/consciousness and the body. (Here, for simplicity, let's take consciousness, mind and soul synonymously.)

    But can we therefore say that everything physical (say a fundamental particle) also must come with some consciousness? Any one who adopts this position quickly runs into difficulty when you ask: What is it that differentiates a corpse from a living being? The most complex parts in the body, for instance the neuronal circuits in the brain, do continue to retain most (if not all) of their complexity (configurational, morphological, local chemical profiles, etc). But why does the same "chunk" of material object now cease to be conscious?

    I am not outright denying the idea that the physical plenum might possibly also act as a plenum for consciousness. My point is, you would still not be able to ascribe a distinguished instance of consciousness to each physically distinguished particle. You would have to posit how is it that a complex enough a structure/configuration begins to experience the consciousness. Most important, the two attributes of the physical and the conscious, even if pertaining to the same plenum, would still have to be regarded as two distinct, separate, attributes. You couldn't use one as being explanatory of the other. Not at least with the level of total knowledge that we have currently.

    2. As to the >2-slit experiments, I don't know, but do anticipate, that it all is a bit of an oversell. I think that the main claim here, while original, still rather relies on physicists not having enough of an intuition when it comes to the field-theoretical simulations. Such a state of affairs, in turn, exists because physicists work with too abstract terms (infinite-dimensional Hilbert space, Dirac formalism, etc.) when it comes to QM. That's why the results appear so unexpected. They probe not so much on a new fundamental aspect of reality as how physicists hold their understanding.

    But ask any computational engineer how he thinks through the following simple question:

    Consider a perfectly isolated system, a box, for the interference chamber. A perfectly impermeable wall in the middle divides the chamber into two halves. Slits can be cut on this middle wall.

    Consider five different configurations, one each corresponding to: no slit, single slit, 2 slits, 3 slits, n > 3 slits (with n to be specified).

    Assume that both the halves are filled with a perfectly still and colorless continuum. Assume that a blob of ink is then introduced in the middle of of one of the two halves, but without disturbing the fluid in any way. Assume that the ink diffuses through the continuum.

    The problem is to predict the diffusion pattern after the lapse of a sufficiently long but finite amount of time.

    (Over infinite time, there will be uniform distribution of the diffusing species in all the five cases; in the first case (no slit), the ink would be confined to the half where it was introduced.)

    It would be very obvious to the computational engineer that the solution for the 2-slit case isn't going to be a superposition of two 1-slit solutions.

    Once this fact is pointed out, even the physicist would agree. (If he doesn't show enough interest right away, mention harmonic analysis, uniqueness theorem, etc., to get him interested. Go ten times as abstract and mathematical than what I just mentioned. Then he will get interested. After NSF grants for 10 years, he would come to agree.)

    A somewhat similar reasoning applies also for the QM wavefunction. It's just that physicists don't have already well-developed intuitions for such things.

    And, if the Sorkin parameter captures something beyond these simple considerations, I would like to know about it. The research would become interesting only in such a case. I think it does not.

    PS: It might be a good idea to move these items to my blog later on.