Friday, November 23, 2012

Oh say can you see, in \(\omega\)'s early light...

This may have been the line that convinced me to write the previous post, but I couldn't actually find a way to work it in:
Some materialists, however, seek to evade this difficulty by suggesting that there is some sort of logical connection between physical states and mental states. It is a logically necessary truth, they say, that when a given physical state occurs, a certain mental state also occurs. If this is true, then the existence of the mental is certainly probable, given our physical world; indeed, its existence is necessary. Nagel himself suggests that there are such necessary connections. So wouldn’t that be enough to make intelligible the occurrence of the mental in our physical world?
I suspect that his answer would be no. Perhaps the reason would be that we cannot just see these alleged necessities, in the way we can just see that 2+1=3.
Does Alvin Plantinga think that he can "just see" the truth of the arithmetic equality he states? If so, could he please enlighten the rest of us as to just what kind of sense-objects the numbers 1, 2, and 3 are? If a number is a class of all objects which are in bijection with one another, as Frege proposed, what is +? If a number is a hereditary set -- an "object" of pure imagination, how can Plantinga see or sense any such relationship?

It is disappointing when a scholar gives such free and open rein to his biases and preconceptions, without even a token nod at the methodological difficulties which others have made entire careers of.

Alvin Plantinga's students need to teach him about the argument from personal incredulity

An old friend of mine, a fellow of the cloth[1], shared this piece on his facebook wall the other day. He's of a decidedly conservative bent, and his share-comment was something along the lines of "even the liberal New Republic!"

Of course, it's irresistible troll-bait when even the liberal New Republic front-pages something entitled "Why Darwinist Materialism is Wrong"; but it feels less fun when I click over and see that the byline is Alvin Plantinga. Or that the piece is a four-page-long review of the new book Mind and Cosmos by Thomas Nagel that the usual suspects are a-twitter about. It feels less fun, because one of the secret thrills of reading pieces with titles asserting that Darwinist whateveritmaybe is wrong is the white-hatted hope that maybe the author will put forward some argument that I've never heard before -- and the black-hatted hope that maybe the author will commit some howler of a logical error and embarrass themself in a new and spectacular fashion, thereby torpedoing their own credibility for the future. (As Greta Christina says, maybe this time will be the time that a serious defender of religion makes an interesting case that I haven't heard before.) But one hardly dares hope for any new errors that someone like Plantinga will make -- he's quite set in his old ones, thank you very much.

TL/DR for what follows: Alvin Plantinga desperately needs to learn that personal incredulity is not a valid argument. He also needs to read some science fiction already. Start with Larry Niven. Then move to Iain Banks.

I'm not quite sure what the intellectual background is that has Plantinga frequently being brought up in reviews of MaC, even before this particular piece by Plantinga came out (maybe they had preprints?), but there is certainly a sense among the reviewers that Nagel's project is not so far off from Plantinga's.

If that sentence were to be written of me, I would immediately phone up at least one of these reviewers, and ask how on earth what I had written could be misconstrued in so ghastly a fashion. But I am of course not Thomas Nagel. I have, in fact, not read any of Nagel's books, including this latest -- I am not sure whether I read his influential article "What is it like to be a bat" sometime in undergrad, but I've read things arguing similar positions. In any case, this post is not to critique Nagel's work, but Plantinga's thoughts upon reading Nagel. (E.g., I will assume that Plantinga fairly represents Nagel's ideas throughout.) What Nagel seems to be arguing against is the materialist consensus of contemporary physics and metaphysics; but also, his book is "a far-reaching broadside against Darwin" according to the Weisberg/Leiter review linked above, which does not provide much hope for anything interesting to be said in it.

Weisberg and Leiter go on to say that it is not evolution itself that Nagel attacks, but rather a sort of reductionism -- reducing living things to their nonliving components, reducing minds to brains, reducing life to chemistry to nuclear physics to quantum physics to who-knows-what-further-smaller-scale-physics. This discussion is hardly new, of course, and hardier souls than myself have waded into Nagel's own arguments and found them wanting. But let's review, just in case, the old argument that "evolution isn't improbable, it's practically inevitable":
Premise 1: there is already life, instances of which reproduce and whose offspring are sometimes different from them (and from each other).
Premise 2: these differences can be relevant to the survival of these offspring, depending on the environments they find themselves in.
Conclusion: the aggregate phenotype of the next generation will be tilted towards those traits which favor survival in the current environment. If the environment changes, the advantageous traits will probably also change. If a population splits and the split pieces find themselves in different environments, it is likely that (given a bit of time) the two populations will end up radically different from each other.


Plantinga wants to bolster Nagel's attacks. He builds his essay around an explication and exegesis of Nagel's phrase "materialist naturalism", by which he intends the orthodox position in the academy from physicians to physicists to metaphysicists, that
there is no such person as God or any other supernatural being. Life on our planet arose by way of ill-understood but completely naturalistic processes involving only the working of natural law. Given life, natural selection has taken over, and produced all the enormous variety that we find in the living world. Human beings, like the rest of the world, are material objects through and through; they have no soul or ego or self of any immaterial sort. At bottom, what there is in our world are the elementary particles described in physics, together with things composed of these particles.
There is, of course, at least one critical undefined term in this definition: "natural". Plantinga does not explicitly define it; suppose for a first attempt we say that a process is natural if it follows a rule which is either deterministic or statistically deterministic[2]; neither such proposition can be experimentally verified, of course, but a deterministic rule can be absolutely falsified and a statistical rule can be shown by experiment to be very likely false (in a precise sense).

I suppose the above definition isn't perfectly satisfactory; for one thing, it introduces an undefined term of its own: "universe". Plantinga is a dualist: he believes that the material world is not all that is; indeed, not only is there an immaterial deity, but that life is different from non-life by virtue of the presence of something immaterial. (Yes, I know that it's not at all clear that the preceding sentence is meaningful.) It is a bit of a dilemma for the dualist-theist whether the immaterial stuff should be taken as part of the state of the universe; if no, then every action by a living being is supernatural, which is a bit of a profusion; if yes, then why should the immaterial stuff making up the deity be arbitrarily written off as outside the universe, when all the other immaterial stuff is also universe-stuff?

No matter. The exact definition of what is "natural", when in dialogue with Plantinga, must certainly exclude his deity, and must exclude whatever it is that produced life in the beginning, and that which produces and sustains consciousness:
So far Nagel seems to me to be right on target. The probability, with respect to our current evidence, that life has somehow come to be from non-life just by the working of the laws of physics and chemistry is vanishingly small. And given the existence of a primitive life form, the probability that all the current variety of life should have come to be by unguided evolution, while perhaps not quite as small, is nevertheless minuscule. These two conceptions of materialist naturalism are very likely false.
Plantinga here, as well as Nagel if his positions are represented fairly, is talking nonsense. He's out of his depth on the math, and making bald assertions about what is likely or not. While it's probably not going to be possible to historically reverse-engineer the exact chemical pathway which first self-replicated in the primordial environment, such chemicals are not really sparsely distributed among molecules once you reach a certain size; in other words, given an environment with a decent amount of energy and chemically disruptive events like occasional gamma showers or other cosmic rays, an array of elements to work with (a bit of hydgrogen, some larger multivalent elements like carbon or nitrogen or what have you) and a few eons of time, the possibility of self-replicating chemicals getting synthesized and then continuing to synthesize themselves isn't a priori unlikely at all.

(Side note to Hollywood: the likelihood of those chemicals being DNA/RNA, on the other hand, is vanishingly small. Aliens will not have DNA. Aliens will have their own replicating data storage/transmission molecules.)

This is familiar ground for anyone who has spent time arguing on the internet about evolution. We have observed speciation events taking place quite rapidly, geologically speaking; even the pace of our own forking from the chimpanzees, say three million years before present, would allow for twenty-two such forkings between the time of the dinosaurs (say 65 MYBP) and today, meaning \(2^{22} \approx 4 \text{ million} \) species descendents of a single species of that time; in reality, since humans and chimps have a relatively slow reproductive cycle, 65 million years represents a lot more generations of a typical species, meaning a lot more potential forkings.
NAGEL GOES ON: he thinks it is especially improbable that consciousness and reason should come to be if materialist naturalism is true. “Consciousness is the most conspicuous obstacle to a comprehensive naturalism that relies only on the resources of physical science.” Why so? Nagel’s point seems to be that the physical sciences—physics, chemistry, biology, neurology—cannot explain or account for the fact that we human beings and presumably some other animals are conscious...
Nagel next turns his attention to belief and cognition: “the problem that I want to take up now concerns mental functions such as thought, reasoning, and evaluation that are limited to humans, though their beginnings may be found in a few other species.” We human beings and perhaps some other animals are not merely conscious, we also hold beliefs, many of which are in fact true. It is one thing to feel pain; it is quite another to believe, say, that pain can be a useful signal of dysfunction. According to Nagel, materialist naturalism has great difficulty with consciousness, but it has even greater difficulty with cognition. He thinks it monumentally unlikely that unguided natural selection should have “generated creatures with the capacity to discover by reason the truth about a reality that extends vastly beyond the initial appearances.” He is thinking in particular of science itself.
Hi, Professors Nagel and Plantinga, I'd like to introduce you to my friends, Professors Dennett, Hofstadter, and Gödel. Specifically, let's think for a moment about the tremendous leap that just happened here. I notice, for example, that Plantinga does not find it at all implausible that, given the existence of complex multicellular life, that such life would evolve or develop some kind of system for coordinating sensory inputs with actions -- a kind of "central nervous system", if you will. And any such system will need routines for filtering the raw datastream provided by the senses down to usable chunks, which might go by a name like "concepts". "Consciousness", then is nothing other than the biological individual having a concept for itself -- an almost inevitable outgrowth (given enough time and computational substrate) of the process of turning a patch of brown in the visual stream into a tree, that is, a unified concept allowing details judged to be irrelevant to be ignored.

Let's think about Nagel's original, famous question from back in the day: what is it like to be a bat? Specifically, Nagel posed the following question: suppose we knew the entire physical state of the brain (or even the whole organism, for that matter) of a bat. Nagel contends, and I agree, that even with all that information, there is no way that a human could have the experience of being a bat. (A human could presumably have all the sensory experiences of a bat, but would, I think, not experience them as a bat would.) However, this is not the end of the story. Suppose, say, we could construct or virtualize a computer network of circuits or connections perfectly isomorphic to the network of neurons in a bat's brain; and that we keep the computer in our office while we let the bat fly, but with a device attached to the bat's neural system which instructs every circuit in the computer to open or close precisely as the bat's does. Then, I contend, that computer will experience what is like to be a bat. Given what we know about bats, it is likely that neither the bat nor the computer will have any experience of "self"; but they will have concepts organizing their sensory experiences and experience of responding to sensory data. These experiences of the bat, and likewise those of a person, up to and including my experience of my own self, are high-level emergent concepts produced by "chunking" raw data into concepts and concepts into higher-level concepts.

The preceding paragaph is a just-so story, not a robust scientific theory. However, it shows how false it is to claim that a scientific paradigm that is even methodologically materialistic (as opposed to the metaphysical materialism that Plantinga so opposes) has no account of selfhood, of consciousness, and of cognition. And it exposes as rank nonsense Plantinga's next claim:
Natural selection is interested in behavior, not in the truth of belief, except as that latter is related to behavior. So concede for the moment that natural selection might perhaps be expected to produce creatures with cognitive faculties that are reliable when it comes to beliefs about the physical environment: beliefs, for example, about the presence of predators, or food, or potential mates. But what about beliefs that go far beyond anything with survival value? What about physics, or neurology, or molecular biology, or evolutionary theory? What is the probability, given materialist naturalism, that our cognitive faculties should be reliable in such areas? It is very small indeed. It follows—in a wonderful irony—that a materialistic naturalist should be skeptical about science, or at any rate about those parts of it far removed from everyday life.
Do you see the straw man? The argument of the naturalist is that if it is possible for a species to evolve the ability to apprehend or generate concepts, and if those concepts correspond in some reliable way to the world, then it is likely that such ability will be evolved. The argument is not that brains will evolve the ability to only conceptualize true concepts, or concepts corresponding to real things in the world. And indeed, this is the dangerous thing about evolution (and the reason why Michael Behe's argument by "irreducible complexity" fails so miserably): an ability or trait, once evolved, will turn out to be useful in completely different circumstances than those which provided the original environmental pressures it evolved in response to. A mind which can make concepts of trees -- and which can treat "concept" as a concept -- can create concepts of basically anything, regardless of whether those concepts correspond to anything in the world, and regardless of whether creating those concepts has survival value.

And it's not saying much to say that, if at the time the human brain evolved the ability to nest concepts to arbitrary depth and abstraction, every such brain decided that it would be more fun to spend all its time concept-building and philosophizing rather than ensuring basic survival, then all those brains would indeed have perished. But since that just-so-story empirically did not happen, it is not advisable to pretend that such is the only possible outcome of brains hitting on the neat trick of arbitrary-depth conceptualization.

Instead, Professor Plantinga, please go read a little bit of recursion theory. Hell, you could even play some Magic, The Gathering and be better off. Your incredulity is not interesting.


[1] Does that phrase imply that the laity are all naked?

[2] A process is deterministic if the outcome of the process is a function of the state of the universe at the time of the process. Typically the state of the whole universe is not needed, just the state of the billiard table on which the process occurs. A process is statistically deterministic (my phrase) if there is a function which takes as input the state of the universe, and outputs a distribution, and the outcome of the process is distributed according to that distribution.

Monday, November 19, 2012

Vacation is...

Lying in bed cleaning out one's RSS reader and occasionally checking to see if one's favored football team is still the first in the land.