Saturday, October 27, 2012
Nagel and his critics, Part II
Whereas my First Things review of Thomas Nagel’s Mind and Cosmos accentuated the positive, the first post in this series put forward some criticisms of the book. Let’s turn now to the objections against Nagel raised by Brian Leiter and Michael Weisberg in their review in The Nation.
First some stage setting is in order. As I indicated in the previous post, Mind and Cosmos is mostly devoted to the positive task of spelling out what a non-materialist version of naturalism might look like. The negative task of criticizing materialist forms of naturalism is carried out in only a relatively brief and sketchy way, and here Nagel is essentially relying on arguments he and others have developed at greater length elsewhere. Especially relevant for present purposes is a line of argument Nagel put forward in what is perhaps his most famous piece of writing -- his widely reprinted 1974 article “What Is It Like to Be a Bat?” -- and developed further in later works like The View From Nowhere.
Nagel on appearance and reality
Nagel’s argument in the article in question is often discussed in conjunction with Frank Jackson’s “knowledge argument” against materialist theories of the mind. Naturally, attention also often focuses on the ideas referred to in the article’s title -- on Nagel’s theme that there is “something it is like” to be conscious, and on his example of the bat as a creature whose conscious states are radically unlike our own. But the heart of Nagel’s argument really goes much deeper than any of this, to the nature of scientific method itself as that has been understood since Galileo and Descartes.
Take a stock example of reductive scientific explanation like the reduction of sound to compression waves, color to surface reflectance properties, or heat and cold to molecular motion. The way these explanations work is by treating the appearance that sound, color, heat and cold present to us in conscious experience as mere appearance, as a projection of the mind that corresponds to nothing in objective, mind-independent reality. What common sense understands by color, sound, heat and cold -- the way red looks, the way a musical note sounds, the way a hot stove feels, and so forth -- is held to have no objective reality, any more than the redness a person unknowingly wearing red-tinted contact lenses thinks he sees in all the objects around him really exists in those objects. Instead, color is for scientific purposes essentially redefined by the method in terms of the surface reflectance properties that cause in us the subjective appearance of color; sound redefined in terms of the compression waves that cause in us the subjective appearance of sound; and heat and cold redefined in terms of the molecular motion that causes in us the subjective appearance of heat and cold.
Thus, as common sense understands color, sound, heat and cold, etc., the reductive method ends up treating the world as essentially colorless, soundless, devoid of temperature, etc. What the method calls “color,” “sound,” “heat” and “cold” is in fact something different from what the man on the street thinks of when he hears these terms. The “red” that the method says exists in the material world is just the tendency of an object to absorb certain wavelengths of light and to reflect others. The “red” that the man on the street thinks exists in the object does not really exist in the object itself at all but only in his perceptual experience of the object. The “heat” that the method says really exists in the material world is just the motion of molecules. The “heat” that the man on the street thinks exists in the object does not really exist in the object at all but only in his perceptual experience of the object. And so forth.
Now, Nagel’s point is not that there is something wrong per se with overthrowing common sense in this way. It is rather that whatever value this method has, it cannot coherently be applied to the explanation of conscious experience itself. If the reductive method involves ignoring the appearances of a thing and redefining the thing in terms of something other than the appearances, then since our conscious experience of the world just is the way the world appears to us, to ignore the appearances is in this case just to ignore the very phenomenon to be explained rather than to explain it. Consciousness is for this reason necessarily and uniquely resistant to explanation via the same method scientific reductionism applies to everything else. For the application of the method in this case, writes Nagel, “does not take us nearer to the real nature of the phenomenon: it takes us farther away from it.” To treat the appearances as essentially “subjective” or mind-dependent is precisely to make them incapable of explanation in entirely “objective” or mind-independent terms.
Hence “in a sense,” Nagel continues, “the seeds of this objection to the reducibility of experience are already detectable in successful cases of reduction.” As I have put it myself in several places, the reductive method in question is like the method of getting rid of all the dirt in the house by sweeping it under a certain rug. While this is a very effective way of getting rid of the dirt everywhere else, it is not a strategy that could possibly be used to get rid of the dirt under the rug itself. On the contrary, it only makes the problem of getting rid of that dirt even worse. And that is exactly why the mind-body problem as it is understood today essentially came into existence with Galileo, Descartes, and Co. and has remained unsolved to the present day. What these early modern thinkers wanted (for certain practical and political ends) was a completely quantitative, mathematical description of the world. Irreducibly qualitative features -- secondary qualities, final causes, and the like -- since they would not fit this model, were thus essentially defined away as mere projections, “swept under the rug” of the mind as it were. But that only makes the idea of dealing with the mind itself in the same manner even more hopeless. For these early moderns, the mind just is, you might say, the holding tank for everything that doesn’t fit their quantitative method. Naturally, then, that method cannot coherently be applied to the mind itself.
Now the lesson Nagel drew from this in the 1974 article was not that physicalism is false so much as that “physicalism is a position we cannot understand because we do not at present have any conception of how it might be true.” But the early moderns who inaugurated this conceptual revolution tended to draw the stronger conclusion. Indeed, writers like Cudworth and Malebranche saw that the method in question can be used to argue for a kind of mind-body dualism. For if you maintain that color, sound, heat, cold, odor, taste, etc., as common sense understands these features, do not exist in matter, then they do not exist in the brain or body any more than they exist in the material world external to the brain and body. If they do exist in the mind, though, then the mind must not be material. Dualism can hardly be refuted by the reductive method, then, precisely because dualism follows from that method.
Now that conclusion is actually a bit too strong, though the Cudworth/Malebranche style of argument has had defenders down to the present (e.g. Richard Swinburne in The Evolution of the Soul). For one could argue instead, as Berkeley did, that only the qualities we know of in conscious experience are real and the mathematically-redefined material world is a mere fiction -- idealism rather than dualism. Or one could argue, as Russellians do, that the sensory qualities presented to us in conscious experience are what “flesh out” the abstract structure described by physics -- thereby putting something like color, sound, etc. as common sense understands them back into matter after all, but in a way that is very different from the way common sense supposes them to be there. (As David Chalmers suggests, though, this really amounts to a kind of riff on property dualism.) And of course, we Aristotelian-Thomistic philosophers would reject the assumptions that lead to this tangle in the first place, dismissing Cartesianism and materialism alike as riffs on the same fundamental error of treating what are really just physics’ useful mathematical abstractions from concrete material reality as if they were the whole of concrete material reality.
What you cannot coherently be, consistent with the reductive method described, is any sort of reductive materialist, which has been at least historically the standard form of materialism. And this, I would say, is why materialism was so rare in modern philosophy before the late twentieth century. It takes real historical ignorance seriously to think that the scientific revolution somehow supports reductive materialism and that Descartes, Locke, Berkeley, et al. were non-materialists merely because they didn’t have the courage and/or foresight to follow out the implications of that revolution. In fact, they were the ones who were consistently following out those implications, while materialists like Hobbes were the wishful thinkers. Perhaps the proud ignorance of the history of philosophy that some (though by no means all) of the early analytic philosophers exhibited made it possible for materialism widely to come to seem plausible by the 1960s. (To paraphrase Newman, to be deep in the history of philosophy is to cease to be a naturalist. That’s certainly what led me away from naturalism, anyway.)
Be that as it may, the only way to be a materialist consistent with the method Nagel describes is to be a materialist of the eliminative rather than reductive kind. Contemporary non-reductive materialism fails as a third option because it fails to be materialist. To acknowledge that higher-level features of the natural world are as real as the lower-level features but irreducible to them is either property dualism or an implicit Aristotelianism -- essentially a recapitulation of Aristotle’s critique of the ancient atomism that is the ancestor of modern materialism.
The most perceptive naturalists -- Nagel, John Searle, Galen Strawson, David Chalmers, and Alex Rosenberg, for example -- see that the only real choice is either to embrace eliminativism or give up materialism in anything like the extant forms. Rosenberg takes the first option, while Nagel, Searle, Strawson, and Chalmers in their different ways take the latter. Eliminativism, however, cannot possibly be right. Insofar as it denies the existence of intentionality, it cannot so much as be coherently formulated. Insofar as it denies the existence and/or reliability of conscious experience, it undermines its own evidential base (a problem which, as I noted in an earlier post, Democritus saw over two millennia ago and Erwin Schrödinger and E. A. Burtt saw in the 20th century). That leaves us with the latter option -- which in Searle leads to an implicit property dualism, in Chalmers to an explicit property dualism, in Strawson to panpsychism, and in Nagel to an implicit Aristotelianism. But whether any of these views are in an interesting sense “naturalist” is something both naturalists and non-naturalists might doubt. And the Aristotelian option is, as I would argue, when followed out consistently going to lead to a Scholastic form of theism.
Leiter and Weisberg on science and common sense
But that is an argument for another time. Let’s turn now to Leiter and Weisberg, who write:
Nagel opposes two main components of the “materialist” view inspired by Darwin’s theory of evolution by natural selection. The first is what we will call theoretical reductionism, the view that there is an order of priority among the sciences, with all theories ultimately derivable from physics and all phenomena ultimately explicable in physical terms. We believe, along with most philosophers, that Nagel is right to reject theoretical reductionism, because the sciences have not progressed in a way consistent with it. We have not witnessed the reduction of psychology to biology, biology to chemistry, and chemistry to physics, but rather the proliferation of fields like neuroscience and evolutionary biology that explain psychological and biological phenomena in terms unrecognizable by physics. As the philosopher of biology Philip Kitcher pointed out some thirty years ago, even classical genetics has not been fully reduced to molecular genetics, and that reduction would have been wholly within one field. We simply do not see any serious attempts to reduce all the “higher” sciences to the laws of physics…
So far so good, though Leiter and Weisberg do not seem to realize how large a concession this is. As I have said, to affirm the reality of irreducible levels of the natural world above the level described by physics is essentially to affirm either property dualism or Aristotelianism -- and as I indicated in my previous post, the neo-Aristotelian implications of anti-reductionism are in fact recognized and embraced by a number of prominent contemporary philosophers of science and metaphysicians. It will not do, then, merely to insinuate (as Leiter and Weisberg go on to do) that Nagel is attacking a straw man when he attacks reductive materialism. For what matters is not whether most contemporary naturalists in fact explicitly affirm the reductionism Nagel rejects. What matters is whether they can consistently reject that reductionism themselves without also moving in a property dualist, neo-Aristotelian, or other non-materialist direction.
Leiter and Weisberg continue:
The second component of the thesis Nagel opposes is what we will call naturalism, the view that features of our world—including “consciousness, intentionality, meaning, purpose, thought, and value”—can ultimately be accounted for in terms of the natural processes described by the various sciences (whether or not they are ever “reduced” to physics)…
This is perhaps a bit too proprietary a use of “naturalism” given that Nagel himself claims to be a kind of naturalist, but let that pass. In their first main criticism of Nagel, our authors go on to write:
Naturalists… defend their view by appealing to the extraordinary fruitfulness of past scientific work, including work growing out of Darwin’s theory of evolution by natural selection. So what should we make of the actual work in biology that supports the “materialist Neo-Darwinian conception of nature” that Nagel thinks “is almost certainly false”? Defending such a sweeping claim might seem to require a detailed engagement with the relevant science, yet in a striking admission early on, Nagel reveals that his book “is just the opinion of a layman who reads widely in the literature that explains contemporary science to the nonspecialist.” And a recurring objection to what he learned from his layman’s reading of popular science writing is that much science “flies in the face of common sense,” that it is inconsistent with “evident facts about ourselves, that it “require[s] us to deny the obvious,” and so on…
[S]urely we have some reason for thinking, some four centuries after the start of the scientific revolution, that Aristotle was on the wrong track and that we are not, or at least not yet. Our reasons for thinking this are obvious and uncontroversial: mechanistic explanations and an abandonment of supernatural causality proved enormously fruitful in expanding our ability to predict and control the world around us. The fruits of the scientific revolution, though at odds with common sense, allow us to send probes to Mars and to understand why washing our hands prevents the spread of disease. We may, of course, be wrong in having abandoned teleology and the supernatural as our primary tools for understanding and explaining the natural world, but the fact that “common sense” conflicts with a layman’s reading of popular science writing is not a good reason for thinking so…
Now, there are three main problems with this line of response to Nagel. First, it is unfair or at least uncharitable to suggest that Nagel’s complaint in the lines Leiter and Weisberg quote here is merely that the views he rejects are contrary to common sense, as if it were a mere prejudice in favor of the opinions of the man on the street that leads Nagel to resist materialist explanations of consciousness, intentionality, etc. As we have seen, the implication of Nagel’s position on the mind-body problem is that it is incoherent, and not merely counterintuitive, to apply to consciousness and the like the methods science employs in the explanation of other phenomena. It is no good, then, merely to point to cases where science has upended common sense, for Nagel has offered reasons to think that such upending could not in principle occur in the case of consciousness, intentionality, and the like. His claim is precisely that the latter phenomena are necessarily resistant to the same mode of explanation, given the nature of that mode of explanation. In fairness to Leiter and Weisberg, and as I have already conceded, Nagel does not recapitulate in Mind and Cosmos all the arguments of his earlier work. All the same, those arguments are well-known, and it is only fair for Nagel’s critics to take account of them when interpreting the claims he makes in the new book.
Second, it is for the same reason a mistake to assume that the dispute between Nagel and his critics is essentially a scientific dispute and that Nagel’s status as a layman dependent on popularizations of science casts doubt on his claims. For in fact the dispute concerns the philosophy of science and the philosophy of nature that contemporary scientists tend to take for granted. Does the essentially mathematical conception of nature we have inherited from Galileo, Descartes, and Co. capture all aspects of nature? Is the methodology associated with that conception an appropriate means of discovering and studying all aspects of nature? These are essentially metaphysical and epistemological questions rather than empirical scientific questions, and Nagel’s position is that they must be answered in the negative. Of course, Leiter and Weisberg might insist (after the fashion of “naturalized” epistemology, metaphysics, etc.) that all such philosophical questions must ultimately be answered through scientific means, but merely to insist on that is simply to beg the question against Nagel rather than to refute him.
Third, it also merely begs the question to suggest that the “fruitfulness” of “mechanistic” explanations in other domains -- where fruitfulness involves the ability to “predict and control” natural phenomena -- gives us reason to think that such explanations might be given of the phenomena at issue in Nagel’s book (consciousness, intentionality, etc.). For one thing, Nagel has, as I have said, given reason elsewhere to think that such explanations cannot succeed. For another, Leiter and Weisberg are here committing a fallacy similar to the one which, as we saw in an earlier post, Alex Rosenberg commits in his book The Atheist’s Guide to Reality. In particular, they are essentially arguing as follows:
1. The predictive power and technological applications of [post-Galilean, post-Cartesian, mechanistic] science are unparalleled by those of any other purported source of knowledge.
2. Therefore we have good reason to think that [post-Galilean, post-Cartesian, mechanistic] science can explain everything that there is to explain.
And that sort of argument is no better than this one:
1. Metal detectors have had far greater success in finding coins and other metallic objects in more places than any other method has.
2. Therefore we have good reason to think that metal detectors can reveal to us everything that there is to be revealed.
In fact, of course, metal detectors are as successful as they are in finding coins, lost keys, etc. precisely because they focus only on those specific aspects of coins, keys, and the like which might be detected via their methods (i.e. the metallic nature of these objects) and ignore everything else (the shape, color, etc. of the objects). And the methods of post-Galilean, post-Cartesian, mechanistic science are as successful as they are in predicting and controlling natural phenomena precisely because they focus only on those aspects of nature susceptible of strict prediction and control (especially those aspects which might be modeled mathematically) and ignore everything else (e.g. any irreducibly qualitative or non-quantifiable features that might exist in nature, such as teleological features, the phenomenal feel of heat and cold, the phenomenal look of colors, and so forth). But just as metal detectors are inevitably going to fail to capture non-metallic phenomena, so too are the methods of post-Galilean, post-Cartesian, mechanistic science inevitably going to fail to capture any aspects of nature not susceptible of prediction and control, nor capable of being captured via the mathematical techniques that make prediction and control possible.
Of course, the naturalist might deny that there are any such aspects, but the point is that to appeal to science in order to support such a denial is utterly fallacious -- as fallacious as appealing to the success of metal detectors in order to support the claim that only metal exists. If there are any non-metallic aspects of nature, you should not expect to find them using metal detectors; and if there are any aspects of nature that elude strict prediction, control and mathematical modeling, you should not expect to find them using the methods of post-Galilean, post-Cartesian, mechanistic science.
Leiter and Weisberg on rationality and consciousness
In response to Nagel’s argument to the effect that rationality, specifically, cannot be explained in purely Darwinian terms, Leiter and Weisberg write:
There is a response to this kind of challenge, one that is widely embraced by philosophical naturalists (though, again, not mentioned by Nagel). This response starts by noting that we determine what is “rational” or “justified” simply by appealing to the most successful forms of inquiry into the world that human beings have developed. Paradigmatic examples of those successful forms of inquiry are, of course, physics, chemistry and biology. They are successful precisely in the way that Aristotelian science was not: they enable us to navigate the world around us, to predict its happenings and control some of them. To confuse one’s intuitive confidence in the logical and epistemic norms that make these sciences possible with some kind of a priori access to the “rational order of the world,” as Nagel puts it, is to forget whence that confidence derives—namely, the very success of these sciences. For philosophical naturalists, the charge of circularity is empty, akin to suggesting that the need for a usable table to have legs requires some justification beyond the fact that the legs actually do a necessary job.
There are several problems with this response. Start with the claim that “we determine what is ‘rational’ or ‘justified’ simply by appealing to the most successful forms of inquiry into the world that human beings have developed” (emphasis added) -- where “success” entails prediction and control of the sort characteristic of post-Galilean, post-Cartesian, mechanistic science. Whom exactly do Leiter and Weisberg mean by “we”? They seem to mean “we, that is, people in general, or at least people who have an opinion about these matters.” But in that case their claim that “we” determine standards of rationality and justification in this way is false, since Nagel and others who reject Leiter and Weisberg’s brand of naturalism do not in fact agree that our standards of rationality and justification are entirely determined by the considerations of what is conducive to prediction and control. Indeed, that claim is (pretty obviously) precisely part of what is at issue in the present dispute between Nagel and his critics. So by “we” do Leiter and Weisberg instead mean “we, namely Leiter and Weisberg, and like-minded naturalists”? In that case they may be accurately representing the views of the “we” in question, but they will have given no reason to think that those views are correct. For why should anyone agree that success vis-à-vis prediction and control alone determines what is rational or justified? Leiter and Weisberg do not tell us (other than by insinuating the fallacious Rosenberg-style argument already criticized above).
A second problem is that this essentially Quinean pragmatist claim about what ought to determine our standards of rationality and justification is simply not at all plausible. Take an inference rule like modus ponens, or a principle like the law of non-contradiction, or an elementary truth of arithmetic such as 2 + 2 = 4. Are we really expected to believe that there was no rational justification for any of these until they were somehow worked into a body of scientific theory that passed the test of empirical prediction and control? Why would anyone take such a proposal seriously for a moment, unless they saw it as necessary to salvaging naturalism from objections like Nagel’s? Say what you will about the empirical credentials of the work of Pre-Socratic natural philosophers, Euclidean geometers, and the like -- they were certainly engaged in a rational enterprise at least insofar as they were capable of evaluating each other’s arguments vis-à-vis standards of consistency, deductive validity, and the like, and we did not need to wait upon the rise of modern science in order to know that much.
Nor would it do to suggest that it was prediction and control in everyday, ordinary non-scientific contexts that justified at least the elementary truths of logic and mathematics before the rise of modern science. For one thing, logically fallacious forms of inference are notoriously useful in everyday life for purposes of control, but they remain fallacious for all that. For another, even in ordinary, non-scientific contexts we judge our beliefs and practices by reference to standards of logic and mathematics, not vice versa. For example, we judge that hypocrisy is bad because it involves a logical inconsistency between one’s expressed opinions and one’s actual practice; we don’t judge that logical inconsistency is bad because it is evident in things like hypocrisy, which we dislike for independent reasons.
That brings us to another point, which is that Leiter and Weisberg’s proposal is not merely counterintuitive and ill-founded; it is also incoherent. For basic logical notions like truth, consistency, entailment, and the like and are simply more fundamental than notions like predictive success or technological control, insofar as the latter presuppose the former. To test a theory’s predictions is to determine what the theory entails, to determine whether what it entails is consistent with what is observed, to judge that this consistency with observation is a mark of truth, and so forth. To apply a theory practically by way of technology or other means of control is also to determine what the theory entails, to determine whether a proposed technology is consistent with what it entails, and so on. And of course both prediction and control also involve measurement, and thus adding, subtracting, and the like. We cannot coherently regard logic and mathematics as deriving their rational justification from empirical science, then, because empirical science itself presupposes logic and mathematics. To Nagel’s objection that Darwinian justifications of standards of rationality are circular, then, Leiter and Weisberg in effect offer in response nothing more than a further circular argument.
We should also note that Leiter and Weisberg’s sweeping dismissal of “Aristotelian science” threatens to run together issues which need to be carefully distinguished. We can agree that geocentrism, the theory of the four elements, and other specific empirical claims to which Aristotelian scientists of the past were committed have been falsified. But it doesn’t follow that the philosophical notions contingently associated with (and sometimes illustrated by reference to) these erroneous claims -- immanent teleology, essentialism, and the like -- have also been falsified. Aristotelian physics is one thing, Aristotelian metaphysics and philosophy of nature quite another. Certainly it would simply beg the question yet again to insinuate that the falsification of the former by itself casts doubt on the latter.
Leiter and Weisberg also object to Nagel’s claim that consciousness cannot be explained in Darwinian terms, and they seem to think that Nagel’s reasons for making this claim have essentially to do with the question of whether the rise of conscious organisms could have been predicted from the state of the material universe prior to their origin. Hence they write:
Philosophers of science have long argued that explanation and prediction cannot be fully symmetrical, given the importance of probabilities in explaining natural phenomena. Moreover, we are often in a position to understand the causes of an event, but without knowing enough detail to have predicted it. For example, approximately 1 percent of children born to women over 40 have Down syndrome. This fact is a perfectly adequate explanation of why a particular child has Down syndrome, but it does not mean we could have predicted that this particular child would develop it. Causes alone are frequently deemed sufficient to explain events; knowing enough to predict those events in advance is an important scientific achievement, but not essential to explanation.
But this simply misses Nagel’s point entirely, at least if we read Mind and Cosmos in light of the earlier work of Nagel’s cited above. If the argument of “What Is It Like to Be a Bat?” is correct, then it is not merely improbable that what Nagel there calls “objective” facts should by themselves give rise to “subjective” facts, but impossible, for they differ qualitatively rather than merely quantitatively. To borrow an example used by the Thomist William A. Wallace in another context, a polygon is just a different sort of thing from a circle, no matter how closely you approximate a circle by adding sides to a polygon. And that a circle might arise from nothing more than the successive addition of sides to a polygon is therefore not merely improbable or unpredictable; it is impossible in principle. Similarly, given the difference between “objective” and “subjective” facts as Nagel characterizes them, you are simply not going to get the latter from the former alone even in principle. At any rate, if Nagel is wrong about this, Leiter and Weisberg haven’t done anything to show that he is, but have merely implicitly assumed that he is.
Leiter and Weisberg also make some remarks about Nagel’s views about ethics which I think do not get to the nub of his position, but since my own views on meta-ethics are perhaps a bit further from Nagel’s than my views on general metaphysics are (and since this post has already run on long enough) I’ll leave it at that for now.