Tuesday, March 12, 2013

Thomas Nagel, Mind and Cosmos, An Unpublishable Review

Thomas Nagel has written a book, Mind and Cosmos, a slim volume mostly of other people’s ideas, on the possibility of reductive physicalist explanations for certain aspects of our minds. He believes these criticisms undermine not just the metaphysical position of physicalism or materialism but also the scientific explanations of the origins of these aspects of our minds. Thus, the general metaphysic is inadequate, but also the specific scientific theories, especially evolution by natural selection, must also fail since they are more a consequence of that metaphysic than evidential support for it.

This last is typical creationist boilerplate made more polite for a sophisticated audience: scientists only believe in evolution because they are already committed to a naturalistic or physicalist metaphysic. Without that assumption, the evidence could never be seen as adequate. The yawning gulf between what stands in need of explanation and what the (current) naturalistic picture is capable of explaining would be too great but for the naturalist’s spectacles which make that gulf seem only a small gap. What, then, are these gaps?

Nagel’s book covers three areas of human mental life, or knowledge, and argues that they cannot be accounted for either reductively (e.g. in terms of underlying or lower-level physical facts such as facts about brain activity) or historically (e.g. in terms of their origin by means of evolution). The three amigos, the Irreducibles we might call them, are consciousness, especially the subjective character of our experience, reason, especially the abstract capacity to think logically, and morality. Each of these resists reductive and/or evolutionary explanation, Nagel claims, and so our current popular conception of evolution as a potentially complete explanation of the origin and characteristics of humanity should be rejected and our worldview must be expanded to include fundamental elements of teleology in the world (even if that teleology is, as noted, elemental rather than derived from the intentions of a designer).

Nagel’s book is a light, but nonetheless baffling, read, not so much in the issues it covers, but in the major fields of research it dismisses. His casual dismissal of entire scientific disciplines and areas of philosophical research makes it much easier for him to propose his own half-baked ideas as legitimate alternatives. For example, in his introductory chapter, Nagel writes, “I would like to defend the untutored reaction of incredulity to the reductionist neo-Darwinian account of the origin and evolution of life. It is prima facie highly implausible that life as we know it is the result of a sequence of physical accidents together with the mechanism of natural selection. We are expected to abandon this naïve response, not in favor of a fully worked out physical/chemical explanation but in favor of an alternative that is really a schema for explanation, supported by some examples” (ibid). Here Nagel runs together questions about biogenesis (origin of living things) with evolution of those living things once they exist and treats them as both equally problematic. He demands a complete chemical and physical (as in, physics) explanation of development of life from non-life and of our current forms from the earliest forms of life. It is hard to believe that Nagel could dismiss the entire field of evolutionary biology as he appears to above, but he concludes this paragraph with two questions, one about biogenesis, and the other about evolution. He writes, “In the available geological time since the first life forms appeared on earth, what is the likelihood that, as a result of physical accident, a sequence of viable genetic mutations should have occurred that was sufficient to permit natural selection to produce the organisms that actually exist?” (p. 6) His implicit answer is that a sufficient number of random mutations is not very likely and so random mutation (even combined with other natural means of introducing genetic variation) cannot be the sole mechanism of evolutionary variation. Nagel really is asserting that the myriad fields of biology devoted to evolution, molecular genetics, and all the rest, constitute only, “a schema for explanation, supported by some examples” (ibid). This description would fit Nagel’s book much better than it does the scientific fields devoted to the study of evolution if only he provided a schema of explanation or some examples. Nagel’s hubris is stunning.

On the specifics of his charges of an explanatory gap, we find consciousness, reason and morality. Nagel says little enough about consciousness that what he does say is easy enough to dismiss here. No doubt he believes that this issue has seen more than its share of spilled ink, and I agree about that, and he would rather not belabor the arguments. Perhaps he is right and consciousness cannot be explained in terms of the underlying physical activities in our brains, or given an adequate functional analysis, but I do not believe that Nagel or anyone else has made this case successfully. Nothing in this short work adds anything to his case. In his famous Bat-paper , Nagel argued that consciousness is essentially subjective and scientific explanation is essentially objective. Hence, there is an essential gap in our understanding of consciousness that cannot be filled from that objective, scientific direction. We all know about the Bat paper and the not-knowing-what-it-is-like-i-ness of bats from our human perceptual and conceptual framework, but nothing in the argument implies that bat-facts are not physical facts or that bat-knowledge is not physical knowledge. Nagel bases his argument on a perceived need for extrapolation or imaginability of the bat’s experience, but that is simply irrelevant to explanation, scientific or otherwise. We may not be able to imagine what it is like to be bat (although, who knows, maybe some of us can), but it does not follow that we would have no explanation of the bat’s experience. If physicalism is true, physical knowledge of the world exhausts all the knowledge of the world. But the extrapolation from bat experience is not knowledge, it is imagined experience. Or if that bat-extrapolation is knowledge, Nagel has never established that this knowledge is not a kind of physical knowledge.

In Mind and Cosmos Nagel leaves all his bats at home (in his belfry?) and relies more on the explanatory gap version of the argument. There is a gap in our ability to understand consciousness that no objective knowledge can fill. Of course, if Nagel is correct, no matter what type of theory one proposes (substance dualist, property-dualist, emergentist, teleological) one cannot fill that gap. Unless Nagel can explain how teleology (which is not subjective either) can fill the gap, then his argument from the explanatory gap must equally undermine his own proposal. In any event, Nagel, from his armchair, demands what he has no right to demand, a full picture of a scientific theory that is still in the earliest stages of development. We don’t yet know how to conceive consciousness in a way that will render it susceptible to reductive explanation, whatever concepts do ultimately allow for the understanding of consciousness will arise from efforts at scientific explanation and we cannot foretell, effectively, whether such concepts are forthcoming without letting that research run its course. Nagel would serve us all better by helping light a candle rather than complaining about the dark, or, worse, trying to blow out whatever candle might be lit on the grounds that what we need is a spotlight.

Nagel’s second Irreducible is reason, such as our ability in logic, mathematics, and other abstract areas, perhaps relating to scientific methodology. Here Nagel relies on the ideas of Alvin Plantinga, arguing that there is no good evidence to think that mental capacities that evolved in order for us to survive would be turned to good use in proving theorems in abstract symbolic logic systems (for example). Thus, since clearly we do have such an ability, and cannot even reasonably doubt that ability, we must doubt that we in fact evolved it. First, if Nagel has ever taught a symbolic logic class, he must know that we are not very good at abstract reasoning. (In a moment we will see him exemplify the very problem here.) It requires considerable effort on our parts to learn these skills, and we are often subject to errors in reasoning that no amount of education can fully excise. We can at best substitute formal judgments based on our symbolic systems for our intuitive judgments; we cannot make the intuitive judgments fully disappear. The types of errors we make even tend to be largely consistent with evolution in that we tend to make errors that would not undermine our survival. For example, we tend to infer causation far too quickly from too little or unrepresentative data. That kind of causal error makes some sense given an evolutionary context since we are unlikely to be penalized by evolution for erroneously thinking there is a causal connection between two things, and we might be penalized (i.e. killed) for failing to draw an inference to a causal relationship that is there. Cognitive psychologists are attempting even now (with, I would say, not enough success) to trace the origins of our ability to reason using conditionals to some capacity evolved by our social, Pleistocene ancestors. People are very poor at reasoning with conditionals in abstract contexts (e.g. the Wason selection task) yet are quite capable of performing formally identical reasoning in other contexts, so the differences in the contexts may tell us what evolutionary forces operated to produce the ability. Although a good theory here is somewhat lacking (in my humble opinion), the fact that we fail at the abstract tasks suggests that we do not, in fact, have natural capacity to reason in this abstract way. Instead we may have learned to apply logical rules in some situations and have only recently learned to extend that reasoning to symbolic and mathematical forms. Evolutionary theory does not undermine the possibility of abstract reasoning. Indeed, it is consistent with the particular abilities we find that humans have. External instruments, language and symbols, have allowed us to turn simple intuitions evolved in a context far different from our current one of abstract logic, mathematics, and science. We did not evolve fingers to type on keyboards, but we have managed to turn those fingers effectively to such a task.

Perhaps the wonder of it is that we are ever capable of good logical reasoning at all, but if we were completely bad at logical reasoning, we would never survive to reproduce at all. For example, if I were incapable of recognizing contradictions in statements, so that when I came to believe, “There is a tiger over there,” I would still be capable of believing, “There is no tiger over there.” (Let’s avoid those cases in which people really are capable of believing inconsistently.) Needless to say, harboring these beliefs together would not improve my chances of survival. And, indeed, we know that people can develop languages and tools to improve this abstract reasoning just as we invented hammers and saws to improve our ability to build houses. Our ability to check with each other might just have spread the popularity of certain errors, but generally intersubjective verification of our logical rules has improved our reasoning as well. How is it possible that checking with other equally fallible people has led to better reasoning? I’m not sure, but I know that my perceptions are not always accurate, and conferring with other, also imperfect perceivers, appears to render my judgments more reliable.

Nagel is aware of the way in which external scaffolding (such as these external tools) can allow for the extension of our more basic abilities, but he rejects the response. He writes, “But the explanation of our ability to acquire and use language in these ways presents problems of the same order, for language is one of the most important normatively governed faculties. To acquire a language is in part to acquire a system of concepts that enables us to understand reality.” (p. 71)

How easy to be a philosopher! Is there research on the evolution of language? Is there any progress on the problem of explaining meaning and language? I’m no expert on this, but no doubt there is work on this question. And that work is not the same work as the study of logic and abstract reasoning. Yet, somehow, Nagel feels no need even to ask whether there is such work or, if there is such work, how successful it is. Nagel is content to assume that whatever problem afflicts one domain must afflict every related domain, so we can have no confidence that the problems can be solved (by decomposing them into simpler problems, perhaps). However, we do not need to pretend that these problems are insoluble just because Nagel remains ignorant of attempts to solve them.

Nagel himself raises the analogy of reason to perception. He writes, “[T]here is a crucial difference: in the perceptual case I can recognize that I might be mistaken, but on reflection, even if I think of myself as the product of Darwinian natural selection, I am nevertheless justified in believing the evidence of my senses for the most part, because this is consistent with the hypothesis that an accurate representation of the world around me results from senses shaped by evolution to serve that function.” (p. 80)

Surely we can, and often do, become less confident in our perceptual judgments if we come to recognize through a more reliable scientific study that our perceptual apparatus is inaccurate. We could even, in theory, discover that evolution would not provide us with the tools for an accurate perception. I may distrust my perceptual judgment that my stove is giving off no light because I recognize from other scientific research that my sensory apparatus is not attuned to the proper frequencies of light to judge accurately. So, perception is the basis for our knowledge of the external world, but it is subject to correction by testing its coherence with others' perceptions and sophisticated theorizing that, itself, is ultimately based on perception. Thus, it is possible that our more sophisticated theorizing about evolution might undermine our trust in our senses in some particular case.

Nagel intends, I think, to argue that natural selection might undermine knowledge claims based on perception (it is, in the language of epistemology, a possible defeater). However, he puts the point in a very odd, indeed logically invalid, way. He claims consistency with evolution supports the accuracy of the senses, but it is more correct to say that a failure of consistency would undermine the accuracy of the senses. These are two logically distinct points, and it is important not to confuse them. Nagel wants to say that evolution could undermine belief in the accuracy of perception, but it does not follow from this that evolution supports the accuracy of perception. The error here is a simple formal fallacy:

If not P (A is inconsistent with B), then not Q (A undermines B).
Therefore if P, then Q.

Here are two analogies to show that this argument form is invalid:
If you are not 16 or older, then you cannot drive legally.
Therefore if you are 16 or older, then you can drive legally.

Alternatively,
If Lassie is not a mammal, then Lassie is not a cat.
Therefore if Lassie is a mammal, then Lassie is a cat.

Obviously, the premise of each of these arguments is true and the conclusion false. So Nagel is guilty of a fairly simple error in formal reasoning. This problem is primarily one of presentation, but it actually nicely illustrates the problem he claims we do not have. We, all of us, even famous philosophers, struggle with formal reasoning, and part of the reason for that may be that we evolved in an environment in which abstract reasoning was not as evolutionarily useful as it is today. Thus, we might see how our belief in evolution might explain errors in abstract reasoning that we already know exist.

As I said above, Nagel considers the analogy to perception but rejects the analogy on the grounds that we could have reasons based on our understanding of evolution to reject a belief or weaken our confidence in a perceptual judgment, but we could not so doubt our logical abilities. He writes, “By contrast, in a case of reasoning, if it is basic enough, the only thing to think is that I have grasped the truth directly. I cannot pull back from a logical inference and reconfirm it with the reflection that the reliability of my thought processes is consistent with the hypothesis that evolution has selected them for accuracy. That would drastically weaken the logical claim.” (p. 80) It is not possible, he thinks, to doubt the validity of the most basic inference patterns we use. (How could we doubt the inference patterns we use to evaluate the validity of an inference pattern?)

The clause, “if it is basic enough”, appears to cover a multitude of sins. If we decide, collectively, that certain popular forms of reasoning, such as affirming the consequent, are invalid, then those must not have been basic enough forms of reasoning.

Unfortunately, Nagel does not explain precisely which logical judgments are such that no amount of evidence about the infirmities of our logical systems could undermine them. We do know, however, that people’s intuitive judgments of validity in the aforementioned Wason selection task are not to be trusted. We teach our students not to trust the inferences of denying the antecedent or affirming the consequent. Perhaps it was only the construction of abstract logical systems using rules and language that weakened our collective confidence in these patterns of inference. In theory it ought to be possible to show by means of an evolutionary argument that some inference pattern we frequently rely on is untrustworthy. I do not know of any such argument, but that should weaken Nagel’s point.

Now Nagel is likely to argue that some logical reasoning patterns are so far beyond dispute that no rational agent could deny them. For example, nothing can both have property P at a time and lack property P at that same time. Is this so intuitively obvious that no amount of evolutionary, or other, reasoning could weaken our confidence in it? Not really. It is possible, but perhaps not likely, that discoveries in quantum mechanics might lead us to doubt even this apparently indisputable belief. We might have difficulty understanding how we could be wrong, but we might still judge that we are. We might be forced to think that some other, better, reasoning must take precedence over the intuitive judgment. However, I think we have good reason to think these most basic inferences cohere with evolution. Any organisms so constituted as to misunderstand basic facts of logic (presuming this is a fact) would have, in Willard Quine’s words, “a pathetic but praiseworthy tendency to die before reproducing their kind.” We find, in fact, that an evolutionary understanding of our reason is probably compatible with these simple logical truths.

Perhaps Nagel means that, while in fact we do not doubt these logical inferences, the real issue is that we could not doubt them while remaining rational agents. But I do not think this is true; I think our systematization of logic (for example) has led to changes in what we take to be intuitively valid, and that we have in fact changed (collectively) what we take to be the basic patterns of inference. And in those cases in which we found good reason to distrust our logical inferences, we might note that they are broadly consistent with an evolutionary account of our origins. We do manage to do better than our forebears because of our construction of elaborate tools such as language, symbol systems, and ‘the’ scientific method. The fact that it took us this long to develop these things, and that they are needed at all, suggests that we are not good at abstract reasoning. This is not inconsistent with evolution.

The third amigo, the third Irreducible, is moral reasoning. Nagel believes there are abstract moral rules that govern us, that oblige us to act in certain ways. These rules could not, according to the argument of one Sharon Street, be discovered or be part of our evolutionary history. Thus, if evolution is true, we have no good reason to think we would know these moral rules. Yet, according to Nagel, we have them nonetheless. It follows that we did not evolve by means of natural selection.

This argument based on the epistemic opacity of morality may just be a special case of the abstract reasoning argument. If moral rules are abstract rules, then they should be just as amenable to discovery as any other abstract truth (about logic or mathematics, say). If morality can be naturalized in some way, then these concerns about moral knowledge would disappear. I am not sanguine about such reductive efforts. If morality can be naturalized, it can be known as other natural facts are, or it is abstract, and thus known as abstract facts are. But, either way, morality seems no less knowable than any other abstract area.

Most of Nagel’s explanation of this issue relates to the abstractness of the judgments, but it does raise the question of how knowledge of morality could enhance our fitness to survive. How does knowing moral right from wrong lead the knower to have a survival advantage over those who know only social rules and customs? This appears to be a puzzle.

First, however, we should question his adaptationism. Why think that every faculty or belief must enhance our fitness in order for it to exist? Perhaps knowledge of morality is exaptive (in Stephen Jay Gould's coinage); it might be a consequence of some other abilities that are adaptive. Just as the ability to hitchhike is not adaptive, but it follows from the ability to use opposable thumbs in other contexts, and those other abilities are adaptive. Nagel should provide some argument that a moral faculty must enhance our survival in order for it to exist.

Second, we might suggest ways in which a moral faculty would be adaptive. It is fairly clear that knowing the customs of our community enhances our survival. Whether we follow the mores of our society or not, we need to know what those mores are when we deal with others in that society. It might turn out that knowledge of morality makes it much easier to understand the customs of society. If, for example, people sometimes do things simply because they are the right things to do, moral knowledge would be essential for predicting their behavior. Altruism, for example, appears not to be an evolutionarily adaptive trait since individuals who work for the common good must give up some benefit to themselves to help others. What really is most beneficial is to appear to be altruistic while not really being altruistic at all. (This is Glaucon’s Challenge, the famous problem that drives Plato’s Republic.) Yet it may not be easy to appear to be altruistic while not being so (just as it is often easier to be honest than to lie consistently and convincingly). Perhaps knowledge of morality makes it easier to discern these rules. Similarly, we might discover when we interact with other cultures that they adhere to only some of our culture’s rules, and that knowledge of genuine morality may allow for productive interaction and cooperation with another culture. Knowledge of morality may enable predictions and explanations of the behavior of others, especially those who do not share our culture, that would not otherwise be possible. Any suggestion along these lines is highly speculative. It may be that only recently have different cultures treated each other as morally important agents, and hence moral knowledge would not help us in our dealings with them. However, speculative or not, these possibilities should be considered, yet Nagel prefers to ignore any theories about the evolution of morality and moral knowledge.

Thus, Nagel’s negative case appears to fail. He offers very weak reasons to think that we did not evolve or that our capacities cannot be explained physically or historically. The arguments are typical arguments from ignorance: we do not know how such a capacity would evolve, and so we must think that it could not. But even with no good reason to reject the naturalistic consensus, Nagel ventures into the uncharted waters of the teleological.

Nagel sometimes suggests that life and humanity exist because it is good that they exist, that the physical world is shaped in some way, not simply by causal, mechanical laws, but by intrinsic value-features of the world. He notes that much evil exists, so whatever this feature of the world is that explains our existence, it cannot be simply a matter of moral value. Nagel might also be aware that evolution is not always a ‘progressive’ process. Some organisms lose complexity in order to survive in a given environment. Some male angler fish, for example, lose a great deal of complexity in order to mate with a female and devolve, one might say, into something like a simple parasite with little of the physical complexity of the female. So the male loses all the complexity necessary to function as an independent fish, and gains in reproductive success by attaching itself to the female. Whatever teleological explanation there is for the world, it must either have rather notable exceptions or it must not lead inexorably to the good, beautiful, or complex. But now we must say that Nagel’s hoped-for explanantia have been drained of all substance. There is, he claims, a natural tendency towards certain ends, but these ends are not definable in terms of any identifiable properties of organisms or the world. So, we must conclude that there is a natural tendency towards those ends that have actually occurred (excluding, perhaps, nature’s many failed experiments and extinct creatures) and away from those that have not. These assertions of teleology are often so hedged with question-begging assumptions that they are effectively empirically meaningless.

While Nagel’s teleology is likely just an unfalsifiable posit, of no use to anyone, the danger of such ideas is that the hedging tends to be forgotten when it comes to explanations. Despite acknowledging in certain contexts that there are innumerable exceptions to the trend toward X (whatever the favored characteristic is supposed to be), the author frequently recurs to teleological explanations when the ends reflect the values of the author. Unfalsifiable claims may play no role in genuine science, but the careful hedging of those claims to avoid empirical disconfirmation is forgotten, and we get to see ourselves as being the logical end-point and meaning of all existence. This kind of egocentrism is conducive neither to humility nor to good science.

Further, while Nagel’s theory might appear vacuous, if it is to have any substance, the theory must posit teleological factors that play some role in the disposition of physical objects. Thus, the entelechies (we’ll borrow the Aristotelian term for the rest of the paragraph to indicate the telos that exists inherently in nature) in our organisms that lead them to survive must somehow cause changes in our DNA and our chromosomes so that the appropriate mutations will occur to lead to the ordained outcome. If the end of the koala bear is to dine only on eucalyptus leaves, the DNA of the koala bear must somehow be modified or constrained so that these features of the bear arise. But how are these entelechies to accomplish this? Do they have the power to cause mutations? Do they break apart DNA molecules and put them back together in line with their end? If so, these entelechies must violate whatever we believe we know about basic physics. If the development of our bodies depends on the structure of our DNA, then that DNA must be susceptible to influence from these entelechies. If these entelechies are to have this power, they must violate the laws of physics and chemistry which govern movements of molecules, or the entelechies must themselves be fundamental forces of chemistry and physics that appear only in certain contexts (presumably untestable contexts) to modify those organisms, but never in a laboratory. Otherwise, why would we never have discovered them? If the telos is endemic to the fundamental constituents of matter, then we should discover them in our particle physics labs. But we have never discovered anything other than laws (perhaps probabilistic) of nature. And, if the world is fundamentally constituted of physical matter, then there is no special reason that laboratory experiments should be exempt from these forces. So, if the telos is to exist, it must either be so subtle that it evades detection in laboratories, or so clever that it only affects the world outside the laboratory. And, again, how does the telos know whether it is operating in a laboratory, in the study of fruitflies, say, or in the natural world? The fact that we have never discovered a telos at the level of fundamental particles, or at the level of experimental genetics and biology in a laboratory, makes these teleological factors either utterly epiphenomenal or comic-book supervillains capable of evading detection yet still working their magic whenever it counts. Neither view makes much sense.

Thomas Nagel, most famous for kicking up dust and then complaining about the acuity of our visual apparatus, in his paper ‘What Is It Like to Be a Bat?’(What Is It Like to Be a Bat? Philosophical Review, pp.435-50, 1974. (Online text) now kicks about in the dirt of the biological sciences and finds himself rather unable to raise much. To conclude that his three cases, consciousness, abstract reason, and moral knowledge, somehow undermine the naturalist and evolutionary consensus in the explanation of human origins would be to prefer hope and self-aggrandizement to evidence. Nagel provides no significant reason to doubt that evolution can explain human origins or that there is a reductive explanation for these human capacities. And Nagel’s proposed alternative is so half-baked that one must speculate that unless it is part of the telos of Mind and Cosmos to be accepted by a wide audience, there is little chance that anyone, outside the circles of evidence-averse creationists (who will always look for useful tools), will take it seriously.