Jerry Coyne on Scientism and Freewill

Jerry Coyne is a popular new atheist and biologist. In his blog, Why Evolution is True, he often defends two positions among others, both of which I think are dubious. The first is scientism in what I call the narrow, academic sense, that science is the only source of empirical knowledge. (Note that this sense of “scientism” is different from my broader use of the word as a synonym for the substitute religion of secular humanism. That broad sense of "Scientism" isn’t relevant to the present discussion. Whenever I refer to scientism in this particular article, then, I have in mind the narrow sense.) The second is that freewill is an illusion, since science shows that determinism is true. I’ll address each in turn

Scientism and Knowledge

Coyne says that he’s ‘always maintained that there are no other reliable ways of knowing beyond science if one construes science broadly--as meaning “a combination of reason and empirical observation.” ’ Again, “The real question is whether there’s any way beyond empirical observation and reason to establish what is true about the world.  I don’t think so…” (see here). In another article, he speaks of his challenge to Keith Ward, which was ‘to give me just one reasonably well established fact about the world that comes from “general philosophical views, moral views, personal experience and judgment” without any verifiable empirical input.’ Coyne summarizes this by saying that he ‘questioned Ward’s contention that faith or other non-empirical “disciplines” could establish facts about the world or universe’ (see here). And in an article on whether the humanities are scientific, he says, “There is only one way of finding out what is true, and that doesn’t involve revelation or making up stories” (see here). Again, his point is that science broadly construed is that only way. Finally, in an article on whether fiction is a way of knowing, he says here,

it’s clear that disciplines like history, archaeology, and even sociology have the capacity to tell us true things about the world, but I have my doubts about the arts.  Either they can present some facts (like the facts peppering historical fiction like War and Peace) that we can independently verify, or they can give us an idea of what someone felt like in a particular situation (as with Gabriel at the end of Joyce’s The Dead).  The latter, though, is not a “truth” in the normal sense, but a rendition of emotions: a way of seeing but not knowing. 

According to Coyne, then, the question of scientism is whether there are ways of knowing besides reason and empirical observation, where “knowing” means the discovery of facts or truths. Where Coyne goes wrong here was shown long ago by Plato: knowledge isn’t just the possession of a true (veridical) belief, since someone can come by such a belief by chance or by being misled and we wouldn’t say this person knows what she’s talking about. Thus, Plato famously added that knowledge requires that the true belief be justified, or supported by reasons. This is to say that the belief must also be acquired in the right way for it to count as knowledge. If knowledge were just the possession of a true belief, where truth is correspondence between the belief and a fact, and a belief is a symbolic representation of that fact, not just lucky people but inanimate objects like books or billboards could be said to know what they represent, which would be absurd. Knowledge is something possessed by a mind, because knowledge must be acquired in a way that only a mind can manage.

Once this is understood, we can see that what motivates the talk of scientism and of nonscientific ways of knowing is an emphasis on the justification side of knowledge, as opposed to the truth side. The assumption is that the humanities and the arts count as legitimate nonscientific ways of justifying true statements. Now, the scientific method of proving a hypothesis is well-known: the hypothesis is confirmed or disconfirmed by clever public tests that isolate the relevant variables, eliminating chance and subjective factors, and letting the facts speak for themselves. Is that basic scientific method the only way of justifying true statements? Note that were there others, these other methods would count as ways of obtaining knowledge because they would amount to nonscientific but still legitimate means of acquiring veridical beliefs.

To answer the question, we have to look at what’s meant by “epistemic justification.” A belief is relevantly justified when the belief is acquired not just by chance but by some sort of respectable mental labour. This labour is what philosophers call the search for rational equilibrium, which means the search for the coherence of our beliefs with each other. The goal is to avoid cognitive dissonance, the fragmentation and incommensurability of our beliefs and thus a split between the sides of ourselves that those beliefs express, and to achieve intellectual integrity which requires deep self-awareness, the classic philosophical virtue. What’s meant by “coherence” here is harder to explain, but one relevant factor is ethical: in attempting to render our beliefs epistemically coherent, we should demonstrate certain virtues such as respect for truth and for those who may be impacted by our beliefs; courage to face harsh truth; skill at handling the complex issues that can arise in learning what’s true; and artistic creativity in expressing or otherwise applying a true belief. The point of epistemic justification is to ensure that the true belief is reliable rather than accidental, and the cognitive virtues are the sources of that reliability.

With this in mind, contrast New Age ideology with modern, naturalistic philosophy. Even were some New Age beliefs to turn out true, we’d have reason to doubt that New Agers know what they’re talking about when they hold those beliefs, because their beliefs wouldn’t be well-justified in the above sense. New Age speculations aren’t currently the fruit of a virtuous search for reflective equilibrium, since those speculations tend to be anthropocentric, whereas modern science decentralizes us. Either New Age myths or modern science must go, as they stand, and by accepting the former, the New Ager shows little willingness to reconcile her worldview with the latter. Moreover, New Age speculations, such as the ones found in the Oprah-approved book, The Secret, cynically spiritualize capitalistic, social Darwinian ideology, holding the consumption of material goods as the ultimate value. According to this sort of “spiritual” worldview, we’re magnets that attract what we most think about, and this notion breeds contempt for sufferers since supposedly they get what they deserve. This worldview is insanely optimistic in concluding that all natural events on this planet are perfectly just, and so the New Ager here doesn’t evince much courage in confronting the abundance of disheartening truths discovered by modern scientists, about the moral indifference of natural forces to our welfare and about our animal rather than angelic nature.

By contrast, naturalistic philosophers arrive at general naturalistic truths through a more ethically respectable process of reaching reflective equilibrium. Modern philosophers think logically, but they also speculate and explore and defend intuitions. But arguably, these latter, nonscientific mental labours are epistemically justificatory, because they attempt to satisfy ethical standards of conduct. For example, when rationalist, empiricist, existentialist, or mysterian philosophers speculate or intuit metaphysical or other philosophical propositions that might turn out true, they do so in a conscious effort to unify modern science with intuitive self-knowledge. They courageously confront the fact that modern science seems to undermine most of our intuitions about our place in the world, and they creatively reflect on how some of those intuitions might be preserved in a rationally respectable manner.

Self-Refuting Positivism

Some naturalistic philosophers, such as the positivists, argued that intuitions or presumptions are cognitively worthless and that only scientific methods yield knowledge. Their recommendation was to dispense with any belief that isn’t supported by scientific methods. This scientism led to a dead end, however, since scientific methods don’t support the positivist’s contention that all worthwhile, “meaningful” cognitive endeavours are exclusively scientific. Positivism presupposes a nonscientific evaluation of science, a pragmatic attitude, or a Philistine prejudice, and these are philosophical rather than scientific issues. In short, a superficially antiphilosophical bit of philosophy is naturally self-refuting. The important point here, however, is that positivist philosophers themselves came to this conclusion, because even they were committed to the Western philosophical tradition which values intellectual integrity.

For example, Rudolph Carnap distinguished between external and internal questions, where external ones are about the choice of a language and internal ones are framed in a way that presupposes the language’s rules. The external questions are answered in what Carnap called a pragmatic, sociological, and nonphilosophical fashion. Thomas Kuhn argued that in the history of science, Carnap’s distinction amounts to that between paradigm shifts and normal, puzzle-solving work. What emerged from these distinctions is greater attention to the values that are presupposed by paradigmatic work and that come to the fore in clashes between theories during a paradigm shift. Suppose a theory’s reign comes to an end, because sufficient amounts of data are rendered anomalous by that theory, and suppose that a new theory gains favour not because of its intellectual qualities, but because its champion holds a gun to everyone’s head and so scares his colleagues into submission. Even were that new theory to turn out true, none of the terrorized scientists could be said to know the facts as told by the theory, for the above reasons having to do with epistemic justification. Again, the more respectable search for reflective equilibrium--even in a power vacuum when there’s great uncertainty about how to explain certain anomalies--is guided by ethical and aesthetic values, including simplicity, beauty, fruitfulness, and so on.

You can stipulate with the positivist that this value-laden mental labour isn’t relevant to the search for knowledge, but then you’ll have to show how science alone warrants that stipulation. Instead, what most analytic philosophers learned from that period of the philosophy of science is that knowledge isn’t as simplistic as theorized by empiricists. Cognitive science, too, supports a broader conception of cognitive processes, by reminding us that reason plays an evolutionary role and by showing, in any case, that we’re not so rational after all, that our most natural modes of thinking are technically fallacious and biased. The point is that assuming that nonscientists possess some knowledge, knowledge had better not be the result just of rigorous logic or hypothesis-testing. And indeed, as long as a nonscientist, or indeed anyone in her daily life, strives to be virtuous and artistically creative in her thought processes, and the result is that those thoughts put her in touch with the corresponding facts, we’ve got the makings of knowledge on our hands.

Other Ways of Knowing

What about pure art, such as painting or fiction? Is either of these a way of knowing, on this picture? Suppose you conclude that life is tragic after you gaze at a sad painting or read King Lear. But suppose you conclude as much only because you’re forced first to consult Coles Notes, since alas the meaning of the artwork otherwise escapes you. Assuming that life does have its tragic side, you’d nevertheless not possess nonscientific knowledge of this fact, since there’s little ethically or aesthetically impressive about repeating the slogans found in a popular commentary on some artwork. But suppose instead that you have a profound emotional experience in an organic response to the art, that your way of perceiving the world is thereby drastically altered. As long as you exhibit some relevant virtues in your struggle to harmonize this new experience with what you thought you knew beforehand, your belief about life’s tragic side is nonscientifically justified, and so you’ve employed a nonscientific way of knowing the facts. Instead of justifying your belief with an argument or an experiment, you've virtuously modified your worldview through an experience of an emotionally-powerful artwork.

Finally, I want to point out that Coyne misconstrues what’s at issue when he challenges folks to present knowledge that has no empirical input or any other overlap with scientific methods. This challenge is quite unhelpful since it can be turned on its head. Why not challenge the scientist, instead, to present scientific knowledge that isn’t arrived at in part by petty squabbles, turf defenses, or other power dynamics? Were all scientific knowledge produced in part by such natural processes, we could then say that there’s no such thing as a distinctly scientific way of knowing, that all knowing is fundamentally an appeal to power.

Instead, the interesting question is whether knowledge is justified only by a scientific brake on our subjectivity or also by the most respectable ways of being subjective in our thinking. The former proposition is self-refuting scientism, and so we’re forced to accept the latter, in which case the alternative ways of knowing remain so despite their overlap with broad notions of rationality. This is because these other cognitive modes are alternatives to the scientistic contention that science (rationality) alone is the only such mode. On the contrary, the humanities and arts provide alternatives if only by subordinating rational standards to ethical and aesthetic ones in an effort to create a coherent worldview, that is, a set of beliefs that includes scientific knowledge as a subset. For example, modern philosophers try to deal virtuously and creatively with the conflict between folk intuitions, which call for some awe as outputs of millions of years of evolution, and with modern science which threatens to bring liberal civilization crashing down around us, mocking the traditional self-images that keep us sane and happy.

One such intuition is that we’re free in the sense of being self-controlling and thus responsible for our actions. Do we know that we’re free even if science implies determinism, that is, the view that every event has a cause? We feel that we’re free, because we feel we have overriding power over our actions. As Coyne says about fiction, though, feeling that something’s true is a way of seeing but not of knowing since, I take it, feeling something doesn’t make it true. What we feel here is irrelevant, since neither is the world round because scientists confirmed this by looking through a satellite’s telescope. The facts are what they are regardless of what cognitive methods we bring to the table (except perhaps in quantum mechanics). The question is whether some subjective labour, such as the everyday experience of being metaphysically free, epistemically justifies a statement that might turn out to be true, such as the statement that we’re free in that respect. Again, this would depend on how much effort is put into harmonizing the one belief with our other beliefs, including those that derive from science which posits natural laws and myriad mindless processes in which we seem to be caught up.   

Freewill and Levels of Explanation

Now, Coyne says we’re not in fact free, since physics says there’s nothing that’s both self-causing and responsible for itself. Particles may pop into existence from a quantum vacuum, but were we self-creating by way of our actions in that respect, we wouldn’t be responsible for them and freewill would be useless for moral purposes. As the philosopher of science, Massimo Pigliucci, points out here, this mechanistic construal of current physics, according to which causes and effects are metaphysically real, may be outdated. Instead, what’s real may be mathematical structures or patterns that can be explained at different levels. In this case, the determinist’s principle that everything has a cause, including our choice to act in a way we consider free, would be neither here nor there.  

But whatever the case with regard to the metaphysical interpretation of physics, Coyne’s determinism is undone by the fact that there are nonreducible levels of explanation. Even were every physical event to have a prior cause, forming a causal chain, this wouldn’t mean that every psychological event is a link in such a chain extending past the person, since a person as such isn’t a physical object. When you think of a person, as such, that is, in either the layperson’s way as a conscious entity with beliefs, desires, rights, and so forth, or in the technical, psychologist’s way as a naturally selected program for processing information, you’re not thinking in terms set just by physics.

You can, if you like, perform a gestalt switch and leap from psychology to physics in your conception of a person, in which case you assume a person is identical with the stuff from which the body is made and with the processes animating that body, and you assume also that the body is identical with a set of physical particles and their interactions. But those pseudoreductions call upon miracles in just the way that a Christian declares that somehow God raised Jesus from the dead. You can wave your hands and say that psychological, biological, and physical events are all surely natural, but that notion of “natural” is philosophical rather than scientific. Otherwise, we’d have an overarching scientific theory that reduces each level of scientific explanation to a more general one. There’s no such theory. For example, no one knows scientifically how to think about psychological categories in purely quantum mechanical terms. No one can translate the one language into the other. No one can capture the full meaning of psychological statements about human behavior using language that refers only to mass, quantum fluctuations, and so forth.

Is this nonreducibility of certain levels of explanation a metaphysical or a mere epistemic matter? That is, are we just currently ignorant of how to understand everything in physical terms or is the notion of such understanding misconceived, considering how the universe is put together? Are there emergent properties, such as consciousness, that can’t be predicted or explained in lower-level terms? Are natural processes genuinely creative in that respect, as the biologist Stuart Kauffman says? I’m going to pass over these important questions here, except to say that there are compelling mysterian arguments that the subjective aspect of consciousness, of what it’s like to be aware of the world, can’t in principle be understood in its entirety from any objective, scientific perspective.

Instead, I’m going to outline how freewill is possible on the assumption that there are nonreducible levels of explanation. The point is just that the concept of freewill might be useful in explaining special, rare phenomena like human behaviours. As for the question of how that concept would fit into the physicist’s picture of nature, there would be no contradiction were physics and psychology incommensurable. Assuming there are emergent properties, a theory that addresses them is just irrelevant to a broader theory that explains the relations between other properties.

But does talk of freewill violate basic naturalistic assumptions? In other words, must morally useful freewill be supernatural? I don’t see why this should be so. Like natural consciousness, natural freewill would surely emerge from the brain, and the trillions of interconnections between neurons and synapses make the brain not just the most complex, but the most self-contained, known thing in the universe. Something that’s highly complex, with highly interdependent parts can be relatively self-contained and thus self-controlling, and the capacity for moral responsibility follows. The brain isn’t perfectly self-contained, since it has senses which connect it with the outside world, but there’s no need for morally useful freewill to be absolute. We’re animals not gods, so our self-control and our moral capacities are limited. We can be freer than other animals, and much freer than rocks and tables, but still be free only some of the time and perhaps often deluded about exactly when we’re free. We can sometimes feel as though we’re free whereas instead our behavior is effectively programmed by commercial advertisements which frame issues for us and exploit our neural networks' inclination towards associative thinking. All of this makes freewill limited and thus natural and quite real. As long as sometimes or to some significant degree our brain controls itself without being significantly influenced by anything else, we can understand how our mind might be naturally free in the sense of being self-controlling and thus how we can be responsible for our actions.

The Naturalistic Fallacy: a Case Study

In a blog entry reviewing a Guardian debate between Julian Baggini and Lawrence Krauss on science, philosophy, and morality, Jerry Coyne hides behind an easily-discerned rhetorical device to conceal his scientistic prejudice against philosophy.

In their Guardian debate, Baggini uses moral questions as prime examples of legitimate, meaningful questions that are irreducibly philosophical. Coyne then says that he’s coming around to Sam Harris’ view that we can already see how science will and should replace the philosophy of ethics. And then Coyne deploys his rhetorical device for the first of three times in his review, saying “People’s view of what is ‘moral’ ultimately must rest on one or more of three things: an appeal to the consequences, an appeal to some authority (like Scripture), or some innate feeling instilled by our genes in combination with our environment (in other words, morality lies in our neurons)” (emphasis added). Coyne is saying here that if there are moral truths, these must be matters of fact which science can naturally discover, in which case philosophical ethics isn’t autonomous or irreducible to psychology or biology.

But notice the ambiguity of the construction “rest on.” “Y must rest on X” could mean either that X causes Y or that X logically implies Y. No one who thinks ethics is irreducibly philosophical thinks science can’t discover the causes of moral judgments, including their evolutionary history and the neural processes involved in our thinking philosophically about moral matters. So of course ethics may “rest on” such scientifically-discoverable facts, but this does nothing to show that there isn’t a big, honking naturalistic fallacy in the way of Coyne’s scientism. The naturalistic fallacy isn’t about whether some natural facts can cause us to answer moral questions in one way rather than another, so that our answers “rest on” those facts in that sense. No, as has been clear since David Hume, the fallacy operates at the epistemic level of whether Y can be inferred from X. The question is whether a moral prescription, for example, follows logically just from factual premises so that, if you like, the prescription “rests on” those premises in this quite different, noncausal sense. The point, then, is that Coyne’s use of the ambiguous construction “Y rests on X” allows him to pretend that he’s overcome the point about the naturalistic fallacy, that he’s addressed the second, inferential issue whereas science surely deals only with the first, causal one. Scientists can explain what causes what, but this has no bearing on whether a prescription follows logically from a description.

Coyne relies on this trusty device a second time, when he asks,

But on what grounds, then, do we determine whether homosexuality is right or wrong? It must rest on an appeal to the consequences (which is an empirical and scientific question), on the way most people feel about homosexuality (something that is a combination of our genes and our environment, and coded in our neurons), on sacred books and dogma, or on a combination of these. Ruling out the third, the first two are, in effect, scientific questions. [my emphasis]

He concludes that “in principle science must be is the ultimate arbiter of moral questions” (sic). For the above reasons this is nonsense. Coyne tries to hide this fact with his needlessly ambiguous phrasing, which apparently prevents him from appreciating the difference between causal and epistemological questions about morality.

After blundering about in this fashion, Coyne arrives explicitly at the question of the naturalistic fallacy:

And for those of you who say that “is” doesn’t produce “ought,” I’d like to ask you this: “well, how do we determine ‘oughts’”?  They don’t come from thin air, and they don’t come from free will.  They come from human judgment, which is a result of our genes and our environments. Why is that not, at least in principle, susceptible to scientific investigation?

Notice that his confusion persists even in his asking of the Humean question. The question isn’t whether facts “produce,” that is, cause moral obligations, but whether statements about the former imply statements about the latter. Then he switches to a third instance of the same type of rhetorical device. Now he relies on the ambiguous construction “X determines Y.” As you’ll know from the freewill-determinism debate, “determine” can mean cause, which is a scientific matter, but it can also mean ascertain from reasoning, which is an epistemological one. Still, the ambiguity of this phrase allows Coyne to attack his strawman when he reminds his reader that surely moral statements don’t come from thin air, but from our judgment, which in turn is a result of our genes and environments. Coyne is once again talking about causes, whereas the naturalistic fallacy is about inferential relations between statements (or thoughts).

All Coyne is entitled to say here is that scientists can explain the psychological or biological patterns in our moral behaviour. So if, as Hume thought, we have a separate sense of morality which he calls “sentiment”, a mental faculty of intuitions about what goals we should pursue, a faculty we might call the conscience, this faculty will “rest on” the brain, meaning that how we use that source of moral reasoning will be caused by our genes and environments. After all, we’re naturally selected to be social, and children are trained to think in moral terms. All of that is fair for scientists to explain. But this leaves entirely untouched the philosophical, epistemological distinction between two different logics, between that which licenses inferences to statements of fact and that which licenses inferences to normative statements. Just because moral statements are caused by brain states, which in turn are caused by the genes and so on, doesn’t mean those statements follow logically from scientific explanations of the causes of morality. Hence the naturalistic fallacy, which indicates that the philosophy of ethics, in which we think carefully about prescriptions without attempting to replace them with descriptions, isn’t reducible to a science of facts.