Naturalism and Normativity*
A philosopher with a larger historical perspective might well think that the current push for naturalism in philosophy is wildly out of date. In the English speaking philosophical world, naturalism became dominant about a century ago. Very little of the central writing in philosophy in the past century has found any explanatory role for transcendent entities, be they Platonic Forms, gods, or Cartesian minds or whatever. Those developments being so, talking about naturalism as prospective can seem extremely strange.
However, it doesn't take long to recognize that the sense of oddity derives from a further development in what is currently intended by those proposing naturalism. Naturalism, as I have understood it above, is a rejection of transcendence. It is the idea that accounts of the world and our place in it must not be couched in terms of entities falling outside the natural world. Clearly, the more recent view that looks on naturalism as something yet to be accomplished involves an extension of the notion of naturalism.
Anti-transcendence is now held to be necessary but not sufficient for philosophical work to be naturalistic. What else is required by those urging a new philosophical naturalism?
What has been added to metaphysical naturalism (as I shall call it) is an epistemological proviso concerning science. This more recent epistemological or scientistic naturalism holds that science is the only genuine, legitimate, cognitive activity.
Consequences about how to proceed in philosophy follow from that fundamental thesis. Criticism of this proposed naturalism might then start by attacking the proviso concerning science. There is not anywhere an attempt to set out a case, even in broad strokes, that science alone occupies the cognitive high ground. On the contrary, vast portions of human knowledge and human understanding, both unsystematic and systematic, are not derived from scientific inquiry. The claim that only science has cognitive value is an expression of an ideology not supported by any empirical or even scientific investigation.
Although that flat out denial of the scientistic criterion for naturalism is in the end the crucial response, I shall not be developing it here. Instead, I shall examine some further shortcomings of this style of naturalism. It must first be noted that there is a split within the current naturalistic movement. Peter Railton in an influential paper has drawn attention to this division. The split I have in mind is the same as he has noticed, although his way of describing the two positions is not the best characterization. It is also important to say at the beginning that the two versions, to which I shall not give names but simply refer to as the more and the less radical, share a common recent philosophical root: Quine is the immediate and powerful ancestor.
Let me begin the examination with the radical version, a view which is probably more strongly represented in the U.S. than in Britain. A good starting place, better than it may look at first glance, is with the claim, sometimes made, that philosophy must become empirically informed. For philosophy to be naturalized it must not only be anti-transcendent but also empirically informed. That specification of the requirement is, however, very odd. Philosophers do inhabit this world and are widely cognizant of its features and operations. All philosophical work has exhibited our empirical knowledge of our surroundings. Only if one has a picture of philosophers being pure spirits inhabiting a timeless world can it be thought that lack of empirical knowledge is a characteristic shortcoming of philosophy.
Moreover, most philosophers are deeply engaged with some empirical discipline or the other, depending upon their specialization. Someone working in aesthetics must have much empirical knowledge of art or arts. It would be a poor philosopher specializing in Greek philosophy who had no empirical knowledge of ancient Greek culture. And so on and on. In short, the idea that philosophy should or must do better with information about our world is not what the new naturalism really wants to say, although epistemological naturalists are tempted to put it that way simply because they hold that there is not any space outside of science for worthwhile intellectual activities.
Fleshing out what this type of naturalism thinks is necessary for philosophy is better put, though still without full accuracy, not as a demand for more empirical knowledge but as a demand that philosophy must be scientifically informed. For philosophical work to be naturalistic there must not only be the absence of transcendence but that it has to be scientifically informed. However, this way of stating the requirement, that philosophers and their work be scientifically informed, is also inadequate. For it is far from obvious how scientifically deficient philosophers, and especially those practicing it as old style naturalists, are. There are, of course, sciences and there are sciences. Epistemological naturalism would not, for instance, be impressed by a philosopher who had extensive geological knowledge, say in the formation of minerals, or was like Nabokov an expert in butterfly description and classification. Such a philosopher might never, in her philosophical work, mention such special knowledge (except perhaps by way of examples for sundry purposes.) For such a scientifically informed philosopher, informed in that way, might well regard it as irrelevant to their philosophical work that they have a special and deep knowledge of, say, owls.
This brings us to the crux of the matter. It is not simply being knowledgeable about some science that epistemological naturalism makes as a requirement for acceptable philosophizing. Rather, the new style naturalism involves the demand that scientific results play a central and crucial role if a philosophical work is to count as naturalistic. That is, it is precisely the idea that old style naturalism might regard cutting edge scientific information as irrelevant to philosophical thought that is being rejected by the new naturalism.
How did this new conception of naturalism and of naturalistic philosophy come to be? What are the sources of its development and acceptance?
Historians of science make a distinction between internal and external factors influencing scientific change. That is, briefly, a distinction is made between social causes and technical causes. If one were to do a thorough job on the question of the rise of the new type of naturalism, one would have to do some historical and/or sociological, work on the external or social causes of the phenomenon. And here the esteem in which science is held by the relevant portions of the social world which philosophy inhabits (and, conversely, the relatively dismal place held by, say, poetry or history in social esteem) helps produce an inclination toward making philosophy scientific on the part of younger philosophers.
But of course I am a philosopher and so much more inclined toward couching my explanation of the rise of a new style of philosophical practice in terms of internal factors, that is explaining its appearance on the intellectual scene by way of citing relatively technical questions arising within philosophy. The chief philosophical source which must be mentioned in explaining the rise of the new bent toward science in philosophy is Quine, both in his 'Two Dogmas of Empiricism' and in his subsequent 'Naturalizing Epistemology'. In examining how Quine has contributed to this new scientistic version of naturalism, it is necessary to take into account both his denial of analyticity and also the semi-independent picture of human belief that is embedded in his notion of the 'web of belief'. Quine's later work involves an explicit rejection of any specifically philosophical issues, distinct from scientific issues, in epistemology, though of course that rejection has been generalized far beyond that particular sub-area of philosophy.
In rejecting analyticity, i.e. necessity generally, in 'Two Dogmas' and thereafter, Quine left standing the other half of the historical contrast: the synthetic. All legitimate claims were now to be understood as synthetic, as empirical. One simply cannot understand the point of today's naturalism without seeing how it has derived from Quine's rejection of analyticity. It is one of the givens of current naturalism that nothing remains legitimate except claims which have empirical backing where 'empirical' is treated as equivalent to 'scientific'. Mathematics and logic alone escape that sanction, though for Quine that they do so is a matter of mere determination not to allow them be false. When in the final section of 'Two Dogmas' Quine comes to present his picture of the structure of human belief – his analogies are to blankets, spider webs and force fields – philosophy is already implicitly squeezed out. Mathematics and logic occupy the center of the web – physics occupies the next part of the circle and then, implicitly, down through the sciences until the periphery is reached with "matters of particular historical fact", e.g. there is a white house on the corner of 5th and Elm. Philosophy has no place in Quine's map of human cognition.
That squeezing out of philosophy is made explicit in 'Naturalizing Epistemology'. The questions philosophers have traditionally asked about human knowledge and human cognition generally are to be replaced by questions which natural science, in that essay specifically psychology, can productively address. Here, though he speaks of naturalization of epistemology, he clearly means that it is to become scientized.
The first and perhaps chief problem with Quine, and consequently with his philosophical descendants, who apportion all legitimate intellectual activities between the scientific and the mathematical/logical is the same as with Hume's division of claims into matters of fact and relations of ideas. If those divisions are accepted, there is no room in the intellectual world for Quine's or Hume's own work. At least Wittgenstein, at the end of the Tractatus, saw that the consequence of applying his distinctions to his own book is that all he has said is nonsense. If Quine and Hume are correct, their own writings ought to be "consigned to the flames for they contain nothing but sophistry and illusion." But on the other hand, we rightly do not think those works ought to be declared nonsense, sophistry and illusion. Instead we conclude that they are not correct in their narrow views of what forms of discourse are legitimate. The world holds more forms of intelligible talk than they (Quine, Hume, Wittgenstein) allow, and that more includes their own.
There are those who function philosophically within Quine's ambit who hold, as Quine did, that philosophy must be dispensed with and replaced by science – say Paul and Patricia Churchland. But other scientistic naturalists adopt a less radical line, thinking that it is possible to have both philosophy and science. It is these less radical versions of the new naturalism that Raillton labels 'substantive naturalism'. There are many versions of how philosophy is to be practiced in a cognitive atmosphere in which science holds the high cards. Let me use Dan Dennett as a major proponent of this kinder (to philosophy) version of epistemological naturalism. Consider the following from Dennett. "One of the happiest trends in philosophy in the last twenty years has been its Naturalization: since we human beings are a part of nature – supremely complicated but unprivileged portions of the biosphere – philosophical accounts of our minds, our knowledge, our language must in the end be continuous with, and harmonious with, the natural sciences."
There is an argument there about the place of philosophy in human intellectual life – although how to spell out the full argument is very difficult. It has the premise that we human beings are part of nature. That, of course, is going to be accepted by metaphysical naturalists as well. However, the conclusion which Dennett draws from the fact about humanity's place in nature does not follow at all. That conclusion concerns limitations on what, as a consequence of our being natural creatures, philosophical work must be like. Philosophical activity, while not banned as it is for the strict Quineans, must be continuous with and also harmonious with the natural sciences.
Let me start with the harmony. I should like to point out that there is a substantial set of assumptions behind that requirement. First, and this is one of the reasons for speaking of Dennett's view as scientistic, it is taken for granted that science is the highest form of intellectual activity and philosophers, whatever their projects, are those who must do the adapting in order to achieve the requisite harmony. I find this a most pernicious requirement. What it leads to is what I find to be a characteristic of the present practice among philosophers, especially young ones, who have been trained in the new naturalism: they are expected to study and incorporate into their work, in whatever way, the latest scientific results, but there is no reciprocal duty on the part of scientists to study the relevant philosophical material and incorporate it into their thinking.
Secondly, it encourages a complete lack of skepticism on the part of philosophers about the results of science. It is not part of the new naturalism to think as Wittgenstein did: "In psychology there are empirical results and conceptual confusion." There can be no philosophical thought that perhaps some science has quite misinterpreted the outcome of some of its inquiries and has done so in ways amenable to philosophical inquiry.
There is another assumption behind Dennett's desire for harmony. Heraclitus long ago challenged human thought on the subject of harmony. We think it desirable that the lion lie down with the lamb. But the world, as Heraclitus saw it, is founded on conflict. Hence harmony is not a desirable goal. An assumption of harmony is found in moral philosophy with the idea that everything that is genuinely valuable ought to be jointly possible. That has been strongly, and I think rightly, challenged by Isaiah Berlin and more recently, Bernard Williams and Martha Nussbaum. They see in the realm of values the same kind of world that Heraclitus discerned long ago, a world of conflict with not every good thing realizable. Yet Dennett's principle, the soft version of epistemological naturalism, treats the goal of harmony between science and philosophy as the chief and unargued desiratum of philosophical practice.
Wilfred Sellars shares the view with Dennett: Sellars speaks of the aim of philosophy as seeing how everything hangs together. But what if it doesn't hang together? What if the world is foxish and not hedgehoggish (borrowing from Isaiah Berlin)? That is, it is common, maybe humanly common if Heraclitus and Kant are correct, to assume that harmony is desirable and attainable, but that is a challengeable assumption nonetheless.
Dennett also says that philosophical work must be continuous with the natural sciences. What does he mean by that? He doesn't say. If, however, one looks at Dennett's philosophical practice, what is found is that he attempts to extend into all reaches of human thought and talk the idea of evolution by natural selection. He is not, in the works I have in mind at least, doing science, but one can see a clear sense in his case in which what he does philosophically is continuous with science, at least with one particular and important kind of science, namely evolutionary theory. He incorporates the central concepts of evolutionary theory into his philosophical outlook. What Dennett is thereby doing is, in one way, not at all different from what many major philosophers have done. Plato, Descartes, Spinoza modeled philosophy on mathematics; Aristotle on biology; Hegel on history; Russell on the new logic. Dennett's wrinkle, what is new with him, is to substitute evolutionary theory for those other kinds of intellectual activities and then to work out philosophical views in light of that piece of science. The question then becomes 'Can we make sense of the various aspects of human life when construing them as Dennett thinks we should?' Obviously about that we may well not be in harmony.
A slightly different account of the philosophical task facing this less radical type of scientistic naturalism can be found in Railton's explanation of 'substantive naturalism'. "A substantive naturalist advances a philosophical account of some domain of human language or practice that provides an interpretation of its central concepts in terms amenable to empirical inquiry." Railton does not say why one should be interested in undertaking that philosophical project. The only thought that comes to me is that he does think that philosophy has a role in intellectual life: philosophers are under-laborers, getting concepts in shape so that scientists can do the important work.
Moreover, are not the existing concepts in whatever domain already cast in terms that enable scientific inquiry? Cannot there be scientific investigations of human knowledge, say, without the naturalistic re-interpreting project? The answer is going to be 'No' – and a major reason is that a significant number of important concepts are going to include a normative element. Normativity, it is thought, is not amenable to science. Since normativity escapes the net of science it cannot be part of the natural world and the philosophical project must be to see how non-normative concepts can replace those infected with normative elements.
As an example, consider Quine and Ruth Garrett Millikan on language. Quine famously attempted to eliminate the concept of meaning from acceptable accounts of language. Why? Over and above his arguments that it is somehow a confused notion, there is a recognition on his part that meaning must go because it is a normative notion. That is, for example, to call that creature a 'cow' is to get it right or wrong, to speak correctly or incorrectly. To say that that is a cow is not simply to do what we mostly do, i.e. it is not simply a regularity. However, in Quine's view of the world, where the objects of science, and that means in his case the objects of physics, are all the objects that there are, there can be no such thing as normativity. For example, it is neither correct nor incorrect for two atoms of hydrogen to mingle with one of oxygen: that is simply what happens under specifiable conditions.
Critics of Quine's philosophy of language hold that, with meaning and thus normativity dropped by the way side, what he characterizes as language simply is not language. That is, language is a basic part of the normative structure of human life.10 It is important here to introduce a piece of clarification.
Where I have spoken about the normativity of language, many philosophers will talk of rules. The standard idea outside of epistemological naturalistic circles is that language is, in a large variety of ways, a rule-governed affair. Consequently there are embodied in any language a large set of rules governing the correctness of doing things this way rather than that. However, it probably isn't correct to speak of all and perhaps not even of the most important normative aspects of language as involving rules. Moreover, in general there are the concepts and practices that cannot be understood in terms of rules but which are nonetheless in various ways normative. Hence my preference is to talk of normativity or standards of correctness rather than rules.
The most important account of language in the less radical version of naturalism, exhibiting Railton's substantive naturalistic program, is that of Ruth Garrett Millikan. She, like Dennett, does not think that philosophy is to be abandoned as Quine does. Rather, she, like Dennett, tries to keep her philosophical work squared with science, where she has in mind by science roughly biology, centrally evolutionary theory. She expresses her overall philosophical aim as follows. "The problem I wish to draw attention [to] is that of reconciling this [the nature of language] with our view of man as a natural creature and a product of evolution." That is, she starts very like Dennett. The idea is that human beings are part of the natural world and a product of evolution. What she is interested in is not how philosophy can be fitted into that view, but how we are to understand the nature of language. Her solution to her problem is that if we accept the background, the scientistic version of naturalism, we have to be able to produce an account of language that treats it as a non-rule governed activity. For there are no rules in the world described by science.
What Millikan and scientific naturalism are committed to is that products of biological evolution cannot be creatures who live and act in light of standards of correctness. There are two questions to be asked. Where does that notion come from? Does evolutionary theory require that conclusion?
As to the first question: part of the background for making that assumption is the old idea, reinstated vigorously by Quine, that physics is the fundamental science – and as the objects studied by physics are not rule following entities, there cannot be norms in the natural world. The reply to that idea would be that the manner in which physics is basic does not confine the natural world to objects subject to the investigations of physics. That is, reductionism is not correct. However, I do not intend to pursue that topic here. I shall assume that the sense in which the claims of physics do constrain other scientific and non-scientific inquiries does not mean that those other investigations must be reducible to physics. The relationship is quite different from that simple picture.
The pertinent question here is whether evolutionary theory, quite independently of a reductionist reading of science and hence of nature, itself forbids the evolution of biological beings who follow norms. One might argue: 'Humans are part of the natural world and are products of evolution. Humans are essentially creatures for whom normativity is a central feature of their form of life. Hence, it must have been possible for such creatures to have come into being through evolution, for that long process to have produced a kind of creature who can create and live by rules and similar standards.'
Does evolutionary theory sanction such a piece of reasoning or does it, as Millikan and the current naturalism assume, preclude so thinking? It is somewhat difficult to produce a definitive answer. What seems behind the recent focus on evolution in philosophy is not some new development in the theory – though there have been those – but a newfound determination to insist that we must take seriously the fact that we are animals, and especially to take it seriously in moral theory. However, writers on evolution, chiefly biologists and social psychologists, have a problem. When the aim is to close the gap between humans and other animals, what is probably distinctive about human beings strongly tends to get lost in the shuffle. What I have in mind is this: suppose we accept that evolution has thrown up a species whose lives are deeply enmeshed with normativity. There are two responses to that. That fact can just be overlooked, ignored. In this picture the objects of biology are not quasi-billiard balls, as they are for Quine, but they, at least our species, are not thought of as normative creatures either. So we get descriptions of the distinctiveness of humans in terms of biological characteristics, hairiness or size of penis or continual sexual receptivity on the part of females, etc. We are treated as and described as social animals, as are termites and bees.
Sometimes Aristotle is said to have recognized us as social animals. But that is a mistaken reading of Aristotle. His claim was that we are, of course social animals, not solitaries, but within that class we are specifically zoon politikon. That is to say we are political animals. (Notice that his precise claim is that we are creatures whose normal (where the norm here is not prescriptive but descriptive) life is in a polis – I'm ignoring that reference to a particular kind of political organization here.) And being a political animal requires that we have rules and institutions, that we are normative creatures. As Mary Midgley points out, the lion is not King of the Beasts: kingship and political authority is not simply having power over a group of social animals. Now Aristotle when he asks what makes us a zoon politikon starts talking about the possession of language. So was instituted the long running and still not completed discussion over the place of language in animal life. My talk of normativity is deeply connected with the idea of language but is not identical with it – though language is, as I claimed above, a matter of normativity (rules if you wish.) Nor do I wish to say that normativity is what distinguishes humans from other animals – my claim is only that we are distinctive in that it is such an extensive feature of our lives.
If you roam through the writings on evolutionary theory as it pertains to human beings, you will have, for the most part, a difficult time finding discussions of rules and normativity. You may, of course, find lots of arguments about altruism, but a great scarcity of arguments attempting to show how altruism might be held to be the root of morality, that is, how the set of normative practices we describe as morality might have arisen out of altruism. In short, the issue is not for the most part explicitly raised by the relevant writers in, or employing, evolutionary theory. Consequently, it is difficult to assess Millikan's and possibly Dennett's, contention that to be harmonious and continuous with biological science (evolutionary theory) we must drop references to normativity in our philosophical work. On the other hand, there is some evidence that experts in evolutionary theory do not assume that rule-following and such is evolutionarily forbidden. Consider Franz de Waal, a primate expert. In Good Natured: the Origins of Right and Wrong in Humans and Other Animals, he is attempting to close the gap between humans and other animals. Among other things he finds that primates (and other animals) are rule followers, though in a more limited way than humans. "That animals follow rules has been known for a long time." 15 Now I'm not thrilled by the cases he presents, but that is not the point here. He, an expert on evolution, sees nothing incompatible between talking of rule following and of being a product of evolution. Let me commend one other book here. Peter J. Wilson has written a book with the delicious title Man, the Promising Animal. Wilson argues, whether rightly or wrongly is not the point, that promising is the distinctive feature of the evolution of homo from other primates. As we know from Searle, the activity of promising is intimately connected with rules (or at least something similar to rules).
The issue is this. The new naturalism holds as a central view that naturalism requires that we must think philosophically without resort to normativity (except in the statistical Millikan variety.) They present that as a consequence of accepting that human beings are a part of the natural world, a creature that has come about through evolution. But it is not in the slightest clear that that claimed consequence of evolutionary theory is really a consequence. In fact, it is more likely than not that evolutionary theory does not forbid human normativity, that the theory is neutral with respect to the question of whether a creature which follows rules could have come into being through evolution.
Sellars famous 'Scientific Image of Man' strikes me as more properly 'The Philosophers Interpretation of the Scientific Image of Man'. Naturalism, metaphysical naturalism, does not demand either the abandonment of philosophy or the uncritical acceptance of the pronouncements of science or the pictures of man and the world attributed to science by philosophers.
Paul Feyerabend had it right: just as philosophers once accepted that it was necessary to make their work continuous and harmonious with religion, we are now being urged to do the same, only with science replacing religion as the touchstone. Both are unacceptable requirements.
* This paper was read – and so has the appearance of a paper intended for oral presentation - at a conference on Ethical Naturalism at Durham University. It is overwhelmingly focused on issues concerning naturalism – the Ethical gets only a very light dusting. It has been much improved by the discussions at that conference. I should also wish to thank Professor William Hyde for his comments.
1. In speaking of the 'new naturalism' I am being far too fast. It did not begin with Quine as what I say below seems to suggest. There was plenty of it in the Positivists: recall Hans Reichenbach's The Rise of Scientific Philosophy. In the U.S. John Dewey would certainly have qualified as an adherent of the view I am describing here. What has happened is not that the view is really new, is not post-Quinean, but that in the last thirty years or so and in consequence of Quine, the view has become much more widely accepted than in the immediately previous period.
2. Peter Railton, 'Naturalism and Prescriptivity', Social Philosophy and Policy, 7:1 (1990), pp. 151-174.
3. Railton names his two types 'methodological naturalism' and 'substantive naturalism'. What he describes as the methodological version is somewhat similar to what I call the radical version of current naturalism, although there are important differences. His substantive version is roughly equivalent to my less radical style of epistemological naturalism.
4. I did not know until I was in attendance at the Durham conference that the radical version of the new naturalism seems considerably more prominent in the U.S.
5. H.H. Price is reported to have been an expert on owls though I do not think this shows up anywhere in his philosophical work.
6. For a recent work which is both scientifically informed yet nonetheless regards such information as irrelevant to philosophical work, see M. R. Bennett and P.M.S. Hacker, Philosophical Foundations of Neuroscience (Blackwell, 2003)
7. For others than the Churchlands accepting the Quinean radical form of naturalism, see the work of Jesse Prinz and of Joshua Knobe – and Knobe's program in experimental philosophy at the University of North Carolina.
8. Daniel Dennett, Foreword to Ruth Garrett Millikan, Language, Thought and Other Biological Categories (MIT Press, 1984)
9. Railton, op. cit.
10. The entire subject is currently quite contentious: Rule-following and Meaning, ed. Alexander Miller and Crispin Wright (McGill-Queen's University Press, 2002)
11. See Hubert Schwyzer, The Unity of Understanding, (Oxford University Press, 1990)
12. See the dispute between Ronald Dworkin and H.L.A. Hart on whether rules exhaust the normative catgegories employed in the law. Dworkin, 'The Model of Rules', University of Chicago Law Review 14 (1967). Incidentally, even the notion of normativity has its problems fort there are norms which do not even begin to look like rules: it is the norm for human beings to have two kidneys. Millikan exploits that way of thinking of norms, treating the norms of language as a case like that of having two kidneys. One might be inclined to rectify matters by talking of prescriptivity, as Railton does, or to rest in the end on 'standards of correctness'.
13. Millikan, op. cit, p. 7.
14. For instance, see James Rachels, Created from Animals: The Moral Implications of Darwinism (Oxford University Press, 1990).
15. Franz de Waal, Good Natured: The Origins of Right and Wrong in Humans and Other Animals (Harvard University Press, 1996), p. 89.
16. Peter J. Wilson, Man, The Promising Primate - The Conditions of Human Evolution (Yale University Press, 2nd ed., 2002).
17. Searle insists that his view in The Construction of Social Reality is "utterly naturalistic" - I take it he is claiming that rule following and the construction of institutions and institutional facts is something which is characteristic of human beings as products of evolution. See Searle, p. 300 in John Searles Ideas About Social Reality, ed. L.S. Moss and D.R. Koepsell (Blackwell, 2003).