Goldilocks Naturalism

Robert B. Talisse, Scott Aikin, Thomas Dabay

Naturalism is obviously a philosophical view. It is a view about what is, about how to inquire and what to inquire into, and about what we can meaningfully talk about. Moreover, because of these features, naturalism is also a view about how to do philosophy. So, not only is naturalism a philosophical view, but it’s also a metaphilosophical view.

Metaphilosophy is philosophy about philosophy. To do philosophy as a naturalist (that is, with the naturalist’s philosophical and metaphilosophical commitments), then, requires that one do two things. First, one must clarify naturalism as a philosophical view, and, second, one must then explain how those naturalist views inform how one does philosophy. The challenge for naturalists, however, is to philosophically account for how one has arrived at any of these commitments. Once taking on that task, naturalists find that they face a dilemma. They have arrived at naturalism either (a) as naturalists or (b) independently of their naturalism. If the first, then they have begged important questions, and if the second, then their view is self-defeating. Hence, there is a philosophical problem for naturalism.

In this essay, we will provide an overview of the controversies involved in clarifying naturalism and observe that most of the controversies arise as disagreements over how to find a just right balance between the demanding and the liberal interpretations of naturalism’s constraints. We are all out to find a Goldilocks version of naturalism. We will then turn to see how the self-defeat problem arises for naturalism, and we will pose our own version of a solution to the problem.

Describing Philosophical Naturalism(s)

Among philosophers, naturalism is generally taken to be the view that the sciences are in some sense authoritative over philosophy. Pithy as it may seem, this formulation suffers from three significant ambiguities: one concerning which academic disciplines are properly scientific, a second concerning which domains are properly philosophical, and a third concerning what sort of authority the former disciplines exert over the latter. For the purposes of this article, we will be bracketing any issues surrounding the second ambiguity, focusing instead on those surrounding the first and third. With this as our focus, we will come to see that there are in fact many disparities between the views that philosophers consider to be naturalistic, ranging from the conservative on one end of the spectrum to the liberal on the other.

Concerning the first ambiguity, philosophers often focus on four candidate disciplines for scientific status. They are, in one philosophically loaded order: physics, chemistry, biology, and the human sciences (psychology, anthropology, sociology, economics, and the like). Nearly everyone agrees that physics is a genuine science, but beyond this there has been little sustained agreement over the past century. Some will defend the extremely conservative view that physics is the only genuine science, while the other candidates are at best disciplines that might become genuine sciences in years to come. Others defend the slightly less conservative view that physics is a genuine science and that key portions of the other candidates are reducible to physics. To the extent that these other candidates can be reduced to physics, they are genuine sciences. However, some philosophers will point out that the demand for reducibility is largely one imposed by other philosophers and not by scientists themselves, which (if true) motivates the more liberal view that chemistry, biology, and maybe even the human sciences can be genuine sciences regardless of whether they are ever reducible to physics. At the most liberal end of the spectrum, most every department in a university could represent a genuine and independent science, despite the markedly different methodologies, commitments, and results of these various disciplines. The end result is that there is no clear-cut answer to which disciplines make up the sciences—hence, the first ambiguity.

When considering the third ambiguity, there is likewise no clear-cut answer as to what sort of authority the sciences hold over philosophy. Perhaps unsurprisingly, the first ambiguity provides a source of debate concerning how to address the third. But in addition to this, philosophers also disagree as to whether the third ambiguity should be addressed in an ontological, methodological, or semantic vein. Defenders of the ontological approach say that, for whatever disciplines we end up grouping under the heading “the sciences,” those are the disciplines that tell us what does and does not exist in the world. On the conservative end, if physics is the only discipline, then the things that exist will be fundamental particles, fundamental forces, voids, and nothing else. On the liberal end, if the human sciences are included as well, then cultures and the general will and the invisible hand of the market could exist in as full and literal a sense as particles and fields of force. As before, there is a range of intermediate positions, all surrounded by persistent disagreement.

Concerning the methodological approach, defenders here say that the sciences dictate what explanatory strategies are appropriate for philosophy. The default position is fairly moderate and involves looking at what methodology all of the sciences share. Usually, this shared methodology is understood as explaining phenomena in terms of their sensory effects or other experimental output and accepting the inevitable fallibility that results. A more conservative position is one that looks to one science in particular, usually physics, and it stipulates that its methodology is the only acceptable one. In such a case, the core of physics’s methodology is generally taken to be causal explanation, meaning that good explanations take the form of subsuming individual phenomena under the scope of universal causal laws. Finally, a more liberal position is one that accepts not only what is shared between different sciences but even the specific methodological differences between the various sciences as being genuine. Most commonly, such a position accepts not only causal explanation as it is practiced in physics but also historical and teleological explanation as they are often practiced in biology. Historical explanations work by tracing the oftentimes contingent or arbitrary events that actually do produce certain phenomena (instead of focusing on the causal laws that necessarily produce them), and teleological explanations work by surveying the purpose or functional role that certain phenomena play within larger systems.

Finally, on the semantic approach, defenders say that the sciences dictate what concepts can be meaningfully used within philosophy. Before we start on this approach, it will be useful to note a distinction philosophers commonly draw between descriptive, modal, and normative concepts. Jargon aside, the distinction is a familiar one—descriptive concepts concern what is or is not the case, modal concepts concern what necessarily or possibly is the case, and normative concepts concern what ought or ought not to be the case. On the semantic approach, one particularly conservative position is that since we use the sciences to describe what has happened, is happening, and will happen, the only concepts that philosophers can use are descriptive concepts. Nowadays, many philosophers see this view as self-defeating, because although the results of science may be descriptive, the way we get to these results is not. We end up with a somewhat liberal position if we focus on physics and recognize that the concepts of causality or lawlikeness are modal concepts (causes necessitate their effects and laws specify what must happen). We end up with an especially liberal position if we focus on biology as well and recognize that the concepts of purposiveness or proper functioning are normative concepts (a purpose is something that ought to be pursued and a proper function is something that ought to be performed).

Thus far, we have highlighted the great diversity of ways one can disambiguate the naturalist’s belief that the sciences are authoritative over philosophy. But might this have been merely a philosopher’s exercise, one with little if any practical payoff? To see that this is not the case, we need only consider naturalism’s relation to theism. Defenders of both views must face up to the question, “Does God exist?” The theist says “Yes, God does exist,” but the naturalist’s answer is less clear-cut.

If naturalists take the ontological approach, then they are liable to give the atheistic answer that “No, God does not exist” because no science testifies to his existence. But if we take the methodological approach, the strongest answer we are liable to give is the agnostic one that “We do not know whether God exists.” This is because we do not have a method for studying a transcendent being such as God, and so we have no way of learning anything one way or another about him.

Finally, if we take the semantic approach, we are liable to give one of two responses, neither of which is often recognized in debates concerning theism. On the one hand, we might be liable to say that the concept of a transcendent being is simply nonsensical because it has no role to play in the sciences. This would mean that the proper response to the original question—“Does God exist?”—is not to answer negatively or to answer agnostically; the proper response is to refuse to answer at all because the question is ill-formed. Surface appearances aside, the string of letters “Does God exist” has no more (and no less) semantic structure than the first stanza of Lewis Carroll’s Jabberwocky.

On the other hand, if this response seems too extreme, then we are liable to say that the people using the string of letters G-o-d cannot possibly mean what they think they mean by it; in fact, they must mean something that naturalists can make sense of in scientific terms. But once we redefine this string of letters to be more scientifically sound, we are just as liable to start answering the original question affirmatively as we are to answer it negatively or agnostically—maybe even more liable. Once “God” is redefined in naturalistic terms, then one cannot, as just a naturalist, determine whether or not the sentence “God exists” is true without all the requisite evidence. Naturalism doesn’t set the evidence in advance—that’s why there are experiments and observations. The arguments that take us to this conclusion are notoriously complicated, so we will leave these considerations here. The pragmatist tradition of philosophy is rich with just such arguments, and the works of Charles Sanders Peirce and William James represent two distinct poles of views about what the evidence supports about God’s existence, once one has redefined the terms. Curious readers should also visit Simon Blackburn’s works on quasi-realism and Huw Price’s on global expressivism, especially those mentioned in the Further Reading section at the end of this article. What is important is that, although naturalism is traditionally seen to be at odds with theism, this, too, is an ambiguous formulation and one that may not even be true.

The Normative Question(s) and Self-Defeat

At this point, you are probably asking yourself some form of the question, Which naturalism is the correct one? Which one ought you to accept, and which ones ought you to reject? The rejection question is the easy one to start with, as you almost assuredly think that some of the naturalisms mentioned are too conservative to be plausible. They cannot fit with moral beliefs (let’s say) or they do not take seriously a science you take to be successful. Similarly, you undoubtedly think that a number of the naturalisms mentioned are too liberal, whether because they open the door to teleological explanation of the sort so dear to theists’ hearts, because they take seriously a science you do not take to be successful, or for whatever other reason you see. In rejecting these forms of naturalism, you are also preparing yourself to answer the acceptance question. You now know that the naturalism you will accept is Goldilocks naturalism, the one that is not too conservative and not too liberal but just right.

Of course, not every reader will have the same sense of what is too conservative or too liberal, and so different readers will be driven toward different forms of naturalism in their search for the Goldilocks form. All of us naturalists recognize that our task is to find the Goldilocks form of naturalism, but we are now faced with the question, How do we go about finding it? How should we be searching for Goldilocks naturalism, and how will we recognize it when we see it? Unfortunately, this is the question that leads us to the brink of self-defeat.

As naturalists, the obvious answer is to say we go about our search naturalistically. Naturalism is a view about how we should approach philosophical questions in general, and we’re now asking a specific philosophical question, so what other response can there be but for us to revert to our naturalism? The obvious problem with this is that our naturalism is itself what is under question, so reverting to our naturalism is as clear a case of begging the question as is the theist’s insistence that God exists because he wrote the Bible.

But what alternative do we have besides reverting to our naturalism? If we do not approach our question naturalistically, we must be approaching it nonnaturalistically—that’s just how the prefix non works. And if we were to pursue this nonnaturalist approach, we would be grounding our naturalism in nonnaturalism. Theists and their ilk would end up being right after all, or at least closer to being right than we naturalists are. What we want is a third alternative besides naturalism and nonnaturalism, but again that’s just not the way non works. We are stuck with two alternatives, and neither alternative is acceptable, so naturalism must either be self-defeating or blatantly question-begging.

A Detour through Kant

Wait a minute, you are probably saying. Surely we’ve been too hasty in writing off the possibility of a third alternative. Our point about non is crafty, yes, but it is crafty wordplay and nothing more. There are many ways of playing with words, you might point out, and what we naturalists are ultimately concerned about is not words but the world—not the philosopher’s armchair but the scientist’s laboratory. To put these thoughts bluntly, although this conclusion that naturalism is self-defeating is the result of one way of speaking, we have yet to show that this way of speaking is the correct way of speaking.

This objection is well taken, and students of the history of philosophy will no doubt be reminded of how Immanuel Kant is traditionally placed in relation to the empiricists and rationalists that came before him. Kant acknowledges that the empiricists are thoroughly naturalistic, but he critiques them for begging most of the important questions against the rationalists. By Kant’s lights, the rationalists have the virtue of not begging questions against the empiricists, but the only reason they have this virtue is because they are willing to let their thinking go unconstrained by the bounds of any possible experience. Kant uses the word transcendent to designate this ph
enomenon of thinking beyond the bounds of possible experience and critiques transcendent thinking as being ungrounded, arbitrary, and possibly nonsensical. At this point, Kant finds himself in a position that is all too familiar to us: choosing between the idleness of begging the question and the excess of thinking that is no longer naturalistic.

One thing that makes Kant revolutionary in the history of philosophy is how he rejects this choice as a false dilemma. For Kant, there are not merely two ways of thinking: one that is within the bounds of possible experience and the other that is outside of this boundary. There is also a third way of thinking: one that consists in the thinker’s drawing the very boundary that defines the other ways of thinking. Kant calls this third kind of thinking “transcendental thinking” and argues that transcendental thinking is neither question-begging nor excessively nonnaturalistic. Armed with a similar notion, we too could escape our choice that leads us to the conclusion that naturalism is self-defeating.

However, arming ourselves with a similar notion is easier said than done. Kant himself offers little help, because understanding his notion of transcendental thinking has proven to be one of the most discussed issues in Kant scholarship. Alternatively, when we try to develop a similar notion for ourselves, we immediately run into the problem troubling Kant scholars. A physical boundary separates two or more physical spaces in the sense that the boundary touches each of these spaces while keeping all of these spaces from touching each other. That’s just what it means when we say a boundary is between these spaces. But a consequence of this way of thinking is that before we can draw a physical boundary, we must have physical access to the spaces being bound; in the simplest case, before we can use a pencil to draw a boundary line, we must first be able to fit the pencil between the spaces we want to separate. This is nothing more than a trite observation about physics and geometry, but from it we can pose a problem.

The problem comes when we stop taking talk of boundaries literally and try to take this as a metaphor for the mental realm. If we are to draw a boundary between empirical and transcendent thinking, as we do when we think transcendentally, then we must have access to both empirical and transcendent thinking. But for Kant and us naturalists alike, the purpose of acknowledging transcendental thinking is so that we can maintain that transcendent thinking is precisely the sort of thinking that humans like ourselves do not have any access to. With this, we are once again at the brink of self-defeat. The Kantian search for a third alternative might seem promising, but it is ultimately just delaying the issue.

Pragmatism to the Rescue

It is at this point in the dialectic where the most recent generation of pragmatists enters the picture. As a group, these pragmatists agree that there is significant debate as to what constitutes naturalism and that the shape this debate has taken in recent years has been an oscillation between more conservative and more liberal forms. They also agree that, given this debate, it is only natural for us to ask which is the correct form of naturalism, and they agree that this question would seem to raise serious internal problems for naturalists. Finally, under different banners1, they each take up the Kantian search for a third alternative to naturalism and nonnaturalism—but in a surprise turn of events, they each welcome the failure of this search!

To see why they greet failure so positively, we need only recap where our search has taken us. We began with a distinction between naturalism and nonnaturalism, but after finding that each leads to problems, we start looking for a third alternative. We note that Kant promises to deliver us this third alternative in his account of transcendental thinking; however, in trying to distinguish transcendental thinking from what Kant calls transcendent thinking, (and what we naturalists would call nonnaturalistic thinking), we realize that Kant’s concept of the transcendental is just as vexed a notion. This being the case, we have concluded that Kant’s transcendentalism does not offer us a viable third alternative.

Seemingly, this leaves us back where we began, with only two alternatives. However, the pragmatists will argue that the failure of our search provides us with a lesson about transcendent nonnaturalism—one we need only recognize to escape our problems. Since transcendent nonnaturalism is beyond the bounds of possible experience, it is literally unthinkable for beings like ourselves. We’ve already noted this point above, but we have yet to appreciate its full significance. This is because if nonnaturalism really is unthinkable, then we could never have been thinking about it while writing this article. Therefore, each time we have used the word nonnaturalism we have been speaking literal nonsense, for we are purporting to be thinking about something unthinkable. Similarly, you could never have been thinking about nonnaturalism while reading this article, so every time you have seen the word nonnaturalism you have been having nonsensical thoughts.

The only way to correct our pervasive non-sense is to excise nonnaturalism from our vocabulary, and once we have done that we are left with only one alternative, namely, the naturalistic alternative. This tactic does not directly answer the question about which form of naturalism is correct, but it does dissolve the worries we have spun out from this question. Once we accept the naturalistic alternative to answering this question (because it is the only alternative), we still have to face up to the worry that we are begging all the important questions. At a bare minimum, even if we naturalists were to be begging the question when we ought not to be, we would still be on better footing than theists who both beg the question and speak transcendent nonsense while doing so. But there are also good reasons to think that any circularity involved in our naturalism is of the benign type.

Since we are finite beings who are always already in the midst of inquiry, circularity is actually quite ubiquitous. Physicists have many conflicting interpretations of the fundamental principles of physics, but no one seriously thinks that the proper response to this disagreement is for physicists to stop doing physics until they can come to a non–question-begging agreement. True, physicists will be begging the question from each other as they continue doing physics, but they will also be producing new theories that can be of use to lay people and providing new evidence for philosophers of physics to work with. If history is any indicator, these new theories and pieces of evidence will help physicists come to an agreement on their current issues, but they will also spawn new issues with new disagreements, and this process will continue so long as people are inquiring into physics.

The same dynamic plays out in innumerable other sciences, and for naturalists, this should be all the evidence we need to accept this dynamic as it is playing out within our naturalisms. After all, anything less would be unscientific.

Philosophy, Dissolutions, and Intellectual Surprise

To many readers, our solution to the self-defeat problem for naturalism will seem to be akin to pulling a rabbit out of a hat. Philosophy regularly comes across as a kind of word-magic, certainly the kind of thing at which a self-described naturalist should blanch. In some ways, this is a regular response to philosophy generally: that its problems are abstruse and its solutions are a web of words. Isn’t it all really gibberish?

Here is why philosophy, at least when done well, is not gibberish. With philosophical work, we are reflecting on the concepts and terms we use to knock about in the world—how we thi
nk about what is, was, will be, might be, ought to be, and must be. Those thoughts aren’t gibberish, and so our coming to grips with what those thoughts are about isn’t either. That’s good philosophy. And what about our strategy for the self-defeat problem for naturalism? If we are right that naturalism also commits us to a view about meaning that makes the self-defeat challenge impossible to pose, then that’s a significant clarification and improvement of our situation. But this requires that we appreciate the revision of what we’d thought was a perfectly intelligible challenge and problem for naturalism. That’s a surprise, for sure, that when properly considered, there is not a consistency problem for naturalism. In some ways, it’s something that should be expected when one takes on the work of clarifying a notoriously controversial concept with a vexing challenge—that a particular instance of the concept renders the challenge inert. In the pragmatist tradition, this strategy is called “dissolving the problem.” One redescribes the challenge to a view in a way that offers not only a way to make the challenge disappear but returns us to using the concept comfortably as we had before. This, we think, is a good pragmatist aspiration for finding the Goldilocks version of naturalism.

Further Reading


  • De Caro, Mario, and David Macarthur, eds. Naturalism and Normativity. New York: Columbia University Press, 2010.
  • ———. Naturalism in Question. Cambridge, Mass.: Harvard University Press, 2004.


  • Brandom, Robert B. Articulating Reasons: An Introduction to Inferentialism. Cambridge, Mass.: Harvard University Press, 2000.
  • Kant, Immanuel. Prolegomena to Any Future Metaphysics that Will Be Able to Come Forward as Science. Translated by Gary Hatfield. In Theoretical Philosophy after 1781, edited by Henry Allison and Peter Heath. Cambridge, U.K.: Cambridge University Press, 2002.
  • Price, Huw. “Moving the Mirror Aside.” In Naturalism without Mirrors, edited by Huw Price. Oxford: Oxford University Press, 2011.
  • Putnam, Hilary. Realism with a Human Face. Edited by James Conant. Cambridge, Mass.: Harvard University Press, 1992.


  • Brandom, Robert B. Making It Explicit: Reasoning, Representing, and Discursive Commitment. Cambridge, Mass.: Harvard University Press, 1998.
  • Davidson, Donald. “On the Very Idea of a Conceptual Scheme.” In Inquiries into Truth and Interpretation, edited by Donald Davidson. Oxford, U.K.: Clarendon Press, 1984.
  • Kant, Immanuel. Critique of Pure Reason. Translated and edited by Paul Guyer and Allen W. Wood. New York: Cambridge University Press, 2009.
  • Macarthur, David, and Huw Price. “Pragmatism, Quasi-realism, and the Global Challenge.” In Naturalism without Mirrors, edited by Huw Price. Oxford: Oxford University Press, 2011.
  • McDowell, John. Mind and World. Cambridge, Mass.: Harvard University Press, 1996.
  • Price, Huw. Expressivism, Pragmatism, and Representationalism. Cambridge, U.K.: Cambridge University Press, 2013.

Robert B. Talisse

Robert B. Talisse is W. Alton Jones Professor of Philosophy, chair of the Philosophy Department, and professor of political science at Vanderbilt University. He is the former editor of Public Affairs Quarterly and a regular contributor to the blog 3 Quarks Daily. He has written several books including, most recently, Engaging Political Philosopy (Routledge, 2016).

Scott Aikin

Scott Aikin is assistant professor of philosophy at Vanderbilt University, where his research focuses on the theory of knowledge, ancient philosophy, informal logic, and the philosophy of religion. He is the author of numerous books and articles, his most recent book is Evidentialism and the Will to Believe (Bloomsbury Academic, 2014).

Thomas Dabay

Thomas Dabay is a lecturer in the Philosophy Department at Vanderbilt University. His research focuses on naturalism, inferential semantics, and pragmatist metaphilosophy. His most recent publication is “Why Peirce’s Anti-Intuitionism Is Not Anti- Cartesian: The Diagnosis of a Pragmatist Dogma,” in the International Journal of Philosophical Studies.

Naturalism is obviously a philosophical view. It is a view about what is, about how to inquire and what to inquire into, and about what we can meaningfully talk about. Moreover, because of these features, naturalism is also a view about how to do philosophy.

This article is available to subscribers only.
Subscribe now or log in to read this article.