Home » Library » Modern Library » Defending Naturalism as a Worldview: A Rebuttal to Michael Rea’s World Without Design

Defending Naturalism as a Worldview: A Rebuttal to Michael Rea’s World Without Design


Summary: Rea claims naturalism “is without rational foundation.” In rebuttal, this essay shows that adopting the “research program” of basic empiricism is universally appealing, and since naturalism as a “worldview” follows from adopting basic empiricism and applying it to the facts of the world, naturalism has a rational foundation. Rea argues to the contrary that naturalism is only a research program and never a worldview. This claim is disproved. Rea’s ultimate conclusion, that naturalism must abandon materialism and realism about material objects and other minds, because naturalism cannot “discover” intrinsic modal properties, is also disproved, by showing naturalists can discover them.

Rea’s Radical Thesis

In World Without Design Michael Rea argues that “naturalism’s status as orthodoxy is without rational foundation,” given that “naturalism is a research program” rather than a worldview.[1] Ultimately, from this Rea aims “to draw out some of the more interesting ontological consequences of naturalism” (225-26), which he claims are the abandonment of materialism, realism about material objects, and realism about other minds. Why must naturalists give up these things? Because they “have made a rationally indefensible decision to restrict themselves to the methods of science alone in developing their theories about the world” (170). And since abandoning such beliefs, he says, is too high a cost to justify remaining a naturalist, his intention is to “persuad[e] some naturalists to jump ship” (17, 173). Naturally, he concludes the book by arguing that Judeo-Christian Theism, or something comparable to it, offers the best chance to support materialism, realism about material objects, and realism about other minds (212-26): “I am entirely convinced,” says Rea, “that some supernaturalistic story … offers our only hope” of rescuing these beliefs (225).

Research Programs and Basic Evidence

Rea argues that “a research program is a set of methodological dispositions” (3), in particular those standing “prior to theory” (4). A research program encompasses the assumptions we make regarding what to trust as evidence before we inquire about what is the best evidence or method. So a research program is a set of assumptions about what evidence we are to regard as basic. As Rea says, “basic sources of evidence [are] sources that are to be trusted even in the absence of positive evidence in favor of their reliability.” He gives as examples “sense perception, reason, memory, [and] rational intuition,” admitting that some philosophers hold that “some of these sources are derivative” instead, meaning “they are to be trusted only after their reliability has been verified by evidence from the basic sources” (2).

Rea analyzes this to the conclusion that no research program can be adopted or abandoned because of evidence, since changing your beliefs about what evidence is basic cannot be based on evidence–for then the new type of evidence you are accepting or rejecting would by definition not be basic, since your trust or distrust in it is based on the other evidence. Therefore, the only reason anyone can have to change beliefs regarding what evidence to accept as basic is purely pragmatic, like picking a research program because it is “most attractive” or “most convenient” or “most irritat[ing to your] enemies” or “whatever” (6-7). You just want to adopt it, for no evidential reason. This leads Rea to declare:

Thus, if naturalism is indeed a research program … there is no basis at all for saying that it is the sort of program that everybody, or every intelligent or right-thinking person, ought to adopt. This, of course, is no special deficiency of naturalism. If it is a deficiency at all, it is one shared by every other research program. But … it goes some distance toward disarming the current presumption in favor of naturalism. (7)

So, according to Rea, one can only persuade people to abandon naturalism either by proving that naturalism is self-defeating (which he says he will not argue, though he points to Plantinga, and actually presents the argument: 182-92) or by giving pragmatic reasons to give it up, like “highlighting its unattractive consequences” (7). Therefore, the whole argument of Rea’s book depends on establishing that naturalism is, and is only, a research program. I will refute that contention. But first, Rea also claims there is no basis for declaring that every right-thinking person ought to adopt a particular research program. However, there is a universally attractive pragmatic ‘ought’ that suggests otherwise. Two arguments join to prove this, which I will relate below.

Everyone Ought to Adopt Basic Empiricism

(1.) As a matter of fact, everyone (hereafter meaning at least very nearly everyone) shares the pragmatic desire to achieve all their goals, many of which can only be reliably achieved by predicting the future as effectively as possible (which includes reconstructing ‘the past’ by applying the same principles retrospectively). Therefore, all evidence that must necessarily be taken as basic in order to develop reliable predictions of the future ought to be adopted by everyone. Therefore, contrary to Rea, I can say there is a minimal research program that everyone ought to adopt, even by appealing solely to pragmatic reasons.

What evidence must necessarily be taken as basic in order to develop reliable predictions of the future? ‘The future’ entails reference to experience. Everyone has goals that depend on future experiences in the sensory, cognitive, and emotive realms, at the very least. This entails a set of basic sources of evidence that everyone ought to accept. This qualifies as a ‘shared’ research program in Rea’s sense, but not an ‘individual’ research program. It proves everyone ought to accept a common set of basic sources of evidence, but not that they should abandon others. Each individual, though sharing a core research program with everyone else, could still have a different individual research program. But…

(2.) It follows from (1.) that everyone ought to have the goal not to be fooled into false beliefs which could undermine their ability to predict the future. Therefore, it is contrary to everyone’s interests to adopt as basic any kind of evidence that could be false. The only evidence that cannot be false is primary, uninterpreted experience, at least in the sensory, cognitive, and emotive realms. For it is impossible to deny that an experience exists when it exists. Such a denial is false by definition, stating that an experience is not present when it is present, thus asserting the contrary of what is. Everything else can be doubted.

For example, that I am experiencing a complex pattern of colors is an undeniable fact for me. Therefore, I must accept this as basic evidence. For it cannot in any intelligible sense be false for me. But the inference could be false that this pattern of colors indicates that I am experiencing a computer sitting in front of me on my desk. Therefore, I should not accept that as basic evidence, because the risk of so accepting it is inscrutable and unnecessary and contrary to my interests. The same follows for everyone.

Therefore, it follows that there is a pragmatic reason for everyone to share one and the same individual research program: that which takes all intrinsically undeniable evidence as basic, and only that and nothing else. And the only intrinsically undeniable evidence available to everyone is primary, uninterpreted experiences, at least of the sensory, cognitive, and emotive kind. Therefore, there is a pragmatic reason for everyone to accept this same set of basic sources of evidence, and to reject all others (though others can still be accepted derivatively). Rea is wrong to claim otherwise.[2]

Is this universally attractive research program naturalism? No–at least not as I understand it, not even in the most limited methodological sense, nor evidently as Rea understands it, since he equates naturalism with scientism, and science is largely comprised of derivative rather than basic sources of evidence (when analyzed in accord with (1.) and (2.)). But neither is this program, which I shall call basic empiricism, equal to intuitionism or supernaturalism, the only other research programs Rea entertains. Since he ignores the world’s most obvious and universally appropriate research program, his book is fatally incomplete.

That I have only established basic empiricism, not naturalism, will prima facie agree with Rea’s contention that we cannot say everyone ought to adopt naturalism–that is, as a research program. But Rea is the one calling naturalism a research program, not naturalists. So this is no problem for us, only for Rea. Real naturalists call naturalism (in the broadest sense, typically qualified as metaphysical naturalism) a worldview, not a research program. Naturalism is a total belief system regarding what exists, typically and appropriately contrasted with Traditional Theism, Philosophical Taoism, and the like. And we certainly can argue that everyone ought to be a naturalist in that sense: all we have to do is (3.) show that everyone ought to adopt the same research program (as I have just done for basic empiricism) and (4.) show how the available facts, in light of that research program, imply naturalism more than any other worldview.

This is why everything hinges on Rea’s contention not only that naturalism is, and only is, a research program, but that, as a logical consequence of this, there is no such thing as a worldview called ‘naturalism.’ For if the latter is false, then so is the former. But this is absurd. To deny the existence of a naturalist worldview is analogous to denying the existence of a Christian worldview. But if naturalism is a worldview, the props are kicked out from under the entire argument of Rea’s book.

Rea’s Misconception of Naturalism

Rea thinks naturalism can only be “a research program wherein one treats the methods of science and those methods alone as basic sources of evidence” (16; cf. 155, 173). I doubt any naturalist believes this, and none of Rea’s evidence definitely supports it, so his characterization is surely a straw man. There are two reasons for this conclusion.

First, I am quite certain that all intellectually informed naturalists regard most if not all of the methods of science as derivative rather than basic sources of evidence. For if you ask them why they trust those methods, they will all start listing evidence of their efficacy (Rea even quotes a naturalist doing this: 61), which one cannot do with basic evidence, as Rea himself says. Thus, Rea’s ‘naturalism’ is not the naturalism practiced by most actual naturalists.

Second, I am quite certain that all naturalists accept many more sources of evidence besides the scientific. They all accept historical evidence, for example, as well as direct personal experience outside the canons of scientific procedure. What sources of evidence naturalists actually accept as basic I can’t say. But I am sure they all accept at least the shared research program of basic empiricism (per above). Either way, naturalists accept many more sources of evidence than science, so Rea’s radical scientism is a straw man.

Rea ignores both these facts and concludes that naturalism does not exist except in his own sense, despite the fact that few naturalists really believe in such a thing. Part of his confusion might stem from misunderstanding naturalists when they say they believe in whatever science proves. Speaking for myself and several colleagues, when we say this, we mean the facts as science has so far proved them, not ‘whatever’ science might one day prove. So far, science has confirmed metaphysical naturalism and nothing else. So we feel comfortable referring inquirers to the findings of science when asked what we believe. But that would change if science proved the existence of the supernatural. Then we would cease being naturalists–yet we would still say we believe whatever science proves, since that then would include the supernatural.

I will address all this later. The point now is that many of Rea’s quoted sources are unclear as to which use of ‘the findings of science’ they really mean. I see a related confusion between a naturalist who advocates a particular method of inquiry because it is entailed by metaphysical naturalism, and a naturalist who defines naturalism as that method. These are not equivalent positions, yet Rea seems unconcerned with the distinction (e.g. 72). Rea also assumes that what naturalists say about evidence is about basic evidence, when few if any of the quotations he uses actually make such a claim.

I should also remark on a particular oddity of Rea’s procedure. Despite the vast quantity of very recent literature (in the note below there are more than sixteen books by naturalists about naturalism in the last ten years alone) Rea relies on two sources for most of his material about naturalism: John Dewey and W. V. Quine (both of them dead, a strange choice for someone attacking a live philosophy). Rea’s assurance that they are naturalism’s “main spokesmen” (16) struck me as odd. I have many colleagues in the naturalist community, yet few regard either Dewey or Quine as their representative. I certainly do not. To be fair, Rea does quote many other naturalists. But his obsessive focus on two old and uncharacteristic members of the current community is an example of the flawed nature of his procedure.

Does Metaphysical Naturalism Not Exist?

A worldview is a philosophical belief system regarding what exists, which informs nearly every aspect of life, including a core system of metaphysics, epistemology, and ethics.[3] Rea does not employ this term, but uses “coherent, substantive philosophical thesis” (22) in essentially the same sense, arguing that naturalism cannot be so formulated, and therefore all “alleged” naturalist worldviews “are nonexistent” because “all are either empirically refutable,” i.e. in theory, “or else vacuous, self-defeating, or otherwise unacceptable from a naturalistic point of view” (52, cf. 53-59).

Contrary to Rea’s claim, there are numerous self-proclaimed naturalists who describe what is unmistakably a worldview and call it naturalism, so that to deny they exist is tantamount to denying the earth is round. That naturalism can be developed as a fully-fledged and coherent worldview is made particularly clear in my forthcoming book Sense and Goodness without God: A Defense of Metaphysical Naturalism. But it can already be seen by surveying naturalist literature.[4] Even Christian critics accept naturalism is a worldview.[5] So Rea has a very heavy burden upon him if he intends to prove all these people fundamentally mistaken.

How does Rea argue that, despite all this evidence, naturalism as a worldview doesn’t exist? His case consists of four arguments. His first is that any naturalism that is “empirically refutable” is unacceptable because naturalists should not believe science can refute naturalism. We will discuss this crucial point in detail later. Second, Rea argues that phrasing naturalism as a metaphysical proposition is ultimately vacuous because it is unfalsifiable. But this argument depends on the first, so a refutation of the former will refute the latter. That leaves two other arguments.

Rea’s Flawed Finding Procedure

Rea argues that:

[T]he characterization of naturalism that is most faithful to the tradition, and the one that best explains both the similarities and the differences one finds among contemporary naturalists, is one which takes naturalism to be not a view but a research program (15) … [for] the methodological dispositions that unite naturalists under a common banner preclude the formulation of naturalism as a coherent, substantive philosophical thesis. (22)

Latent here is a fundamentally flawed procedure, and therefore a fundamentally flawed argument. First of all, Rea includes in his set of all naturalists those who merely defend methodological or epistemological naturalism (59-65). But by definition those people need not adopt naturalist metaphysics and therefore they need not be naturalists[6], nor must their views about anything be informative of naturalism as a worldview. They could be skeptics or positivists, and thus not endorse any worldview, or they could endorse a nonnaturalist worldview that was nevertheless consistent with naturalistic methods or epistemology.

For example, a devout Christian can adopt methodological naturalism, so long as she believes either that there are ways of knowing not accessible to science (such as faith), or that historical-scientific evidence can justify her belief in the basic tenets of Christianity. And many Christians believe both–personal religious experience is believed by many to be largely inaccessible to scientific study, and much of contemporary apologetics is premised on the belief that atheism is implausible in the face of scientific evidence of design and historical evidence of the resurrection of Jesus. Thus, Rea cannot base conclusions on naturalism as a worldview by limiting himself to aspects shared by all the people he examines. Such a procedure will inherently leave out of account the very evidence he should be looking for: shared theses about life, the universe, and everything, common to all metaphysical naturalists, constitutive of a “coherent, substantive philosophical thesis” called naturalism.

Yet even that procedure would be flawed. For if Rea sought to characterize naturalism as only those beliefs held in common by all metaphysical naturalists (and he doesn’t even do that) he might still end up with a worldview that no person actually believes in. Consider an analogy. Just because there is no actual person who believes only those things held in common by all Christians does not mean Christianity doesn’t exist as a worldview. There are hundreds of complete, coherent Christian worldviews, every one different, no one of them exactly corresponding to what is shared by all of them. Every one adds to some minimal set of beliefs important premises that cannot be left out of account, and there might not even be a minimal set of beliefs universally held in common by all Christians. The same could be true for naturalism.

This means Rea has not even tried to find naturalist worldviews. His method for locating them is inherently incapable of finding them. Thus, anything he has to say about what he finds to be ‘naturalism’ is irrelevant to actual naturalists, and his conclusion that there is no “coherent, substantive philosophical thesis” called naturalism is without basis.

Is Naturalism Inconsistent or Self-Defeating?

Rea next argues that “characterizing naturalism as a view rather than a research program inevitably portrays naturalism … as a self-defeating thesis” or a thesis “inconsistent with the core dispositions of the tradition” (16). The key word here is inevitably. If that claim is false, then Rea’s entire project unravels. There may indeed be self-defeating or inconsistent conceptions of naturalism. But Rea cannot pretend to have refuted naturalism if he does not address every form of it, and not every form is a ‘self-defeating thesis.’ Rea can eliminate inept versions with this approach, but competent ones would remain, and remain untouched by the argument of his book. Otherwise, Rea is like some village atheist who ‘proves’ Catholicism inconsistent and then declares that Christianity doesn’t exist.

For example, Rea charges that Alex Rosenberg’s definition of naturalism is ‘self-defeating’ because it includes “the repudiation of first philosophy” along with the premise that “the sciences … are to be the guide to epistemology and metaphysics” (65).[7] He implies that the second premise is an example of the very ‘first philosophy’ being denied in the first premise. But that is not a correct interpretation of Rosenberg’s meaning. Rosenberg means by ‘first philosophy’ the attempt to reason to facts prior to making any relevant observations pertaining to those facts. There is nothing incompatible between this assertion and Rosenberg’s assertion of a broad ‘scientism,’ wherein the sciences are to be our guide, not our sole trusted source of evidence (nor even a basic one). Why are they to be our guide? Because they present the only methods that work. This is not a conclusion arrived at by first philosophy. It is observed, making it an a posteriori discovery.

Rea also cannot assume Rosenberg is declaring science a basic source of evidence. Rosenberg is clearly not doing so. Nor does any naturalist as far as I know. Ultimately, there is nothing self-defeating in Rosenberg’s definition, even as Rea presents it. Rea does challenge the naturalist rejection of first philosophy by arguing that “it is still quite obvious that we would need some justification for believing that only empirical methods issue in justified belief” (64, sic). Therefore, supposedly, a naturalist cannot abandon first philosophy without contradicting himself. But Rea’s error here lies, again, in wrongly presuming that naturalists treat scientific methods as basic sources of evidence. In reality, we treat them as derivative and a posteriori, which means we do have justifications for our rejection of other methods: for we observe other methods to fail, or to be inconclusive.

Nor can Rea mean to attack naturalists here for adopting basic empiricism as their research program, because his claims of self-defeat, by his own admission, only apply to metaphysical or empirical theses, which is not what a research program is according to Rea. And though his book amounts to an attack on radical scientism, a research program probably no one on earth adopts, his book never addresses basic empiricism, where the only evidence that can be accepted as basic is that which cannot be denied. Basic empiricism is the default position every rational being must adopt, as at least part of their actual research program, since to reject it is to deny the undeniable. Therefore it is the addition of basic sources of evidence that demands justification (or at least explanation), not basic empiricism. So while Rea’s own research program might need first philosophy to avoid self-defeat, basic empiricism does not. And yet that is the real research program (or a part thereof) adopted by all naturalists.

Finally, Rea charges naturalism with self-defeat by following Plantinga’s “Evolutionary Argument against Atheism” (182-92). Though he soft-pedals the point (he “will not attempt to defend” its conclusions), he does say he believes the argument is sound (192), then merely cites some critics of the argument and moves on. This seems careless. For both of the ‘possible solutions’ to Plantinga’s argument that Rea proposes and then asserts false are, to me, obviously not false:

  1. There is no reason to suppose that a naturally evolved ability to perceive and reason would probably be fundamentally misleading about the facts of the world. To the contrary, reasonably reliable faculties are a highly probable outcome of such a selection process (and Rea is aware of this argument: 193). Plantinga’s famous examples to the contrary are actually highly improbable on evolutionary theory and thus do not argue against it.
  2. Even if such a development of reliability were improbable, there can be no doubt that we won that cosmic lottery. There is copious evidence that our faculties could not possibly be fundamentally unreliable. If they were, mankind could never have developed a modernized industrial civilization, much less have sent people to the moon and back. Maybe there are sentient beings who have evolved on planets throughout the cosmos with unreliable faculties. Even so, the fact that we alone advanced into our present situation is clear evidence that we, unlike them, got lucky. And this would not be unreasonable. There are over a trillion known galaxies, each with over a billion stars. Even if life arises only once per trillion stars (which means only once in every thousand galaxies) there must still be nearly a billion biospheres in the cosmos. So even if the odds against developing reliable faculties were an astonishing billion to one it would still be highly probable that one species got reliable faculties. From evidence of our success, it would be more than reasonable to conclude that this was us.[8]

Is Naturalism Just Scientism?

Rea says “my own view is that the methodological dispositions shared by naturalists are more central to their naturalism than the ontological commitments” (23) such that whatever science ever proves will simply become a part of naturalism. As a result, Rea rejects the application of the term “naturalist” to ancients like Lucretius because Lucretius did not adopt the scientific standards of modern-day naturalists (24). This is odd, since the ancients invented the concept of a ‘naturalist,’ in the term physicus or physikos (from physis, ‘nature’), meaning, in one common sense in use, someone who believes everything is a material and causal part of nature and there is nothing else. Lucretius was called a naturalist in the vocabulary of his own day, and there is little substantial difference between his worldview and that of modern naturalists. Sure, they differ on physical details and methods. Nevertheless, they are more similar than different, certainly in terms of those beliefs that now distinguish, for example, a scientist who is a naturalist from a scientist who is a Christian.[9] So again Rea’s finding-procedure fails to grasp what it really means to be a naturalist.

My own view, and that of my colleagues, is that naturalism is an ontological thesis first and foremost. I only accept the methods of science because they work and are explicable on my naturalistic worldview, whereas, say, the methods of magic neither work nor, if they did, would their working be explicable on naturalism. But science is also explicable on other worldviews, and thus could prove them, refuting naturalism. Naturalism could even be consistent with methods different from science. For instance, we could learn more about the universe by studying the records or using the mysterious devices of an advanced alien race rather than conducting scientific research. And unlike Rea, I am a naturalist, so I have an implicit authority to define who I am that Rea utterly lacks. He cannot claim I am not a naturalist any more than he can claim I am not a man.

Nevertheless Rea tries. He argues that “naturalists are united at least in part by methodological dispositions that preclude allegiance to views that cannot be called into question by science” (51) yet he never really demonstrates this. He shows that modern naturalists all accept science as authoritative. But the universal acceptance of the current methods and findings of science is a contingent, not a necessary fact of naturalism. Quoting methodological naturalists won’t rebut this, since they are not metaphysical naturalists, and only metaphysical naturalism is a worldview, not methodological naturalism–the latter can be a part of a worldview, but even worldviews other than naturalism (like some versions of Christianity).

Even if there are naturalists who think, as Rea says, “that developments in science could [never] force someone to reject naturalism” (51), that says nothing about the vast majority of other naturalists who disagree. Rea is aware his sources might simply be wrong, mischaracterizing their own worldview by failing to think things through or word themselves carefully (a problem ethnographers are familiar with, and Rea’s activity here is ethnographic and should be informed by the methods of that science). But he brushes this aside as if it were insignificant. Rea seems unwilling to admit that even if there are such naturalists as he alleges, an argument against them is still irrelevant to the rest of us. In short, Rea’s book is irrelevant to most naturalists.

A Brief Ethnography of Contemporary Naturalism

Rea makes factual claims about the beliefs of naturalists that can be empirically falsified. For instance, he says “it would be completely absurd … to think that empirical investigation could overthrow naturalism without overthrowing the scientific method itself in the process” (52), because “naturalism, whatever it is, must be compatible with anything science might tell us” (55), so “theses refutable by science cannot plausibly count as versions of naturalism because naturalism involves, first and foremost, a commitment to follow science wherever it leads” (63). Yet Rea knows that “the scientific method” contains no premises that entail the non-existence of gods, angels, immortal souls, magic, or almost any other supernatural entity, therefore discovering such things could not in itself “overthrow” the scientific method. But Rea’s point is that (or so he argues) naturalism is compatible with gods, angels, immortal souls, magic, and almost any other supernatural entity!

Obviously, this is wrong. Contrary to Rea (55), “a commitment to follow science wherever it leads” (63) logically could lead us to a new worldview. Even if naturalism contains the premise “follow science wherever it leads” it does not follow that it contains the premise “wherever science leads is naturalism.” A worldview may contain the seeds of its own demise: a faithful following of its own principles may lead us to abandon it in light of what we uncover. Indeed, a premise I believe to be inherent in most forms of naturalism is: “naturalism might be false, and if it is, it is your job as a naturalist to find out.” Again, Rea’s myopic obsession with a narrowly defined “research program” leads him to ignore or deny naturalism as a worldview.

A little lay ethnography bears this out. To find out what naturalists really believe, I conducted an informal survey of my colleagues in the naturalist community during the month of January 2003. I asked two questions of naturalists around the country:

  1. As a naturalist, do you believe it is logically possible (even if the current evidence makes this highly improbable) for science to disprove naturalism some day?
  2. As a naturalist, do you believe that if naturalism were overthrown by a purely empirical investigation, that the scientific method would necessarily be overthrown as well?

Rea’s arguments entail that the answers should be No / Yes. Yet I got the exact opposite results. Evan Fales, Keith Parsons, Charles Echelbarger, and Greg Klebanoff, all Ph.D.’s in philosophy and top defenders of naturalism, answered Y / N. So did Taner Edis, physics professor and author of the latest comprehensive defense of naturalism.[10]

Fales has written on related matters before.[11] He wrote to me, “I see no reason why we could not … scientifically discover supernatural entities–however you define ‘supernatural’–provided only that those entities causally interact with our environment in suitably ramified ways.” Parsons cited Schick in support of the same conclusion.[12] Echelbarger wrote, “If controlled experiments in psychic research could only be explained well on the theory that there are immaterial beings or forces, that would be a major blow against naturalism,” and far from such a discovery overthrowing the methods of science, “on the contrary, it would represent an extraordinary victory for the power of scientific method.” Klebanoff argues that science could disprove, for example, any version of naturalism that entails “the causal closure of the physical,” which means “the thesis that for any physical event x that has a cause there is another physical event y such that y is the cause of x.” That is a thesis science could refute, and doing so would not overthrow scientific method. He did note, and I concur, that there may be other ‘versions’ of naturalism.

Edis, too, cites parapsychology as an example: were it a successful scientific enterprise, he explains, “The bottom-up picture of reality which I consider central to naturalism would need to be radically revised.” He notes that “some might still want to shoehorn the result into ‘naturalism,'” citing Griffin as an example[13], “but I think it would be best to just abandon naturalism at that point instead of trying to redefine the term.” And doing so would not overturn scientific method. As Edis says, “we would quite likely have to seriously revise much of what we assume in doing science. Yet I think it would recognizably remain science,” whereas naturalism would not recognizably remain naturalism.

Other experts agreed. James Lippard has a masters degree in philosophy and cognitive science and has published papers related to our present question.[14] He also answers Y / N, writing:

Science could demonstrate the existence of supernatural entities, such as disembodied minds which survive death, materializing fish that fall from the sky, or other phenomena described in places like the Journal of the American Society for Psychical Research, the Journal of Scientific Exploration, or Fortean Times.

Though he notes that “if such things are demonstrated … they would thereby count as part of the natural order,” he nevertheless says this is moot “if naturalism is taken to be a claim about what types of entities exist, along the lines of physicalism.” In which case, “I certainly think it is possible for scientific investigation to disprove it, or at the very least show it to be highly likely to be false.”

Keith Augustine, whose master’s thesis at the University of Maryland was actually on the definition and defense of naturalism[15], not only answered Y / N, but gave me a slew of quotations of other naturalists who appear to agree. His thesis also presents copious arguments for and explanations of his belief that naturalism is scientifically falsifiable, even against critics. There, Augustine demonstrates the point that “some imaginable events, if they occurred, would be so suggestive of a supernatural cause that it would be unreasonable to hold out indefinitely for a natural explanation of them.” He gives several examples and various lists of criteria. In personal correspondence he writes:

I like Alan Lacey’s definition: “What [naturalism] insists on is that the world of nature should form a single sphere without incursions from outside by souls or spirits, divine or human …” My own definition is: “[N]aturalism is a metaphysical position about what sorts of causal relations exist–it is the position that every caused event within the natural world has a natural cause … [i.e.] naturalism is the position that everything that exists within nature is itself natural and is solely influenced by natural causes.”

One then need only define “natural” as against “supernatural,” and how you do so determines what sort of metaphysical naturalism you subscribe to. In contrast, Augustine says, if we define naturalism as Rea wants, as simply a belief in ‘whatever there is’ which is scientifically investigable, then we would be discussing “a trivial version of naturalism,” one that “is not in any way informative–it does not, for example, tell you what sorts of event to expect to happen or not to happen.” In other words, since science could demonstrate the existence of gods, angels, immortal souls, magic, and many other supernatural entities, to say that naturalism is the belief in whatever science demonstrates is to make naturalism vacuous, reducing it to a useless synonym for ‘scientism.’

In support, Augustine cites two well-known defenders of naturalism:

Graham Oppy: “[A]ll that the naturalist needs to contend is that, as a matter of contingent fact, there is no good reason to believe in any naturalistically unacceptable entities. (Many naturalists–consider, for example, Lewis and Jackson–allow that there could have been spooks and gods, and that we could have had good reason to believe that there are spooks and gods; however, they insist that, as a matter of fact, we don’t and there aren’t.)”[16]

Ernest Nagel: “It is possible, I think, to conceive without logical inconsistency a world in which disembodied forces are dynamic agents, or in which whatever happens is a manifestation of an unfolding logical pattern. In such possible worlds it would be an error to be a naturalist.”[17]

So we have seven expert naturalists saying exactly the opposite of what Rea contends, plus at least three others who appear to. Rea even ignores his own evidence here: though he presents Alex Rosenberg’s definition of naturalism as including Darwinism (65), this means Rosenberg would probably abandon naturalism if science disproved Darwinism (which it could). Finally, in personal correspondence to me, noted philosopher Paul Draper also answered Y / N, and though he is agnostic about naturalism, he is a sympathizer who is soon to go on record defending the scientific falsifiability of naturalism.[18]

Among lay naturalists, too, opinions are the same. Jeffery Jay Lowder answered Y / N. He has defended naturalism in a public debate with theologian Phil Fernandes[19], and is co-founder and president emeritus of the Internet Infidels, a non-profit organization dedicated to running the Secular Web for the very purpose of “promoting metaphysical naturalism.” Computer scientist Steven Carr concurs, declaring “even the crudest scientific methods (naked-eye observation) could prove that there are supernatural beings.” Darrel Henschell agrees science could refute naturalism, though he wasn’t sure what effect that would have on scientific methods.

Only one lay respondent, Charlie Clack, answered N / N, only because, as he put it: “I don’t see any way to test supernaturalism. I can’t imagine any evidence that would tend to prove or disprove it.” But as we have seen, Clack’s lack of imagination is at fault here, not an ideological commitment to the dogma alleged by Rea. Hence Sue Strandberg answered Y / N, because she does have a fairly clear idea of what a supernatural fact would be, and knows science could discover such things. She points to the worlds presented in the movies Ghost Busters and Star Wars. In the former, Dan Akroyd’s character “could measure the ghost ‘residue’ and make successful predictions of ghost activity,” and in the latter “the Jedi knights could learn … how to use ‘the Force'” which was comparable to a “technology,” and “by being able to repeat demonstrations of the Force,” they “could at least in theory have tried to figure out how and why it worked the way it does” and that is, quite simply, “science.” Yet neither world would be one naturalists could believe in and still call themselves ‘naturalists’ in any intelligible way.

Strandberg also gave the most eloquent response to the second question:

Science is a method of trying to gain reliable knowledge by searching for consensus among competent observers who commit themselves to using procedures which seek to eliminate personal bias. I see no reason why if ESP, angels, ghosts or God exist that scientists could not investigate the deeper laws behind such phenomena, given a new and better perspective.

We now have at least twelve experts and five laymen (counting myself) all against Rea.[20] Only one layman came close to Rea’s assumptions, yet even he did not support them. So it should be clear by now that scientism is not incompatible with supernatural beliefs, nor is naturalism in any sense synonymous with it.

This ethnography of naturalist beliefs, although by no means thorough or rigorous, still proves that many, if not most naturalists do not believe the things Rea says they do. Which means the naturalism that Rea’s book challenges is not the naturalism popularly defended today. Perhaps Rea would argue that all the naturalists I have surveyed must be adhering to a view that is inconsistent or self-defeating. But we have already seen this charge is baseless.

The Natural-Supernatural Distinction

Every form of metaphysical naturalism asserts that the natural world is all there is, at least within the closed system of our universe, and that nothing supernatural exists.[21] This has been a primary meaning of the term for at least three hundred years. Lexica in the 18th and 19th centuries defined naturalism in various ways as “a philosophical system that saw man living solely in … a universe devoid of transcendental, metaphysical or divine forces” or as “the system of those who find all primary causes in nature,” set in contrast with “spiritualism.”[22]

Some naturalists create a middle category, of the nonsupernatural and nonnatural, but I will include this among ‘natural’ things, so a dichotomy can be drawn between the natural and the supernatural allowing naturalism to be defined as above. But either way, it can be summed up as the denial of (or disbelief in) the supernatural. Methodological naturalism says the same, though not about what exists (it is not a worldview), but merely what can be studied by science. As noted earlier, a Christian Theist can be a Methodological Naturalist without contradiction. So our only concern here is with the worldview of Metaphysical Naturalism. That is all ‘naturalism’ shall mean from here on.

Rea rightly points out that the natural-supernatural distinction is not easy to draw (54-55). He concludes from this that naturalism cannot be defined this way. But he doesn’t consider whether the way this distinction is drawn might be a demarcation point for different varieties of naturalism. Different kinds of naturalists will often regard different kinds of things as supernatural, just as different kinds of Christians will regard different kinds of things as essential or contrary to Christianity. Even so, there ought to be some minimal set of criteria that is common to all demarcations of the supernatural by all naturalists, or else the word may indeed be a bit of meaningless jargon.

In many cases, disagreements about what counts as supernatural stem from the confusion, muddled thinking, or mere lack of imagination of one or another naturalist, and naturalism can easily be mischaracterized even by its own adherents (a fact Rea concedes: 53). Setting that aside, we still see a common factor among naturalist identifications of the supernatural, which derives from the etymological basis of the word ‘natural.’ Since ancient times the natural has been distinguished from the artificial (the ‘products of man’), like technology, culture, and law. Of course, naturalists do not regard the artificial as supernatural, but as natural, insofar as such things derive entirely from and are entirely dependent on the natural. Since man is a natural animal, everything man does is a natural thing, no less than a bee hive is natural, yet also an artifact of the activity of bees. But as soon as artifice breaks free of its origin in or dependence on nature it becomes supernatural.

In my forthcoming book I define ‘nature’ as “a non-sentient universe, with all its properties and behavioral principles,” such that “anything involving sentient beings and powers beyond nature” or “anything purely mental” excludes naturalism. By “beyond nature” I mean “not grounded or formed naturally” or “existing prior to or independently of nature.” In short, pure or uncaused mental entities, and the effects of these entities, are supernatural, as are all events involving pure mind over matter. Even a natural mind can produce a supernatural effect if it can do so directly, ignoring every other aspect of nature but the mind’s will to affect it. All such entities and effects are purely artificial, not caused by nature, nor dependent on nature to happen or exist. My brief ethnography of naturalists helps bear this out as a view probably held by all naturalists, providing a successful natural-supernatural demarcation for defining ‘mere naturalism.’

Obviously, my definition rules out a Traditional God, for clearly He is a supernatural mind (neither created nor limited by, nor dependent on nature in any way). Since all naturalists are nontheists, this stands as one major confirmation of my criterion. As noted earlier, there are some who try to naturalize god, but this is not at all common or commonly approved, and such people certainly are using the term ‘naturalism’ in a way that is trivial or meaningless, rather like someone claiming to be a Christian while asserting that salvation has nothing to do with Christ.

Rea’s own demarcation also confirms mine. He argues that supernaturalism (which excludes naturalism) is a ‘research program’ wherein religious experience, by which he means “an apparent direct awareness” of a divine mind, communication, or testament (68, 221), is accepted as a ‘basic’ source of evidence (67). Such a research program entails believing the metaphysical thesis that supernatural causation exists. For religious experience could only plausibly be a basic source of evidence if at least one supernatural mind existed. Religious experience could come to be trusted as a non-basic source of evidence, so that science (even naturalism, if one limits the ‘divine’ to natural entities) can in principle accept religious experience as evidence[23], but this is not what Rea means, since he is talking about basic, not derivative evidence (68). And though one can imagine natural mechanisms by which religious experience could be a reliable source of evidence for the facts of the universe, no such mechanism can be imagined where religious experience would be acceptable as basic evidence (i.e. trustworthy in and of itself, without other evidence in support of it). Only supernatural minds or powers in my sense would make such a thing conceivable.

Another critic of naturalism, Mark Steiner, confirms this.[24] He argues that we can refute naturalism if we prove that nature is anthropocentric, meaning that nature anticipates or somehow ‘makes special’ man’s place in the universe–or more precisely mind’s place in the universe. Therefore, he believes naturalism entails that the universe is not anthropocentric. This agrees with my criterion, since, just as Steiner himself argues, the universe could only be anthropocentric in his sense if mind was necessarily prior to nature, and not a contingent product of nature as naturalists all believe. Another theist, David Foster, has argued the very same thing.[25]

Keith Augustine surveys many definitions of naturalism and the supernatural in the philosophical literature and comes to a conclusion that is very close to my own.[26] He argues that the distinction is about what sorts of causes exist, whereby a supernatural cause is one that is not caused by any physical or spaciotemporal fact of the universe and yet “exhibits apparently purposive or intelligent behavior.” Augustine’s definition entails that anything not caused by nature, but caused by a mind independent of nature, is supernatural. So my criterion fits his.

Unprompted, Sue Strandberg gave almost the very same answer. She said that, for her, something is supernatural:

If it both (1) used laws different than those which apply to the familiar form of the universe we share in common experience–i.e. matter, energy, time arrow, etc., and (2) was directly related in a significant way to the existence and direct causal power of thought, desire, personhood, Mind, Will, Intention, Emotion or any combination thereof.

The fact that she had exactly the same insight as several other naturalists confirms my point: this is probably what all naturalists generally agree is supernatural. Klebanoff, too, gave an example involving scientifically acceptable proof of a “nonphysical mind or soul” that could and would disprove “the causal closure of the physical” and thus provide adequate grounds to abandon naturalism. Thus ‘mind’ divorced from physics looms prominent in his conception of supernature too. Therefore, there is a widely accepted criterion after all.

When we go to Webster’s College Dictionary and find naturalism defined there as “the belief that all phenomena are covered by laws of science and that all teleological explanations are therefore without value” we see the same pattern. First, even common dictionaries agree that naturalism is a belief about what exists, a belief that specifically excludes pure teleological causation–in other words, the wills of gods, angels, spirits, and us, insofar as they are effective without or in defiance of scientific laws. In other words, again, the supernatural is any example of mind prior to nature. Proponents of supernaturalism say the same thing, e.g. William Dembski writes that “the fundamental divide” between theism and naturalism “is this: Is reality fundamentally mindful and purposive or mindless and material?”[27]

In conclusion: substantive metaphysical naturalists agree that naturalism entails the view that mind is dependent on nature, not the other way around, and that nature precedes mind: ontologically, metaphysically, and physically (i.e. causally). So if any mind or purely mental property precedes nature ontologically, metaphysically, or causally, then supernaturalism is true. And this holds regardless of whose mind is involved: whether there is only one supernatural mind (like God), or if all or any beings have supernatural minds (or natural minds with supernatural powers), or if the universe itself is in some way fundamentally mental (e.g. the cognocentric properties of thought, will, desire, value, intellection, etc., precede or ground the natural universe in some way). If any of these statements is true, then naturalism is false.

The Discovery Problem

After deploying the above tactics, Rea moves to his argument that “intrinsic modal properties seem not to be discoverable by the methods of science” (16). He calls this the Discovery Problem. “Having persistence conditions,” for instance, “is part of our concept of a material object” (82), but science (because it is empirical) can’t justify belief in them, according to Rea, therefore it can’t ground materialism (etc.). Even if Rea is wrong about naturalism being a research program, this argument should still concern naturalists since it might be reformulated against basic empiricism. Also, since this is the core focus of Rea’s entire book, it deserves special treatment.

Rea argues that “we cannot see, hear, smell, taste, or touch the fact that a rabbit cannot survive being flattened by a steamroller” because “we cannot perceive the fact that the resultant mess is not the same object as the rabbit, just in a different form” sic (105). Of course, every naturalist would say it was the same object–all the same material is there–it just isn’t a rabbit anymore, since a rabbit is by definition a particular arrangement of material, and this material is no longer arranged that way. For instance, there is no longer a rabbit’s circulatory system.[28] Thus we actually can “see” the fact that the rabbit did not survive the steamroller. We can also “touch” and, such as by applying a stethoscope to where its heart should be, even “hear” the fact that the rabbit is no more.

So it isn’t clear what Rea expects here. Is he objecting to the theory that rabbits will never survive steamrollers, on the contrary theory that rabbit-arrangements of matter might bounce back into shape or crush a hole into the ground and thus resist being rearranged? That isn’t what Rea seems to mean. Is he objecting to the theory that matter no longer in a rabbit-arrangement will not have the properties entailed by that arrangement, like a heartbeat, on the contrary theory that ‘the resultant mess’ might still have the properties of a rabbit, like a heartbeat? That isn’t what Rea seems to mean either. And in both cases scientific procedure proves these alternative hypotheses false: all the evidence confirms that a rabbit-arrangement of matter is necessary for rabbit-arrangement properties to exist, and that a rabbit-arrangement of matter will not survive being steamrolled. For it is never any other way and the alternatives are thoroughly falsified.

So what does Rea mean? It’s hard to tell. The only alternative he offers is an explanation for the destruction of the rabbit via extrinsic modal properties, rather than the intrinsic ones essential to material objects. So it seems as if he is arguing that science cannot adjudicate between the two and therefore cannot justify belief in one over the other (e.g. 11). He seems to accept that science can prove that there are objects and that those objects have properties, and he also seems to assume those properties must be either intrinsic or extrinsic. For if there is any other kind, he doesn’t say what it would be (160). And since he also claims to ‘prove’ that naturalists are committed to constructivism (15, 156, etc.), it seems the only issue left is whether modal properties are intrinsic or extrinsic, not whether they exist. For the fact that modal properties exist seems empirically obvious. At least, I cannot imagine any plausible explanation of human experience that excludes them.

Given this conclusion, is science really incapable of demonstrating that modal properties are intrinsic rather than extrinsic? To be extrinsic in Rea’s sense, properties require “relations between a … mind … and the stuff that supposedly constitutes the objects in question” (159) such that, absent the activity of human thought, all objects are intrinsically “unidentifiable” (12). In contrast, to exist, intrinsic properties do not require any relationship to a mind. Rea offers ‘being a statue’ as an example of an extrinsic property (13), since there is no intrinsic difference between a statue and a rock carved by natural forces. Rea does not address the fact that the two differ in physical-causal history, so he seems to assume that properties are to be taken e tempore, and if that assumption is defensible (I am skeptical), then he would be correct that there is no difference here: the distinction between statues and rocks would then be entirely an arbitrary human convention. But even granting that, it is unclear how such naming rituals could ever make modal properties explicable. The choice to call a rock a statue certainly has causal affects on human behavior, but it makes no observable changes in the behavior of the natural world. Nature treats rocks and statues the same. But she doesn’t treat live and crushed rabbits the same. Thus, intrinsic modal properties (IMP’s) explain our observations far better than extrinsic modal properties (EMP’s) seem able.

Consequently, Rea admits that scientists probably think they believe in IMP’s by an ABE, or Argument to the Best Explanation (105).[29] Rea considers in detail only two ABE’s for IMP’s: Crawford Elder’s arguments from proper function (106-27) and microstructure (128-38). I will not address the former since I see no reason for naturalists to appeal to proper function. One need only appeal to function to explain all the world’s phenomena, and Rea accepts that science can prove the existence of functions (113).

But Rea doesn’t do justice to Elder’s argument from microstructure. It seems a simple matter to show how Elder’s world-models entail IMP’s in ways that are explanatorily superior to models that exclude them. Besides, one can argue, e.g.: IMP’s are entailed by any universe that exhibits physical structure and causal consistency; all observations so far suggest that our universe exhibits physical structure and causal consistency, and no observations so far suggest otherwise; therefore IMP’s probably exist. No argument of comparable force can be made for EMP’s. And while we have detailed models that thoroughly explain and predict IMP’s, we have no such models for EMP’s. The latter therefore, at present, have no scientific viability as an explanation. Likewise, the thesis “modal properties are intrinsic” makes the things we observe highly probable, whereas the thesis “modal properties are extrinsic” does not make current observations nearly as probable (since many other outcomes are possible on the theory that the properties of objects are in any way mind-dependent–for instance, the world could behave like a cartoon).

Likewise, while “modal properties are intrinsic” requires only the existence of physical structure and causal consistency, two things whose existence is highly plausible and evidentially supported, “modal properties are extrinsic” requires an inscrutably complicated set of unproven assumptions and mechanisms merely to explain why we observe what we do rather than something else–e.g. Why do we all observe the same things the same way, even before we learn why they are the way they are? Or, if there is no one else but us, how do our minds subconsciously design a world of such vast complexity and consistency far exceeding our conscious abilities? How do minds obtain and employ the power to change the universe by conceptualizing it in the first place? Etc. In like fashion, IMP’s explain more facts than EMP’s seem capable of–like why we can’t change the world by reconceptualizing it, or why some people’s conceptions (even sometimes our own) are demonstrably false. So Rea’s position seems indefensible to me.

Consider persistence conditions. Persistence is the necessary conjunct of the naturalist’s commitment to causal closure, since persistence entails that nothing changes without a cause, and that, all else being equal, specific causes always effect specific changes. Basically, if metaphysical naturalism is true, then every change must have a cause; every change open to thorough examination so far is observed to have a cause; therefore, naturalism is confirmed by observation. This holds even for random phenomena at quantum scales, since they have a cause in the nature of quantum systems and fields. The existence of a mere predictable propensity is causal, so that natural laws or quantum fields, even if immaterial, nevertheless ’cause’ things to be one way rather than another. The same reasoning applies to material objects: if materialism is true, then persistence is true; everywhere we look we observe persistence (at least until observable causes bring about predictable changes, which is what persistence conditions entail); therefore, materialism is confirmed by observation.

Naturally, we have much more evidence than this, for both naturalism and materialism, all converging on the same conclusion, and no evidence challenging it. Rea seems unable to touch this, so he tries to bypass all this evidence by falsely characterizing naturalism as scientism and thus attacking its methods rather than its hypotheses. Consequently, Rea’s accusation that we cannot solve the Discovery Problem, even if it defeats his false notion of naturalism as a ‘research program,’ does not much touch naturalism as a metaphysical thesis, and he never claims it does.

But I don’t think he is successful in attacking even his straw man. Rea argues that we cannot justify belief in IMP’s because “there is no naturalistically acceptable basis for thinking that reflecting upon conceptual or conventional truths is a way of acquiring information about the world’s intrinsic modal structure” (87) and that even “scientific methods [cannot] generate evidence in support” of any belief in IMP’s. But Rea seems unaware of the nature of scientific methods, which necessarily combine “reflecting upon conceptual or conventional truths” with basic empiricism to produce a derivative source of evidence.

The Scientific Discovery of Intrinsic Modal Properties

If we accepted Rea’s own premise that the methods of science provide ‘basic’ sources of evidence, even then all we would need is to show that the methods of science can justify belief in IMP’s like persistence conditions. Now, to engage the scientific method means literally this: we adduce an explanatory model, deduce what follows from that, then observe what actually happens, arriving at a conclusion about the nature of things through inductive logic. This method entails that “to do science” we not only can but must imagine facts and conditions obtaining in the world, reason from them to conclusions about what would be observed if that model were true and what would be observed instead if that model were not true, then look to see which observations bear out. If the former, then we have ‘confirmed’ to some degree the imagined facts and conditions–even without observing them directly. This is how we discovered atoms, for example, long before we could ‘see’ them in any sense of the word. This is how all science is conducted.

Rea doesn’t seem to acknowledge this. For example, he oversimplifies the nature of Elder’s hypothesis (134), ignoring, for example, the role of falsification criteria, acting as if hypotheses only have verification criteria. Consequently, Rea fails to notice the real power of arguments like Elder’s for solving the Discovery Problem. So let’s apply genuine scientific reasoning to one of his own examples. Rea asks:

Why think that where we find matter arranged both in the way distinctive of H2O molecules and in the way distinctive of mere collections of subatomic particles we find something that is essentially an H2O molecule rather than something that is essentially a collection of subatomic particles? (104)

This question can be taken two ways: either he means that a collection of subatomic particles arranged as H2O might also be a ‘mere collection of subatomic particles,’ or that there might be no difference between two collections of particles, one arranged as H2O and another arranged differently (presumably at random or with no discernible structure–as implied by the word “mere”). Rea cannot be asking the former, since then the word “mere” is gratuitous and inappropriate: for no one denies that H2O is a collection of subatomic particles, and everyone agrees that what distinguishes it from other (‘mere’) collections is its physical arrangement. So Rea must be asking the latter: Why should we think that a collection of subatomic particles arranged as H2O is any different from any other collection? The obvious answer is: geometrically they are observably and categorically different, just as a square is observably and categorically different from a triangle. And science gives this answer an even more concrete basis.

Consider the scientific fact that the physical structure of H2O entails that microwave ovens can cook food.[30] This modal property of microwave ovens is so widely and routinely observed that it cannot be denied. For some of us, our daily lives depend on it being true. Thus, there is no plausible way we can pretend that these ovens don’t have this property, or that the presence of water in an object doesn’t confer to it the property of being cookable in a microwave. This is easily demonstrated by microwaving substances with different quantities of water in their composition: those with no water will hardly heat at all, while those with a lot of water will heat the quickest (all else being equal, e.g. discounting the effect of heating any humid air around the object).

Rea would say, if I understand him right, that we have no scientific evidence that these properties are intrinsic rather than extrinsic. But that is untrue. For we know why microwaves can cook substances with H2O: (1) the physical, geometric structure of H2O is peculiarly triangular, in a way many other molecules are not; (2) this arrangement is lopsided: two light atoms bound to one heavy one, producing a mechanical fulcrum around which the atom can physically ‘spin’; and (3) this produces a special physical arrangement of charged particles (electrons and protons). The latter is especially important: due to the nature of molecular bonding, there is an excess of negative charge physically located near one nexus (where the oxygen atom is located) and an excess of positive charge opposite (since the oxygen atom draws inward the electrons of the two hydrogen atoms). Consequently, H2O molecules are magnetic dipoles, and that means they will physically rotate to align with a magnetic field. If such a field alternates in polarity, the molecule’s rotation will continue back and forth, and the resulting friction (for lack of a simpler word) heats up the surrounding matter. A magnetic field is rarely strong enough to have a noticeable effect, but a shower of electromagnetic particles called microwaves, of a size precisely matched to the physical size of the structures in an H2O molecule, will produce magnetic effects that rotate those molecules quite rapidly. And that is how microwave ovens cook food.

Now we must ask Rea: Can he explain this without appealing to IMP’s? I don’t see how. The entire success of microwave cooking technology utterly depends on the physical arrangement of subatomic particles in H2O. If that arrangement were any different, H2O would not rotate, and microwaves would not cook our food. Therefore it seems undeniable to me that science has proven that there is a physical difference between H2O and other collections of particles, and that this difference is essential to the unique properties of H2O, and essential to the function of microwave ovens. I cannot fathom how microwave technology could ever have been predicted, developed, or perfected without this explanatory model.

It follows that the physical structure of H2O is absolutely indispensable to science, not to mention modern industry. This physical structure is demonstrably necessary for water to have the property of cookability in a microwave–at least, there is no plausible argument, consistent with current scientific evidence, how this property could survive the removal of this structure. So when Rea asks if there is anything empirically verifiable that an object exhibiting modal properties (e.g. the property of being cookable in a microwave) “could not possibly exist without” (103), we can easily answer him here: an object with the property of being cookable in a microwave cannot exist without the physical, verifiable structure of a magnetically dipolar molecule. There is no reason to believe otherwise–for no evidence has suggested any other possibility.

I doubt Rea can provide any sort of explanatory model that accounts for the operation of microwave ovens by appealing solely to EMP’s. In other words, can he ignore the role of physical structure and the causal consistency of this system and yet still produce the exact same predictions? Or produce more efficient microwave ovens? Could he ever have even invented a microwave oven? There is no reason to suspect that models incorporating solely extrinsic modal properties for H2O would ever have led to the prediction of the effects now exploited by microwave oven technology, much less to our present ability to predict the physical differences between efficient and inefficient microwave ovens. Therefore, Rea is wholly out of order here. And this is just one example for one material structure–there are hundreds of properties derived from the structure of water that have been confirmed in hundreds of comparable ways, and there are millions of atomic, subatomic, molecular, and other physical structures that have likewise been confirmed in the same kinds of ways as these.

If the properties of H2O are extrinsic, then its structure is irrelevant, and therefore science is entirely wrong about why microwave ovens work. That seems immensely improbable. In support of this, it is immensely improbable that wholly unrelated discoveries (e.g. those made in quantum mechanics vs. those made in chemistry) should converge on a common truth, unless there is a truth out there apart from human concepts. And many well-proven scientific theories entail IMP’s. For example, both Darwinism and the Big Bang Theory entail that things happened when there were no minds to conceptualize them. So IMP’s had to exist if these theories are true. Indeed, I can’t fathom how a constructivist could explain how minds came to be, since only modal properties could cause something as complex as a mind to develop, yet that could never happen if modal properties require minds. An EMP universe is thus physically impossible. Science, in contrast, explains how minds came to be through Big Bang Theory and Darwinism, which makes for a powerful ABE for IMP’s. Therefore science, but even more certainly naturalism (as an actual worldview), can justify belief in IMP’s.

Rea’s own expectations confirm this. He allows that we are automatically justified in such a belief unless “accepting it involves substantially greater metaphysical commitment” than the alternative and basic evidence does not support it (98). Basic empiricism entails that basic evidence cannot support any beliefs about the world outside our mind–only derivative evidence (theories based on our basic evidences) can do so. But Rea requires both criteria to be met. So he claims that IMP’s are “mysterious,” and therefore more ‘metaphysical commitment’ is needed to believe in them than in EMP’s (99). But IMP’s are not at all mysterious. As we saw for microwavability, they are often fully explicable and evidentially verified. Far more mysteries spring from the attempt to explain the same phenomena with EMP’s. For instance: How did minds come to be? How did they acquire their power to create EMP’s? How do they effect that power? Why does that power work only the way we observe it to and not in more expected ways? For example, why does the world behave as if IMP’s exist, rather than, say, like a cartoon? And so on. Science resolves all these questions beautifully, but only by appealing to IMP’s. Thus, on Rea’s own expectations we are justified in believing in IMP’s.

Crawford Elder’s ABE for IMP’s

Rea does not address the above line of reasoning, but only Crawford Elder’s version of the ABE from microstructure for IMP’s, hereafter ME. Rea’s rebuttal to ME consists of two principal arguments: (1) Rea claims that ME fails to support “realism about individual organisms” (136-37) and (2) Rea claims that ME requires an unnaturalistic trust in the intuition that “what is metaphysically necessary and what is necessary simply as a matter of natural law” are the same (138). Both objections suffer from the defect of assuming IMP’s must be essential properties, when in fact neither modal nor intrinsic properties must be ‘essential’ properties in Rea’s sense (101-02), so attacking essentialism does not really refute Elder’s microstructural argument for IMP’s. But even granting essentialism, there are still defects in Rea’s case.

As for (1), Rea allows that ME “may support realism about biological organs (like hearts).” But organs are individual organisms, just on a different scale. An organ is an organism of cells, while an organism is an organism of organs. There is no relevant difference here. So if ME justifies realism about organs, it does so for organisms. Just as the interlocking symbiotic relationship of a collection of cells is essential to an organ’s modal properties (take its organization of cells away and a heart will not beat), so is the interlocking symbiotic relationship of a collection of organs essential to an organism’s modal properties (take away its heart and an animal will not live).

Rea might say there are no necessary arrangements here. A heart will function without some of its cells, and an animal hooked to a blood pump can survive without a heart. But that is a red herring. Without something that moves its blood around, an animal will not live. Therefore, some physical structure is necessary for that modal property, and though many structures can do it, it can’t be just ‘any’ structure (contrary to what EMP’s entail), therefore an animal’s modal property of being able to live is probably intrinsic. Likewise for a heart’s dependency on the structure of its cells. Hence, contrary to Rea’s assertions that no properties are essential to organisms (137), a living human cannot be a living human without a circulation of blood, just as a human cannot be a human without cells. Therefore organisms do have essential properties after all.

As for (2), Rea argues that ME requires belief in the necessity of natural laws. But, he says, we have no empirical evidence supporting belief in the necessity of natural laws, so we can only believe such a thing because of intuition or pragmatism, neither of which is acceptable if naturalism is to solve the Discovery Problem. But Rea is wrong. ME does not depend on the necessity of natural laws. It depends only on physical structure, and therefore it only supports conclusions about metaphysical necessity, requiring no assumptions about the equivalence of that to natural law. Though I believe one can reason from that to the necessity of natural law, this is unnecessary for solving the Discovery Problem.

Consider water microwavability again. (1) No assumptions about natural ‘law’ are needed to see that the physical shape of water molecules and the physical shape of microwaves together entail the modal property of being cookable in a microwave–one need only theorize causal consistency, but such a theory has been amply confirmed by empirical evidence, so it is justified. (2) Though it may seem that microwaves will spin water molecules only if physical laws are ‘necessary,’ that is not the case: if natural law changed tomorrow in any relevant respect, then water would cease to be microwavable. Natural laws are a part of the structure of the universe, and therefore a part of Elder’s microstructural argument for IMP’s. Just as changing the physical structure of H2O will change its modal properties, so will changing the structure of physical laws. This has no effect on the Discovery Problem. Even if natural law is mutable, ME still stands.


Rea’s claim that naturalism “is without rational foundation” is false, because adopting basic empiricism is universally rational, and the worldview of naturalism follows from adopting basic empiricism and applying it to the facts of the world as most carefully observed so far. Rea attempts to bypass this line of reasoning by arguing that naturalism is only a research program and never a worldview, but evidence and argument have proven this claim false. Rea’s ultimate conclusion (that naturalism must abandon materialism, realism about material objects, and realism about other minds) derives solely and entirely from his argument that naturalism cannot solve the Discovery Problem. But I have demonstrated that basic empiricism as a research program and naturalism as a worldview can both solve the Discovery Problem. Therefore, Rea’s ultimate conclusion is false. Naturalism need not abandon materialism or realism, and naturalism remains the most reasonable worldview.[31]

See Also Appendix Below for Three Ancillary Issues Raised by Rea: Intuitionism, Supernaturalism, and Pragmatism.


[1] Michael Rea, World without Design: The Ontological Consequences of Naturalism (Oxford: Clarendon, 2002), 1. Lone page numbers in parentheses refer to this. After I wrote my critique a similar review by Troy Cross was brought to my attention, who agrees with some of my general points, in Notre Dame Philosophical Reviews.

[2] Two different points should be made here by way of crucial digression–first regarding what form of naturalism I am defending, and second regarding the distinction between basic and derived evidence:

(1) I am here defending the epistemological groundwork for metaphysical naturalism. “Naturalists” who reject ‘basic empiricism’ also reject metaphysics, like Quine or Sellars. Thus, Quine, for example, is not really a “metaphysical” naturalist and thus has never advocated a naturalist “worldview,” but only a naturalist “methodology.” In short, Quine held that we can only justify belief in the empirical findings of the natural sciences (and what coheres therewith)–anything beyond that cannot be asserted, but that includes propositions like “There is no God” or “Nature is all there is.” This is why Rea sees in Quine a model for his advocacy of naturalism as a research program. That may be true for Quine (I will leave it to his fans to defend him), but Quine’s philosophy is not what Rea thinks he is arguing against, namely, a belief that there is a natural universe and only a natural universe and that all phenomena can be thereby explained. Quine, for example, would gladly accept the existence of God if scientific evidence confirmed it, and he would not regard that as changing his philosophy at all, which would remain the methodological naturalism it had always been.

(2) Many philosophers mistake derived for basic sources of evidence. The simple rule is: if I ask you “Why should I trust P?” and you can give me an answer, which consists of appeals to some other evidence besides P, then P cannot be basic. It is instead derived from that more basic evidence. This certainly holds for the sense of the term “basic” which Rea uses, as can be seen from the way “basic evidence” features in his argument against naturalism. For the converse test is: if you claim P is basic despite your ability to base it on something else, I would ask how you will ever know when P is false, if in fact P is false. If you answer by an appeal to evidence other than P, then you cannot maintain that P is basic, at least not in Rea’s sense. In my experience, I have never found any defensible evidence that was not really derived from some set of raw, uninterpreted experiences, even if its advocate nevertheless thought it was basic.

[3] For surveys of the meaning and application of the concept of a ‘worldview’: David Naugle, Worldview: the History of a Concept (2002), based on his 1998 dissertation at the University of Texas, Arlington: A History and Theory of the Concept of ‘Weltanschauung’ (Worldview); William Cobern, World View Theory and Science Education Research (Manhattan, Kansas: National Association for Research in Science Teaching, 1991); Leon McKenzie, Adult Education and Worldview Construction (Malabar, Florida: Krieger, 1991); Michael Palmer and Stanley Horton, eds., Elements of a Christian Worldview (Springfield, Missouri: Logion, 1998); Ronald Nash, Worldviews in Conflict (Grand Rapids, Michigan: Zondervan, 1992).

[4] Leading works on naturalism by naturalists, in reverse chronological order: Andrew Melnyk, A Physicalist Manifesto: Thoroughly Modern Materialism (New York: Cambridge University Press, 2003); Daniel Dennett, Freedom Evolves (New York: Viking, 2003); Taner Edis, The Ghost in the Universe: God in the Light of Modern Science (Amherst, New York: Prometheus, 2002); John Shook, Pragmatic Naturalism and Realism (Amherst, New York: Prometheus, 2002); Simon Altmann, Is Nature Supernatural? A Philosophical Exploration of Science and Nature (Amherst, New York: Prometheus, 2002); Joseph Rouse, How Scientific Practices Matter: Reclaiming Philosophical Naturalism (Chicago: University of Chicago, 2002); Matt Young, No Sense of Obligation: Science and Religion in an Impersonal Universe (2001); Robert Nozick, Invariances: The Structure of the Objective World (Cambridge, Mass.: Belknap Press of Harvard University, 2001); Lewis Edwin Hahn, A Contextualistic Worldview: Essays (Carbondale: Southern Illinois University Press, 2001); Kai Nielson, Naturalism and Religion (Amherst, New York: Prometheus, 2001); Barbara Forrest, “Methodological Naturalism and Philosophical Naturalism: Clarifying the Connection,” Science and Society, H. James Birx and Eduard I. Kolchinsky, eds. (St. Petersburg: Russian Academy of Sciences, 2000); J. T. Fraser, Time, Conflict, and Human Values (Urbana: University of Illinois Press, 1999); Roy Bhaskar, The Possibility of Naturalism: A Philosophical Critique of the Contemporary Human Sciences, 3rd ed. (New York: Routledge, 1999); Willem Drees, Religion, Science and Naturalism (New York: Cambridge University Press, 1996); Kai Nielson, Naturalism Without Foundations: Prometheus Lectures (Amherst, New York: Prometheus, 1996); Sidney Hook, The Metaphysics of Pragmatism (Chicago: Open Court, 1996); Richard C. Vitzthum, Materialism: An Affirmative History and Definition (Amherst, New York: Prometheus, 1995); Paul K. Moser and J. D. Trout, Contemporary Materialism: A Reader (London: Routledge, 1995); Jeffrey Poland, Physicalism: The Philosophical Foundation (Oxford University Press, 1994); Peter French, Theodore Uehling, and Howard Wettstein, eds., Philosophical Naturalism (Notre Dame, Indiana: University of Notre Dame Press, 1994); John Ryder, American Philosophic Naturalism in the Twentieth Century (Amherst, New York: Prometheus, 1994); David Papineau, Philosophical Naturalism (Cambridge, Mass.: B. Blackwell, 1993); Jeffrey Walther, Religious Naturalism (Santa Rosa, CA: G. Throwkoff, 1991); Paul Kurtz, Philosophical Essays in Pragmatic Naturalism (Buffalo, New York: Prometheus, 1990) and In Defense of Secular Humanism (Buffalo, New York: Prometheus, 1983); Wilfrid Sellars, Naturalism and Ontology (Reseda, CA: Ridgeview, 1979); Sterling Lamprecht, The Metaphysics of Naturalism (New York: Appleton-Century-Crofts, 1967). Also relevant are works like that of Ilya Prigogine’s Order out of Chaos: Man’s New Dialogue with Nature (Bantam,1984), Jason Casebeer’s Natural Ethical Facts: Evolution, Connectionism, and Moral Cognition (MIT, 2003), and Yervant Krikorian’s Naturalism and the Human Spirit (New York: Columbia University Press, 1944).

[5] For example, leading critiques of naturalism as a worldview are included in: James Beilby, ed., Naturalism Defeated? Essays on Plantinga’s Evolutionary Argument Against Naturalism (Ithaca: Cornell University Press, 2002); David Noebel, J. F. Baldwin and Kevin Bywater, Clergy in the Classroom: The Religion of Secular Humanism, 2nd ed. (Manitou Springs, Colorado: Summit Press, 2002); Frederick Olafson, Naturalism and the Human Condition: Against Scientism (New York: Routledge, 2001); William Lane Craig and J. P. Moreland, eds., Naturalism: A Critical Analysis (New York: Routledge, 2000); David Noebel and Tim LaHaye, Mind Siege: The Battle for Truth in the New Millennium (Nashville, TN: Word, 2000); Phillip Johnson, The Wedge of Truth: Splitting the Foundations of Naturalism (Downers Grove, Ill.: InterVarsity, 2000) and Reason in the Balance: The Case Against Naturalism in Science, Law & Education (Downers Grove, Ill.: InterVarsity, 1998); David Noebel, Understanding the Times: The Religious Worldviews of Our Day and the Search for Truth (Eugene, Oregon: Harvest House, 1994); Steven Wagner and Richard Warner, eds., Naturalism: A Critical Appraisal (Notre Dame, Indiana: University of Notre Dame Press, 1994); Ronald Nash, Worldviews in Conflict: Choosing Christianity in a World of Ideas (Grand Rapids, Michigan: Zondervan, 1992); William Shea, The Naturalists and the Supernatural (Mercer University Press, 1984).

[6] Cf. Edward B. Davis and Robin Collins, “Scientific Naturalism,” Science & Religion: A Historical Introduction, Gary Ferngren, ed. (Johns Hopkins University Press, 2002), 322-34.

[7] Cf. Alex Rosenberg, “A Field Guide to Recent Species of Naturalism,” British Journal for the Philosophy of Science 47:2 (1996), 1-29. I have noticed several Christian writers who cite this essay as if it contained a complete accounting of all live versions of naturalism. Rosenberg doesn’t even claim to have done that, and he hasn’t.

[8] Plantinga argues that the probability is inscrutable and therefore we cannot claim it was within such a bound of possible outcome. This argument does not succeed, because the reliability of our faculties is already established beyond doubt, and therefore must be a possible outcome. Given this, in conjunction with other evidence collectively supporting naturalism over supernaturalism, it is reasonable to conclude that the probability, even if inscrutable, nevertheless must fall within the bounds of possibility. This conclusion is supported by the fact that there is no reason to believe the probability is any lower (e.g. Plantinga offers no real evidence that it is, and his hypothetical evidence entails outcomes of evolution that are far less probable than the outcome he wishes to deny, which is prima facie highly probable: the production of reliable faculties).

[9] Indeed, this very point is defended at length by another Evangelical critic of Naturalism: Benjamin Wiker, Moral Darwinism: How We Became Hedonists (Downers Grove, Ill.: InterVarsity, 2002). I am preparing a critical review of Wiker’s book at the present time.

[10] Taner Edis, The Ghost in the Universe: God in the Light of Modern Science (Amherst, New York: Prometheus, 2002).

[11] Evan Fales, “Can Science Explain Mysticism?” Religious Studies 35 (1999), 213-27; “Divine Intervention,” Faith & Philosophy 14 (1997), 170-94; “Mystical Experience as Evidence,” International Journal for Philosophy of Religion 40 (1996), 19-46; “Scientific Explanations of Mystical Experiences II: The Challenge to Theism,” Religious Studies 32 (1996), 297-313; “Plantinga’s Case Against Naturalistic Epistemology,” Philosophy of Science 63 (1996), 432-451.

[12] Theodore Schick, “Methodological Naturalism vs. Methodological Realism,” Philo 3:2 (2000), 30-37.

[13] David Griffin, Parapsychology, Philosophy, and Spirituality: A Postmodern Exploration (Albany, New York: State University of New York Press, 1997) and Religion and Scientific Naturalism: Overcoming the Conflicts (ibid., 2000).

[14] James Lippard, “Historical but Indistinguishable Differences: Some Notes on Victor Reppert’s Paper,” Philo 2:1 (Spring-Summer, 1999), 45-47; “Charles Fort,” in Gordon Stein, ed., Encyclopedia of the Paranormal (Amherst, New York: Prometheus, 1996), 277-281.

[15] Keith Augustine, A Defense of Naturalism. Master’s Thesis. (College Park: Graduate School of the University of Maryland, 2001). The text is available on the Secular Web.

[16] Graham Oppy, “Review of Naturalism: A Critical Analysis ed. by William Lane Craig and J.P. Moreland” (The Secular Web, 2000).

[17] Ernest Nagel, “Naturalism Reconsidered,” in Houston Peterson, ed., Essays in Philosophy. (New York: Pocket Books, 1960).

[18] Paul Draper, “God, Science and Naturalism,” in Bill Wainwright, ed., Oxford Handbook of Philosophy of Religion (Buffalo, New York: Prometheus), due to appear in 2004. For Draper’s definition of (Metaphysical) Naturalism, see Jeff Lowder’s note: Metaphysical Naturalism?

[19] Jeffery Jay Lowder and Phil Fernandes, “The Lowder-Fernandes Debate: Naturalism vs. Theism—Which Way Does the Evidence Point?” (1999), debate held at the University of North Carolina at Chapel Hill and sponsored by the Internet Infidels and the Institute of Biblical Defense.

[20] I am excluding one respondent. Ted Drange, best known for his book Nonbelief and Evil: Two Arguments for the Nonexistence of God (Amherst, New York: Prometheus, 1998), gave me a mixed reply, answering Y / N if science is conceived ‘broadly’ and N / [N/A] if conceived ‘narrowly.’ Since I could not ascertain which of Drange’s conceptions of science Rea has in mind, Drange’s reply could not be understood in any way relevant to this inquiry.

[21] For this definition, and the fact that the word “naturalism” alone often carries this meaning: s.v. “naturalism,” Oxford English Dictionary, §2; also, Chris Rohmann, “naturalism,” A World of Ideas: A Dictionary of Important Theories, Concepts, Beliefs, and Thinkers (New York: Ballantine): 274-75.

[22] Lilian Furst and Peter Skrine, “The Term ‘Naturalism,'” Naturalism (London: Methuen, 1971): 2-3.

[23] Fales, op. cit.

[14] Mark Steiner, The Applicability of Mathematics as a Philosophical Problem (Cambridge, Mass.: Harvard University Press, 1998). For a critical review: Richard Carrier, “Fundamental Flaws in Mark Steiner’s Challenge to Naturalism in The Applicability of Mathematics as a Philosophical Problem” (The Secular Web, 2003).

[25] David Foster, The Philosophical Scientists (London: C. Hurst, 1985). For a critical review: Richard Carrier, “Bad Science, Worse Philosophy: The Quackery and Logic-Chopping of David Foster’s The Philosophical Scientists” (The Secular Web, 2000).

[26] Augustine, op. cit.

[27] In Benjamin Wiker, op. cit.: 13; same sentiment appears in Victor Reppert, C.S. Lewis’s Dangerous Idea: In Defense of the Argument from Reason (Downer’s Grove, Ill.: InterVarsity Press, 2004); C. S. Lewis, Miracles: A Preliminary Study (New York: MacMillan, 1978): 23; and Ian J. Thompson, “How to Distinguish the Natural and Spiritual” (TheisticScience.org, 2002).

[28] Given some misunderstanding on the part of one reviewer of this article, I shall include some digressions here that I don’t think are necessary, but should certainly make what I am saying in the body of this article absolutely clear:

  1. There is a different geometric shape, and a different relationship in space and time between parts and a different pattern of causal connection between and among parts, between a rabbit and a mere collection of particles. This physical arrangement is what makes a rabbit a rabbit. For example, a gold ring will roll down an incline, but a gold block will not–even if the two are made of exactly the same gold. That is, take a gold ring and sit it on its side on an incline and it will roll down. Then smash that ring into a block and sit it back at the same starting point and it won’t roll down. Yet what has changed? No amount of the substance has changed. The substance itself hasn’t changed. The nature of that substance hasn’t changed. Both objects are made of exactly the same atoms, not just the same kind and number of atoms (though that, too, obviously). So what is different? The physical arrangement of those atoms. In other words, the geometric arrangement of those atoms allows the ring to roll but not the block. And this property follows simply from the axioms of geometry (as was proved by Archimedes in his geometric study of mechanics), which follow necessarily from the physical facts of space-time as a geometric space.

    This same point carries beyond space-time geometry into causal space as well. Take a circulatory system: the system of veins and arteries and nerves and the way they are physically connected to each other and various organs changes the “causal” space of the system (geometrically, this is closer to the study of knots than Archimedean mechanics, but it involves both). For example, if the nerves that control arterial contraction (and thus regulate blood pressure) were physically connected to a rabbit’s skin instead, then the circulatory system would be incapable of regulating blood pressure as efficiently as it would if the nerves were connected to the arterial muscles, as they usually are. Thus, simply physically rearranging the parts–and doing nothing else but that–changes the way the system causally functions. Thus, there is a difference–a purely physical difference–between a mere pile of arteries and nerves, and a pile that is physically connected in the way a rabbit’s circulatory system is connected. Thus, the difference cannot be merely conceptual. And this difference is not merely about physical relation in space, since one can scrunch, stretch, twist or change the overall shape of a circulatory system to some degree without changing its causal powers. But there is still a fixed geometry to the system that never changes throughout such manipulations, just as there is for a knot, unless that geometry is destroyed (the knot broken or untied)–as when a rabbit is smashed by a steamroller.

  2. My reviewer also suggested that origin plays such a part in defining an object that arrangement is insufficient to demarcate categories. I disagree. This may sometimes be true, but is not always necessarily true. At any rate, Rea rejects the role of historical origins: as I noted already, he thinks there is no difference between a statue and a natural rock formation that happens to look like a statue. But he can only say that if he rejects the causal history of the object. For once we take that into account, there is a physical difference between a statue and a natural rock formation, and Rea’s claim that there is no difference must be rejected. Thus, either way, Rea loses. Either causal history is not relevant to categorization (in which case physical arrangement is sufficient, e.g. a rabbit that we assemble atom-by-atom in the lab is still a rabbit), in which case Rea cannot claim it is insufficient, or causal history is necessary to categorization (in which case we need only add that to our consideration, e.g. where a rabbit comes from is a purely physical fact, with location and pattern in space and time, and thus not an arbitrary concept of the human mind), in which case Rea cannot claim there is no physical difference between two objects we categorize as different, since such objects will always differ in causal-historical origins (and in just that respect that is relevant to our categories, e.g. a statue was physically caused by the intelligent agency of a human being, while a natural rock formation was not caused by any intelligent agency nor by any action of a human being).

[29] Rea also tries to develop an argument against allowing naturalists to even use an ABE at all. See Appendix below: Pragmatism.

[30] Himanshu Jain, “Learning about Electric Dipoles from a Kitchen Microwave Oven” (Materials, Science, and Technology Online, 1998), available online as a PDF Document. Many other properties of water have been traced to its physical geometry: see Peter Weiss, “Wet ‘n’ Wild: Explaining Water’s Weirdness,” Science News 165(4), 24 January 2004, pp. 58-60.

[31] Elsewhere Rea also attacks “Moral Realism” using essentially the same kind of arguments: Michael Rea, “Naturalism and Moral Realism,” in Knowledge and Reality, Thomas Crisp, David VanderLaan, and Matthew Davidson, eds. (Dordrecht: Kluwer, 2004), which is available online as a PDF Document. The main flaw in the argument he presents there is an over-constrained definition of “moral realism,” and consequently his argument is mooted by the ethical theory I defend in my forthcoming book Sense and Goodness without God: A Defense of Metaphysical Naturalism.

Appendix to Defending Naturalism

There are three issues that come up in Rea’s book that are not essential to my conclusion above, but which may be of interest to naturalists and opponents of naturalism. Those issues are “Intuitionism,” “Supernaturalism,” and “Pragmatism” I will devote a short discussion to each, with the assumption that the reader has already read everything above.


Naturalism, according to Rea, rejects ‘intuitionism,’ which he defines as “treat[ing] rational intuition as a basic source of evidence” (67) insofar as a “rational intuition” is “a conscious episode not involving sense perception, memory, or inference” whereby “a proposition seems to be necessarily true” sic (68; cf. 201), and “a basic source of evidence” is one that is “trusted even in the absence of positive evidence in favor of [its] reliability” (2). He distinguishes a rational intuition from a “physical intuition” (175) without giving examples of the latter, though one can infer that these are intuitions of probable rather than necessary truths.

Rea’s line of argument seems to allow naturalism to be compatible with physical intuitions. At any rate, he only asserts that naturalism cannot treat “the appearance of necessity as evidence in favor of a proposition” since this “presupposes that it is possible for some propositions to be justified a priori,” and naturalism, being based on empiricism, denies this (175). He must mean only propositions about the facts of the world, since naturalists certainly do believe that propositions about other propositions (or sets of propositions) can be justified a priori, and Rea spends a lot of time agreeing that rational intuitions about such things are indeed justifiable on naturalism, at least as derivative sources of evidence (e.g. “we have empirical evidence in favor of the belief that our intuitions are reliable in the domain of logical, mathematical, and conceptual truths,” 211; cf. 203; he also implies that the ‘rational intuitions’ naturalists cannot appeal to are those of synthetic a priori truths: 210).

I see no need to dispute Rea’s argument here (192-99). Given his extremely narrow definition of intuition, I agree there is no reason naturalists should accept that as a basic source of evidence about the facts of the world (and Rea’s focus is on modal properties, which are facts of the world). However, unlike Rea, I can imagine naturalistic mechanisms that would make such intuitions available to naturalists as basic sources of evidence. For example, if humans were really engineered by naturally-evolved aliens, naturalism could remain entirely true, but all manner of cognitive abilities could have been intelligently designed in humans. And if we took this assumption as part of our research program, we would have a version of naturalism that, even on Rea’s assumptions, would be justified in taking rational intuitions as a basic source of evidence. Perhaps this would not properly be ‘naturalism’ on Rea’s view, unless we acquired scientific evidence that this alien-design scenario was true, in which case we could trust rational intuitions derivatively. But since there is no evidence of this, nor any reason to assume it, naturalists today cannot appeal to this in defense of treating rational intuitions as sources of evidence about the world.

But that doesn’t seem very relevant to Rea’s objective. For example, can Rea really exclude physical intuition, e.g. as the basis for Elder’s assumptions (138)? Even in general, whether taken as basic or derivative, why can’t naturalists base their belief in Intrinsic Modal Properties (IMP’s) on a physical intuition? Rea never addresses this question, and that is a major gap in his book’s argument. After all, ‘physical intuition’ seems to explain well the real nature of scientific procedure. Consider the microwavability of water: physical intuitions about atomic structure and behavior certainly do support belief in IMP’s there, and though they are backed up empirically in that case, from such examples one can justify some measure of trust in unverified physical intuitions as evidence to be considered in comparable cases science has yet to explore. Likewise, there is evidence that the reliability of such intuitions increases both in proportion to experience with the relevant kinds of circumstances (e.g. a physicist’s intuition about ball lightning is sooner to be trusted than a tailor’s) and in conjunction with a successful explanatory model (e.g. atomism has produced many more successful physical intuitions than Platonism). Moreover, even rational intuitions about formal truths can be mapped onto observed phenomena and thus become predictive of them, and doing just this is in fact an indispensable part of the scientific method. Though this still only produces probable truths (since the appropriateness of such mapping is precisely what science seeks to ascertain empirically), rational intuitions of formal truths still play a role in science, and play a valid part in justifying IMP’s on naturalism.


Rea concludes his book by arguing that ‘supernaturalism’ (the belief that religious experience is basic evidence) can support belief in IMP’s and thus in everything else he says naturalism is committed to abandoning (like materialism and realism about other minds), but in actual fact he spends most of his time defending Traditional Theism (by which he means Judeo-Christian Theism: e.g. 223) as the only form of supernaturalism that is successful at this, even though he admits he isn’t even going to do any of the “substantive theology” necessary to make the point (223). So he really only outlines a possibility.

When Rea faces the troubling fact that Traditional Theism entails the prediction that everyone would have a basic “religious experience” of God, yet in fact “very few people” do (222), he tries to escape with an ad hominem against atheists, naturalists, and all religionists who reject Traditional Theism. “Many theists hold,” he says, that “everyone has some experience or other that they recognize as an experience of the God of traditional theism, but many people suppress those experiences.” He says “it is at least easy to see why people might be inclined to” do that, quoting Nagel’s statement that he actually hopes there is no God and that this desire, rather than justified unbelief, “is responsible for much of the scientism and reductionism of our time” (212). He explains away the fact that we honestly tell him we haven’t had such experiences by insisting that if we are suppressing them we wouldn’t know it (222-23). In effect, Rea builds an unfalisfiable interpretation of the facts so that any evidence will confirm his view. He actually makes this the standard defense against all charges that religious experience is not reliable. For example, faced with the problem of religious diversity, he says any such diversity can be explained away by theory and thus isn’t a problem (224), which is basically circular, unfalsifiable reasoning.

Rea offers only one credible approach to shoring up the reliability of religious experience (RE), though he does not say this has been done (because it hasn’t). He compares RE to vision: though there are discrepancies in the reports of sighted people, we nevertheless don’t abandon trust in our sight. Rather, we develop “a story to tell” that explains these discrepancies, thereby distinguishing reliable cases of seeing from unreliable ones (225). He says this must be valid because “there is no vision-independent way of resolving the matter,” but that isn’t entirely true, nor does it justify his attempt to defend RE. For example, through the construction of instruments we can convert light to sound and thus ascertain where there really is light and where not, and what its qualities are. This has actually been done: light has been converted to vibrations that are felt on the skin of the blind, who can thus see objects and navigate around them. More obviously, the very reason we have multiple senses is precisely to check each other. For instance, we can tell that our ‘blind spot’ does not really correspond to a hole in the universe because we can still feel our finger when we put it there and it otherwise disappears from view. And we can tell our arm isn’t ‘bent’ when we put it in water because we can feel it is really not located where we see it.

Even so, most vision-corrective behaviors do rely on vision, usually in conjunction with the verbal reports of others or a more careful visual examination of the relevant facts. A notable example directly relevant to Rea’s attempt to defend RE is synesthesia, the disorder that causes people to ‘see’ sounds–whatever they hear produces colors and patterns in their visual field, and they can actually use this ability to accurately discriminate among sounds just as the rest of us do with our ears. How do these people come to believe that the light they see isn’t really there? Because (1) other sighted people verify that it isn’t, (2) we can show them the physical differences in brain structure and activity between synesthetes and nonsynesthetes, (3) they can physically block, reflect, or refract all real light approaching their eyes, but not the apparent light of their synesthetic visions (in fact, while a lamp will illuminate a dark room, a radio will not–yet it will still cause colors to appear in their visual field), (4) we can build a light-detecting machine that reports the presence of all light, and they can observe that machine failing to report the light appearing in their synesthetic visions, and, theoretically, (5) we could even show them the pattern of electrical activity in their optic nerve which confirms that there are no signals there corresponding to the apparent synesthetic visions. In short, there are at least five ways to empirically demonstrate that the synesthete isn’t seeing real light.

Is anything comparable available for RE? No. Rea doesn’t even try to come up with any. Instead, he simply declares that we can trust RE so long as we can make up any story that works. We don’t need any empirical evidence that our story is the correct one. If that is true, then synesthetes would be justified in calling all of us partially blind (just as we call some people ‘color blind’). All they need is a ‘story’ that makes sense of the accusation. They don’t need any evidence. But that is clearly incorrect. When faced with disagreeing perceptions we cannot assume we are right and everyone else is wrong, no matter what story we cleverly devise. We must have a reason to believe our story is correct, and contrary stories incorrect. And it is hard to see how such a reason could be anything other than empirical. Either way, Rea has done none of the work actually required to make reliance on RE plausible, much less credible.

Finally, Rea doesn’t address alternative supernaturalisms adequately enough to support any kind of confidence that Traditional Theism is the ‘best’ solution. For example, he concludes that “though ultimately I reject it … naturalism is the most viable research program apart from a brand of supernaturalism that warrants belief in a suitably developed version of traditional theism” since “most of the nontheistic religions that have a coherent concept of divinity do not recommend belief in a mind of whose existence, behavior, or characteristics we could be directly aware” (226). He says most. But most does not mean all. Therefore Rea apparently acknowledges there are nontheistic religions that are as viable a research program as his traditional theism, yet refuses to discuss them for some reason. Isn’t it incumbent upon him to explore those before declaring traditional theism the best?

Taoism, for example, can offer a research program that satisfies all of Rea’s concerns. We are a part of the Tao and thoroughly akin to it and thus can be aware of it directly. Taoism in fact undergirds both religious experience and rational intuition. Moreover, Taoism has many arguments in its favor over against theism as a metaphysical view of reality. For example, it resolves the Argument from Evil with minimal ad hoc baggage, as well as the Arguments from Divine Hiddenness, Religious Diversity, and Intrareligious Confusion. At the same time, all the arguments for generic theism actually support Taoism with equal force, especially the Ontological, Cosmological, and Design Arguments. Yet Taoism is nontheistic–or, at least, it cannot be camped with Traditional Theism. For though the Tao is describable as a ‘mind’ of sorts, it is more a ‘vegetable mind’ or ‘mental force,’ lacking reason or desire in any conventional sense, yet capable of creating and sustaining a coherent reality in accord with its inherent nature.

Taoism also lacks a heaven or hell (it allows no personal survival of death), and it lacks a personal being to whom it would make any sense to pray, so it is certainly not what Rea is rooting for. Yet it seems evident to me that Taoism is a much better research program than Traditional Theism. For it requires commitment to a lot less metaphysical baggage and fits far better all the observed evidence. It certainly looks a lot more attractive to me. Just about the only thing Rea can offer for theism that Taoism lacks is the prospect for immortality. So is that all this really boils down to? Is Rea a theist for no greater reason than that he wants to live forever and thinks wishing will make it so? I won’t claim to know. But it is easy to see how a Taoist can tell just as good a story about Rea’s denial of the truth of Taoism as Rea spins for us atheists.

Pragmatic vs. Epistemic Criteria for Truth

Rea also tries to undermine all ABE’s by suggesting that they require pragmatic criteria, and that pragmatic criteria cannot lead to epistemic conclusions (135, n. 3; cf. 138-57). Rea argues that even if taken as basic evidence pragmatic criteria entail things about the universe naturalists won’t want to accept. But that is moot, since I reject Rea’s claim that such methods are basic for any naturalist. Insofar as they are accepted at all, they only provide derivative evidence.

Rea notes “the mere fact that it is convenient for us to think and act if [certain] beliefs are true is no evidence for their truth” (145), yet scientists seem to treat convenience as evidence of truth. This requires assuming “someone or something in the universe is somehow guaranteeing that pragmatic criteria for theory choice will be truth-conducive,” which would be either god or us–the former entailing theism, the latter entailing constructivism (156). Rea does not explicitly connect this to the formal logic of an ABE. He only hints that there may be a connection, and then proceeds under cover of a flurry of qualifications (“suppose it is true … seems to be … on the supposition that …” etc.). He thus attacks pragmatism simply to cover all bases, e.g. Rea argues that ME either relies on intuition or pragmatism, and if intuition produces truth then naturalism is false, and if pragmatism produces truth then naturalism is either false or else committed to constructivism–so naturalists must abandon materialism.

There are two things wrong with Rea’s attack on pragmatism. He doesn’t really consider mind-free world models that would nevertheless align certain pragmatic criteria to the truth. He just assumes there are none. Yet isn’t it obvious that believing in true propositions is useful, in a way that believing in false propositions is not? Even if there can sometimes be pragmatic reasons to believe false propositions, it still follows that pragmatic criteria can often lead us to the truth even without a mind making that so, and one can easily see how a statistical argument would validate a pragmatic basis for truth discovery. But I won’t defend that here, since I think pragmatic criteria only have a trivial function in naturalist epistemology: for Rea is wrong that science depends on pragmatic criteria as he seems to conceive them. For instance, when Rea finally lists some actual “pragmatic criteria” employed in science, it isn’t clear what makes them pragmatic rather than epistemic. He offers only “simplicity, elegance, and compatibility with entrenched views” (156), but these don’t look pragmatic, at least insofar as they are genuinely accepted in science.

Simplicity, for example, is an expected feature of any mechanical, non-sentient universe: obviously to explain complex phenomena we must look for the many simpler causes that accumulate to produce it. And this conclusion is epistemic, not pragmatic: we observe it to be a fact that complex phenomena are comprised of simpler ones. No mind is needed to make that so. And besides that, much of what science settles on is not simple: the periodic table is vastly more complicated than the simple system of four or five elements early chemists tried to describe the world with. Thus, scientists readily abandon simplicity as a criterion, recognizing that the truth is often not simple after all. The same can be said of “compatibility with entrenched views.” Not only are scientists quite ready to go against this criterion when the evidence so dictates, but their trust in “entrenched views” is entirely empirical, not pragmatic: those views are empirically confirmed so thoroughly that to go against them would be a decidedly un-empirical behavior–unless evidence very strongly leads to a different conclusion.

Finally, one might think “elegance” is surely pragmatic. But “elegance” is not really regarded as scientific evidence. As an undefinable criterion, it is accepted by no scientific journal as grounds for asserting a theory’s factual truth. Insofar as “elegance” is accepted by anyone as “evidence,” this acceptance is wholly based on the prior empirical success of employing that criterion, so “elegance” is not regarded as basic evidence as Rea must contend. Finally, what scientists have come to call “elegant” looks nothing of the kind to nonscientists (e.g. taken as a whole, the equations of Relativity Theory are a nightmare of monstrous complexity bringing cold sweats to any uninitiated observer) and thus “elegance” is probably a learned response to the truth-bearing properties of the brute structure of the universe (i.e. no mind made the universe elegant, we just learned to call ‘elegant’ the way the universe is). So Rea cannot even establish “elegance” as a pragmatic criterion inherent to science.

Rea does essentially argue at one point (162) that: (1) there is no epistemic way to distinguish a real world from the illusions of a Cartesian Demon; (2) therefore, only pragmatic grounds can be given for rejecting the latter; (3) but, per Rea’s argument, such pragmatic grounds assume the existence of a Cartesian Demon; (4) therefore there must be a Cartesian Demon (which, according to Rea, would be God or, according to constructivism, us). The premises most likely dubious then would be (1) and (3).

As to (1), naturalists can always imagine epistemic differences between a Cartesian Demon’s world and a real world, differences that could in theory be looked for and observed, unless the demon is absolutely and eternally consistent in presenting a world entirely consistent with metaphysical naturalism and nothing else. But if a demon is doing that, in what sense would naturalism be false? Would it be false in any way relevant to us? Not really. On such a view, only belief in naturalism would allow us to achieve all our goals, for no other worldview would be credible except naturalism, given that no proposition not true on naturalism would be made true by the Demon. So there would be no reason not to be a naturalist on this view. For then, believing in the demon is pointless–it gains us nothing, offering us no predictions not already true on naturalism alone, nor does it change the truth of any of our propositions (e.g. objects still ‘exist’ and still have ‘properties’ wholly apart from their relation to our minds, etc.). When we combine this conclusion with the fact that naturalism is already sufficiently intelligible and capable of being true without any such demon making it so, we arrive at a valid epistemic conclusion that there is no reason to believe in that demon, but plenty of reason left to believe in naturalism alone. Therefore (1) is false.

As to (3), if we chose to believe in naturalism alone, sans Cartesian Demon, merely because it was convenient (e.g. believing in the demon predicts nothing whatsoever about the world and is thus useless), would this really in any way imply that there must be a Cartesian Demon? No. For this line of reasoning is entirely consistent with the proposition ‘there is no Cartesian Demon.’ Rea argues that the only way we could trust that this ‘convenience’ corresponded with the truth is if some mind was making the truth suit our convenience. But that isn’t so. The world we observe is so thoroughly consistent, with absolutely no signs of a Cartesian Demon, that the mere possibility that a demon is at fault is empirically indefensible (i.e. there is absolutely zero empirical basis for such a belief). In that situation, no Cartesian Demon is needed to justify our pragmatic belief that what is not probable is not credible. Even though ‘there is a Cartesian Demon’ is still consistent with the evidence we do have, so we do not have epistemic grounds to deny it, we do not need to appeal to a Cartesian Demon to validly say that there is no reason to believe it, so we won’t. Therefore (3) is false, unless Rea would say this was not a pragmatic argument, in which case it is epistemic and (1) is false.

all rights reserved