We saw in part I of this series that Alex Rosenberg’s new book The Atheist’s Guide to Reality is less about atheism than it is about scientism, the view that science alone gives us knowledge of reality. This is so in two respects. First, Rosenberg’s atheism is just one implication among others of his scientism, and the aim of the book is to spell out what else follows from scientism, rather than to say much in defense of atheism. Second, that it follows from his scientism is thus the only argument Rosenberg really gives for atheism. Thus, most of what he has to say ultimately rests on his scientism. If he has no good arguments for scientism, then he has no good arguments either for atheism or for most of the other, more bizarre, conclusions he defends in the book.
So, does Rosenberg have any good arguments for scientism? He does not. In fact, he has only one argument for it, and it is quite awful.
What is scientism?
Before we look at the argument, let’s consider how Rosenberg characterizes scientism:
“Scientism”… is the conviction that the methods of science are the only reliable ways to secure knowledge of anything; that science’s description of the world is correct in its fundamentals; and that when “complete,” what science tells us will not be surprisingly different from what it tells us today. (pp. 6-7)
As I’ve noted elsewhere (e.g. here, here, and here), the trouble with the claim that science is the only reliable source of knowledge is that it is either self-defeating or trivial -- self-defeating if we narrowly construe what counts as “science” (since scientism is itself a metaphysical and epistemological theory and not a view that physics, chemistry, or any other particular science has established) and trivial if we construe “science” broadly (since in that case philosophy, and in particular metaphysics and epistemology, count as “sciences” no less than physics, chemistry, and the like do). Rosenberg certainly avoids the second horn of this dilemma. For his construal of what counts as “science” is very narrow indeed:
If we’re going to be scientistic, then we have to attain our view of reality from what physics tells us about it. Actually, we’ll have to do more than that: we’ll have to embrace physics as the whole truth about reality. (p. 20)
To be sure, he does not deny that chemistry, biology, and neuroscience also give us knowledge. But that is only because he thinks they are reducible to physics: “The physical facts fix all the facts. [This] means that the physical facts constitute or determine or bring about all the rest of the facts.” (p. 26)
Now some naturalists will demur at this point, preferring a “non-reductive physicalism,” or “emergentism,” or some other such doctrine to Rosenberg’s radical reductionism. As a number of chemists and philosophers of chemistry have argued in recent years, it is at the very least debatable whether even chemistry is really reducible to physics. (For a useful overview of the literature, see chapter 5 of J. van Brakel’s book Philosophy of Chemistry. Also useful is the Stanford Encyclopedia of Philosophy article on the philosophy of chemistry.) Reductionism in biology is even more obviously open to challenge. And of course, whether consciousness and human thought and action can be accounted for in physicalist terms is notoriously controversial even among naturalists themselves -- Fodor, McGinn, Searle, Nagel, Levine, Strawson, and Chalmers are just some of the prominent naturalistic philosophers of mind who have been critical of existing attempts by their fellow naturalists to explain the mind in purely materialistic terms.
Now I sympathize with such arguments, but I don’t think they establish an alternative form of naturalism. For what they show, I would argue, is that higher-level features of material reality are no less real than the lower-level features, that the lower-level features are not somehow ontologically privileged. And in that way they show (even if only inchoately, and even if their proponents often do not realize it) that something like an Aristotelian, holistic conception of material substances is correct after all. Talk of “emergence,” “non-reductive physicalism,” and the like fudges this, because it insinuates that the lower-level features described by physics are still somehow more fundamental than the higher-level ones, even though the higher-level ones are acknowledged to be irreducible. The latter, it is implied, somehow have to “emerge” from the former. Such views are bound to sound obscurantist precisely because they amount to an unstable halfway position between reductionistic naturalism of the Rosenberg variety and traditional Aristotelian anti-reductionism.
I would say, then, that one has either to go the whole hog for Rosenberg-style reductionism or chuck out the whole naturalistic framework altogether (along with “emergence” and other such half-measures) and return to a full-blown Aristotelian metaphysics of material substances. To that extent I think Rosenberg is right to hold that if someone is committed to scientism, then he should hold that “the physical facts fix all the facts.” (Obviously some will dispute this conditional, but since it constitutes a point of agreement between Rosenberg and me, I won’t pursue it further here.)
If Rosenberg avoids the one horn of the dilemma, though, he thrusts himself headlong onto the other. For how exactly has scientism been established by physics, chemistry, biology, or even neuroscience (if we allow for the sake of argument that neuroscience is reducible to physics)? Does scientism make predictions that have been rigorously confirmed? Is there something like a Michelson-Morley experiment that scientism makes sense of in a way no rival theory does? To ask such questions is to answer them. The fact is that neuroscience hasn’t come close even to discovering exactly what it is that goes on in the brain when scientists form hypotheses, construct theories, make predictive inferences, develop experimental tests, write up their results, submit them for peer review, etc. That is to say, neuroscience hasn’t even explained the practice of science itself in purely neuroscientific categories, much less shown that no other practices can yield genuine knowledge. Scientism remains what it has always been -- a purely metaphysical speculation and not an empirical theory at all, much less a confirmed empirical theory.
No doubt we will be treated at this point to some hand-waving to the effect that even if neuroscience has not “yet” fully explained scientific practice, neither has it turned up any evidence that there are sources of knowledge other than science. But whether neuroscience is the only genuine source of knowledge about how we come to have knowledge is itself part of what is at issue in the dispute between scientism and its critics. Hence, to argue “We have no neuroscientific evidence that there is any genuine source of knowledge other than science, therefore there are no grounds at all for believing that there are any such alternative sources” would simply be to beg the question.
All of this might seem moot if Rosenberg had a really powerful argument in favor of scientism. But he does not. David Stove once gave the ironic label “the Gem” to a Berkeleyan argument for idealism he regarded as especially bad. Rosenberg’s argument for scientism gives Berkeley a run for his money, for it is a real Gem. He states it several times in the book:
The technological success of physics is by itself enough to convince anyone with anxiety about scientism that if physics isn’t “finished,” it certainly has the broad outlines of reality well understood. (p. 23)
And it’s not just the correctness of the predictions and the reliability of technology that requires us to place our confidence in physics’ description of reality. Because physics’ predictions are so accurate, the methods that produced the description must be equally reliable. Otherwise, our technological powers would be a miracle. We have the best of reasons to believe that the methods of physics -- combining controlled experiment and careful observation with mainly mathematical requirements on the shape theories can take -- are the right ones for acquiring all knowledge. Carving out some area of “inquiry” or “belief” as exempt from exploration by the methods of physics is special pleading or self-deception. (p. 24)
The phenomenal accuracy of its prediction, the unimaginable power of its technological application, and the breathtaking extent and detail of its explanations are powerful reasons to believe that physics is the whole truth about reality. (p. 25)
Rosenberg’s argument, then, is essentially this:
1. The predictive power and technological applications of physics are unparalleled by those of any other purported source of knowledge.
2. Therefore what physics reveals to us is all that is real.
How bad is this argument? About as bad as this one:
1. Metal detectors have had far greater success in finding coins and other metallic objects in more places than any other method has.
2. Therefore what metal detectors reveal to us (coins and other metallic objects) is all that is real.
Metal detectors are keyed to those aspects of the natural world susceptible of detection via electromagnetic means (or whatever). But however well they perform this task -- indeed, even if they succeeded on every single occasion they were deployed -- it simply wouldn’t follow for a moment that there are no aspects of the natural world other than the ones they are sensitive to. Similarly, what physics does -- and there is no doubt that it does it brilliantly -- is to capture those aspects of the natural world susceptible of the mathematical modeling that makes precise prediction and technological application possible. But here too, it simply doesn’t follow for a moment that there are no other aspects of the natural world.
Those who reject Rosenberg’s scientism, then, are not guilty of “special pleading or self-deception,” Rosenberg’s condescending bluster notwithstanding. Rather, they are (unlike Rosenberg) simply capable of recognizing a brazen non sequitur when they see it. Unfortunately, condescending bluster is all Rosenberg ever offers in addition to his favorite non sequitur. Here’s some more of it:
“Scientism” is the pejorative label given to our positive view by those who really want to have their theistic cake and dine at the table of science’s bounties, too. Opponents of scientism would never charge their cardiologists or auto mechanics or software engineers with “scientism” when their health, travel plans, or Web surfing are in danger. But just try subjecting their nonscientific mores and norms, their music or metaphysics, their literary theories or politics to scientific scrutiny. The immediate response of outraged humane letters is “scientism.” (p. 6)
According to Rosenberg, then, unless you agree that science is the only genuine source of knowledge, you cannot consistently believe that it gives us any genuine knowledge. This is about as plausible as saying that unless you think metal detectors alone can detect physical objects, then you cannot consistently believe that they detect any physical objects at all. Perhaps someone who thinks that metal detectors give us exhaustive knowledge of the world could write up a Metallicist’s Guide to Reality and “argue” as follows:
“Metallicism” is the pejorative label given to our positive view by those who really want to have their stone, water, wood, and plastic cakes and dine at the table of metallic bounties, too. Opponents of metallicism would never charge their metal detector-owning friends with “metallicism” when they need help finding lost car keys or loose change in the sofa. But just try subjecting their nonmetallic mores and norms, their music or metaphysics, their literary theories or politics to metallurgical scrutiny. The immediate response of outraged humane letters is “metallicism.”
Of course, “metallicism” is preposterous. But so is Rosenberg’s scientism.
Those beholden to scientism are bound to protest that the analogy is no good, on the grounds that metal detectors detect only part of reality while physics detects the whole of it. But such a reply would simply beg the question once again, for whether physics really does describe the whole of reality is precisely what is at issue.
I am being hard on Rosenberg, and he deserves it for putting forward such transparently bad arguments, and with such arrogance. But it is only fair to note that he is hardly alone in the delusion that his Gem is some kind of knockdown argument for scientism. One hears this stupid non sequitur over and over and over again when arguing with New Atheist types. It is implicit every time some Internet Infidel asks triumphantly: “Where are the predictive successes and technological applications of philosophy or theology?” This is about as impressive as our fictional “metallicist” smugly demanding: “Where are the metal-detecting successes of gardening, cooking, and painting?” -- and then high-fiving his fellow metallicists when we are unable to offer any examples, thinking that he has established that plants, food, works of art, and indeed anything non-metallic are all non-existent. For why on earth should we believe that only methods capable of detecting metals give us genuine access to reality? And why on earth should we believe that if something is real, then it must be susceptible of the mathematically precise prediction and technological application characteristic of physics? I submit that there is no answer to this question that doesn’t beg the question.
As always, earlier generations of skeptics were wiser than the intellectually backward Dawkins generation. For instance, Bertrand Russell was well aware that, far from giving us an exhaustive picture of reality, physics in fact gives us is very nearly the opposite, and is unintelligible unless there is more to reality than what it reveals to us:
It is not always realised how exceedingly abstract is the information that theoretical physics has to give. It lays down certain fundamental equations which enable it to deal with the logical structure of events, while leaving it completely unknown what is the intrinsic character of the events that have the structure. We only know the intrinsic character of events when they happen to us. Nothing whatever in theoretical physics enables us to say anything about the intrinsic character of events elsewhere. They may be just like the events that happen to us, or they may be totally different in strictly unimaginable ways. All that physics gives us is certain equations giving abstract properties of their changes. But as to what it is that changes, and what it changes from and to—as to this, physics is silent. (My Philosophical Development, p. 13)
Moreover, physics’ tremendous success at prediction and technological application is precisely the result of its deliberate neglect of any aspect of reality that does not fit its mathematically-oriented methods. Early modern thinkers like Bacon and Descartes sought to reorient science in a practical, this-worldly, technological direction. Mathematics facilitated this; aspects of the world that couldn’t be mathematically modeled were a distraction. Hence they were relegated to the status of mere “secondary qualities,” or treated as features that are the proper study of metaphysics rather than physics. That was less a metaphysical discovery, though, than a methodological stipulation. If you set out to study only those aspects of reality that might be rigorously predictable and controllable, then you are bound to find that those are the only ones you discover. But it is preposterous to pretend that you have thereby shown that there are no other aspects of reality, just as it would be preposterous for the “metallicist” to pretend that his exclusive focus on those objects that might be detected electromagnetically shows that there are no non-metals. (See The Last Superstition for more detailed discussion of this theme.)
What Rosenberg and others beholden to scientism have done, then, is simply to confuse method with metaphysics (an occupational hazard of post-Galilean science and post-Cartesian philosophy, as E. A. Burtt warned in his classic book The Metaphysical Foundations of Modern Physical Science). The fallacious blurring of epistemology and metaphysics is, of course, also a feature of many idealist arguments, which is why Stove thought they merited our scorn. All the more appropriately, then, might we label Rosenberg’s argument a “Gem.”
Scientism versus teleology
Among the features of the world physics deliberately ignores for its purposes are those that involve final causality. As Rosenberg writes:
Ever since physics hit its stride with Newton, it has excluded purposes, goals, ends, or designs in nature. It firmly bans all explanations that are teleological… (p. 40)
As the words “exclusion” and “ban” indicate, though, this is, yet again, merely a methodological stipulation. By itself it tells us nothing at all about whether teleology is real. Again, if the designer of a metal detector says “For purposes of metal detection, let’s ignore every feature of the objects we’re after except their electromagnetic properties,” then he is naturally going to pay no attention to whether this or that object is a coin, or a key, or a thumbtack, or even whether it is made of iron as opposed to nickel. But it obviously does not follow that the only real properties of the objects the metal detector finds are their electromagnetic properties, and that we should be eliminativists about coins, keys, thumbtacks, iron, and nickel. Similarly, since teleological features cannot be modeled mathematically, the early moderns – thinkers who, following Bacon and Descartes, wanted to turn science in a practical, this-worldly direction and thus toward a focus on prediction and control – decided to ignore them. But (as it cannot be repeated too frequently) it simply doesn’t follow that such features do not exist.
Rosenberg no doubt thinks an appeal to Ockham’s razor justifies such an inference. He writes:
Since Newton 350 years ago, [physics] has always succeeded in providing a nonteleological theory to deal with each of the new explanatory and experimental challenges it has faced. That track record is tremendously strong evidence for concluding that its still-unsolved problems will submit to nonteleological theories. (p. 40)
The implication is that since physics hasn’t ever needed to postulate final causes, we can infer with confidence that it will not need to do so in the future; and if it does not need to do so, the principle of parsimony should lead us to conclude that final causes don’t exist.
But there are several problems with such an argument. For one thing, Rosenberg’s main reason for denying the existence of teleology, plans, purposes, designs, intentionality, and the like at the biological level and even at the level of the human mind, is that physics has ruled teleology and cognate notions out of science altogether. But in that case an appeal to Ockham’s razor of the sort just considered would lead Rosenberg into a “No True Scotsman” fallacy. He will be saying, in effect: Physics can explain everything that exists without appealing to teleology. So, by Ockham’s razor, teleology must not be a real feature of the world. Of course, biological functions, human thought and action, and the like cannot be understood except in teleological terms. But that just shows that they must not really exist, because teleology doesn’t exist, because physics can explain everything that exists without it!
Another problem is that something like teleology is necessary to explain the facts that physics describes, at least if we regard any of them as embodying genuine causal relations. That is, in any event, the view of a number of contemporary philosophers of science and metaphysicians – George Molnar, C. B. Martin, John Heil, and other “new essentialist” writers – who have no theological ax to grind, but who regard dispositions as “directed at” their manifestations and thus as exhibiting what Molnar calls a kind of “physical intentionality.” This is (as historian of philosophy Walter Ott has noted) essentially a return to an Aristotelian-Scholastic understanding of final causality as a precondition of the intelligibility of efficient causality. Unless we suppose that an efficient cause A inherently “points” beyond itself to its typical effect (or range of effects) B as toward an end or goal, we have no way of making sense of why it is that A reliably does in fact generate B rather than C, D, or no effect at all.
Rosenberg doesn’t see the possibility of such a view because he has only the crudest conception of teleology -- he evidently thinks that a teleological explanation is one that simply postulates that “God designed it that way.” No one familiar with the Aristotelian and Scholastic traditions would make such a mistake, though someone who supposes that teleology and natural theology stand or fall with Paley-style “design arguments” is likely to. (As I have noted before, Rosenberg’s knowledge of natural theology seems to derive mostly from whatever was in the anthology his undergrad PHIL 101 teacher was using.)
Rosenberg also supposes that the second law of thermodynamics is incompatible with the existence of teleology. For “the second law tells us that the universe is headed to complete disorder” (in particular, heat death) and “no purpose or goal can be secured permanently under such circumstances” (p. 41). But the existence of teleology doesn’t require that an end or goal be realized permanently. And insofar as the second law of thermodynamics describes causal regularities -- and in particular a tendency toward disorder -- it would itself be an instance of teleology, not a counterexample to it.
(The subject of teleology is one I have devoted much attention to elsewhere , e.g. in chapter 6 of The Last Superstition, chapter 2 of Aquinas, and in a great many blog posts on the dispute between Aristotelian-Thomistic philosophy and “Intelligent Design” theory. I won’t repeat myself here -- interested readers are directed to these sources.)
So, Rosenberg has no good arguments for scientism, and thus no good arguments either for atheism or for the other, more bizarre conclusions he derives from scientism. As we will see in the remaining posts in this series, some of those conclusions are in any event incoherent, and thus constitute a reductio ad absurdum of the premises that lead to them.
Before turning to these conclusions, though, it will be worthwhile examining Rosenberg’s brief attempt to counter kalam-style arguments for God as the cause of the Big Bang, with some alternative cosmological speculations of his own. We’ll do so in the next post in this series.
[Addendum: A reader calls attention to this critique of Rosenberg by Timothy Williamson, which dovetails with some of the points made above. A key line: “Those most confident of being undogmatic and possessing the scientific spirit may thereby become all the less able to detect dogmatism and failures of the scientific spirit in themselves.”]