Ed Feser responds to Sean Carroll

Ed Feser has a recent piece here.  I reproduce it below with comments interpolated.

People have been asking me to comment on the remarks about causation made by atheist physicist SeanCarroll during his recent debate with William Lane Craig on the topic of “God and Cosmology.”  (You’ll find Craig’s own post-debate remarks here.)  It’s only fair to acknowledge at the outset that Carroll cannot justly be accused of the anti-philosophical philistinism one finds in recent remarks by physicists Stephen Hawking, Lawrence Krauss, and Neil deGrasse Tyson.  Indeed, Carroll has recently criticized these fellow physicists pretty harshly, and made some useful remarks about the role of philosophy vis-à-vis physics in the course of doing so.
It is also only fair to note that, while I have enormous respect for Craig, I don’t myself think that it is a good idea to approach arguments for a First Cause by way of scientific cosmology.  I think that muddies the waters by inadvertently reinforcing scientism, blurring the distinction between primary (divine) causality and secondary (natural) causality, and perpetuating the false assumption that arguments for a divine First Cause are essentially arguments for a “god of the gaps.”  As I have argued many times, what are in my view the chief arguments of natural theology (i.e. Aristotelian, Neo-Platonic, Thomistic and other Scholastic arguments) rest on premises derived from metaphysics rather than natural science, and in particular on metaphysical premises that any possible natural science must presuppose.  For that reason, they are more certain than anything science itself could in principle ever either support or refute.  Arguments like Aquinas’s Five Ways, when properly understood (as, these days, they usually are not), no more stand or fall with the current state of play in scientific cosmology than they stand or fall with current gastroenterology or polymer research.  (See chapter 3 of Aquinas, my ACPQ article “Existential Inertia and the Five Ways,” my Midwest Studies in Philosophy article “The New Atheists and the Cosmological Argument,” and many other articles and blog posts.  Or, since we’re linking to YouTube, see my lectures “An Aristotelian Proof of the Existence of God” and “Natural Theology Must Be Grounded in the Philosophy of Nature, Not Natural Science.” )

Carroll’s remarks are largely directed at the question of whether scientific cosmologists should regard theism as a good explanation for the sorts of phenomena they are interested in, given the standard criteria by which models in physics are judged.  Since I don’t find that a terribly interesting or important question, I have nothing to say about his criticisms of Craig on that score.

Having said all that, Carroll’s remarks, where they touch on philosophical matters, are pretty shallow, and he does clearly think that what he has to say somehow poses a serious challenge to theism in general, not just theistic arguments grounded in scientific cosmology.  So those remarks are worth a response.  The key passage concerns Carroll’s criticism of Craig’s claim that “If the universe began to exist, then there is a transcendent cause which brought the universe into existence.”  Carroll says:

The real problem is that these are not the right vocabulary words to be using when we discuss fundamental physics and cosmology. This kind of Aristotelian analysis of causation was cutting edge stuff 2,500 years ago. Today we know better. Our metaphysics must follow our physics. That’s what the word “metaphysics” means. And in modern physics, you open a quantum field theory textbook or a general relativity textbook, you will not find the words “transcendent cause” anywhere. What you find are differential equations. This reflects the fact that the way physics is known to work these days is in terms of patterns, unbreakable rules, laws of nature. Given the world at one point in time we will tell you what happens next. There is no need for any extra metaphysical baggage, like transcendent causes, on top of that. It’s precisely the wrong way to think about how the fundamental reality works.  The question you should be asking is, “What is the best model of the universe that science can come up with?” By a model I mean a formal mathematical system that purports to match on to what we observe. So if you want to know whether something is possible in cosmology or physics you ask, “Can I build a model?”

End quote.  Now, it would take a book to explain everything that’s wrong with this.  And as it happens, I’ve written such a book; it’s called Scholastic Metaphysics: A Contemporary Introduction.  Since I’ve already said so much about these issues both in that book and elsewhere, I’m not going to repeat myself at length.  Let me just call attention to the key begged questions, missed points, and non sequiturs in Carroll’s remarks.

Well, of course this is 'wrong' from Ed's Aristotelian point of view.  But there is hardly any point in his saying it.  Carroll simply has a rival metaphysics.  The only sensible criticisms of a metaphysics are internal ones, from consistency, say, or external ones from relative explanatory power, say.  I will try to bring this out as we go through the post.  But it enables Ed to get in a neat puff for his latest book. 

Carroll tells us that explanation in physics proceeds by way of building a “model” that describes a “mathematical system” reflecting “patterns, unbreakable rules, laws of nature.”  Fine and dandy; I’ve pointed this out many times myself.  If Carroll’s point were merely that, to the extent that theism can’t be formulated in such mathematical terms, it just isn’t the sort of thing the physicist will find a useful explanation for the specific sorts of phenomena he’s interested in, then I wouldn’t necessarily have any problem with that.  That’s not what classical theism, properly understood, is all about in the first place.

But Carroll goes beyond that.  When he says that once you’ve hit upon the best mathematical model, whatever it turns out to be, “there is no need for any extra metaphysical baggage… on top of that,” he evidently means not just that you don’t need anything more for the purposes of physics, specifically, but that you don’t need anything more than that, period.  For he says that asking for more is “precisely the wrong way to think about how the fundamental reality works” and that “our metaphysics must follow our physics.”  The idea seems to be that once you’ve answered all the questions in physics, you’ve answered all the questions that can be answered, including all the metaphysical questions.  There’s nothing more to be done, not just nothing more for the physicist to do.

Now, why should anyone believe that claim (which is essentially just a version of scientism)?  Carroll gives no argument for it at all; he just asserts it with confidence.  This is a step down from Alex Rosenberg, who in The Atheist’s Guide to Reality did give an argument for a similar claim -- an argument which, as we saw, is extremely bad, but is at least still an argument. 

I'm not sure Carroll is being scientistic here at all.  It's just that as physics has developed over the past three hundred years, and as it has investigated the behaviour of matter on smaller and smaller scales, the concept of cause has disappeared from the subject.  In its place came, as Carroll says, differential equations.  These merely describe, in a general way, how matter behaves at small scale.  Taking such descriptions as axiomatic, we infer the behaviour of matter at macroscopic scales.  This is now seen as explanatory of macroscopic behaviour and replaces the common sense notion of causation.  The physics of the early twentieth century taught us that the behaviour of matter at the atomic scale simply cannot be understood by projecting the behaviour of macroscopic bodies onto smaller and smaller pieces of matter.  The atomic world is not the everyday world writ very small.  We cannot understand the small in terms of the large.  But we can understand the large in terms of the small.  So physics has abandoned causes.  The metaphysics of cause, central to an earlier Aristotelianism, is redundant.  All that is left is a ghost of formal causation to be found in the differential equations.

Nor could there be a good argument for Carroll’s scientism, because scientism is demonstrably false.  For one thing, “scientism” is more poorly defined than Carroll claims theism is.  However we tighten up our definition of notions like “science,” “physics,” and the like, the resulting scientism is going to be either self-refuting (since it will turn out that scientism cannot itself be established via the methods of physics or any other natural science), or completely trivial (since, to avoid the self-refutation charge, “science,” “physics,” etc. will have to be defined so broadly that even the metaphysical notions Carroll wants to dismiss will count as “scientific”). 

This is not the place to discuss that all-weather put-down,  'scientism'.  I do not concede that Carroll is being scientistic.  In any case, nobody is going to advance a 'demonstrably false' thesis.  That which is demonstrably false is not that which those accused of scientism are advocating.  A topic for another post perhaps. 

For another thing, to suppose that since physics confines itself to mathematical models, it follows that there is nothing more to reality than is captured by such models, is fallaciously to draw a metaphysical conclusion from a mere methodological stipulation.  The problem is not just that, if there are features of reality which cannot be captured in terms of a mathematical model, then the methods of physics are guaranteed not to capture them (though that is bad enough).  It is that there must in fact be more to reality than is captured by those methods, in part because (as Bertrand Russell noted) physics gives us only structure, and structure presupposes something which has the structure and which a purely structural description will of necessity fail to capture. 

This is interesting.  But Ed does not say what aspects of reality he has in mind, so we can't form a judgement as to whether this is relevant to physics or cosmology.  We can think of physics as the dynamics of matter in the small and cosmology as dynamics of matter at the very largest of scales.  If dynamics is the study of where the matter is over time, then it seems ripe for mathematical treatment.  What's the problem?

I develop these points in detail in Chapter 0 of Scholastic Metaphysics.  I also show, in that chapter and throughout the book, that the appeal to “laws of nature” so routinely and glibly made by naturalists like Carroll, simply does not and cannot do the work they suppose it does, and papers over a mountain of begged metaphysical questions.  In fact the very notion is fraught with philosophical difficulty, as writers like Nancy Cartwright and Stephen Mumford have shown.  As I have noted many times, the notion of a “law of nature” was originally (in thinkers like Descartes and Newton) explicitly theological, connoting the decree of a divine lawmaker.  Later scientists would regard this as a metaphor, but a metaphor for what?  Most contemporary scientists who pontificate about philosophical matters not only do not have an answer but have forgotten the question.

Allow me to pontificate.  Whatever 'law of nature' may have meant in the past, the phrase now, if it is used at all, merely refers to the differential equations which are thought to encode the patterns of possible motions.  They thus have the status of axioms.  One can ask, I guess, philosophical questions about axiomatic mathematical and physical theories, but I don't see how the concept of 'divine lawmaker' possibly fits into this, nor do I see a 'mountain of begged metaphysical questions'.  Such questions only arise against a background of metaphysical assumptions such as Ed's Aristotelianism.  Quite sensibly, in my view, Carroll simply refuses to talk this language.

One contemporary scientist who does see the problem is physicist Paul Davies, who, in his essay “Universe from Bit” (in Paul Davies and Niels Henrik Gregersen, eds. Information and the Nature of Reality: From Physics to Metaphysics), writes:

The orthodox view of the nature of the laws of physics contains a long list of tacitly assumed properties.  The laws are regarded, for example, as immutable, eternal, infinitely precise mathematical relationships that transcend the physical universe, and were imprinted on it at the moment of its birth from “outside,” like a maker’s mark, and have remained unchanging ever since… In addition, it is assumed that the physical world is affected by the laws, but the laws are completely impervious to what happens in the universe… It is not hard to discover where this picture of physical laws comes from: it is inherited directly from monotheism, which asserts that a rational being designed the universe according to a set of perfect laws.  And the asymmetry between immutable laws and contingent states mirrors the asymmetry between God and nature: the universe depends utterly on God for its existence, whereas God’s existence does not depend on the universe…

Clearly, then, the orthodox concept of laws of physics derives directly from theology.  It is remarkable that this view has remained largely unchallenged after 300 years of secular science.  Indeed, the “theological model” of the laws of physics is so ingrained in scientific thinking that it is taken for granted.  The hidden assumptions behind the concept of physical laws, and their theological provenance, are simply ignored by almost all except historians of science and theologians.  From the scientific standpoint, however, this uncritical acceptance of the theological model of laws leaves a lot to be desired… (pp. 70-1)

Now the naïve atheist reading this blog for the first time may suppose that at this point I am going to exclaim triumphantly that there cannot be law without a lawgiver and proclaim victory for theism.  But in fact, like Davies I don’t accept the theological account of laws.  I think it is bad metaphysics and bad theology (insofar as it tends toward occasionalism).  I want rather to make the following two points.  First, when scientists like Carroll confidently proclaim that we can explain such-and-such in terms of the laws of physics rather than God, what they are saying, without realizing it, is: “The explanation isn’t God, it’s rather the laws of physics, where ‘law of physics’ originally meant ‘a decree of God’ and where I don’t have any worked-out alternative account of what it means.”  Hence the “alternative” explanation, when unpacked, is really either a tacit appeal to God or a non-explanation.  In short, either it isn’t alternative, or it’s not an explanation.  The utter cluelessness of this stock naturalistic “alternative explanation” would make of it an object of ridicule if it were not so routinely and confidently put forward by otherwise highly intelligent, educated, and widely esteemed people.

Well, I do have an account of what 'law of physics' means.  It means a mathematical description of the behaviour of the simplest constituents of matter.  Just what Galileo sought in his experiments with inclined planes.  Together with principles of composition, such descriptions are explanatory of the behaviour of complex organisations of matter by means of deductive inference.  What, pray, is ridiculous about this?

Second, the original, explicitly theological Cartesian-Newtonian notion of “laws of nature” was intended precisely as a replacement for the Aristotelian-Scholastic metaphysics of nature.  The Scholastics held that the regularities in the behavior of natural phenomena derived from their immanent essences or substantial forms, and the directedness-toward-an-end or immanent teleology that followed upon their having those forms.  In other words, regularities reflected the formal and final causes of things.  The early moderns wanted to get rid of formal and final causes as immanent features of nature, and thus replaced them with the notion of “laws of nature” conceived of as externally imposed divine decrees.  To keep talk of “laws of nature” while throwing out God is thus not to offer an alternative to the Aristotelian-Scholastic view at all, but merely to peddle an uncashed metaphor.  So, whereas Carroll glibly asserts that “now we know better” than the Aristotelians did, what is in fact that case is that Carroll and other contemporary naturalists have not only chucked out Aristotelian metaphysics but have also chucked out the early moderns’ initial proposed replacement for Aristotelian metaphysics, and have offered nothing new in its place.  This is hardly a problem for the Aristotelian; on the contrary, it is a problem for anyone who wants to dismiss Aristotelian metaphysics.

Aristotle has a shallow physics and a deep metaphysics.  We moderns have the reverse.  The sense in which 'we now know better' is that our experience of the world is wider.  Telescopes can see to the ends of the universe and particle accelerators can look inside nucleons.  Neither space, time, nor matter are what they seem to the naked eye.  In one sense contemporary physics is itself a replacement for Aristotle's metaphysics.  For another way, see below.

Like other contemporary Aristotelians, I would say that the right way to interpret a “law of nature” is as a shorthand description of the way a thing tends to operate given its nature or substantial form.  That is to say, “laws of nature” actually presuppose, and thus cannot replace, an Aristotelian metaphysics of nature.  (Again see the discussion of the metaphysics of laws of nature in Scholastic Metaphysics.)  There are other accounts of laws, such as Platonic accounts and Humean accounts, but these are seriously problematic.  Platonic accounts, which treat laws of nature as abstract entities in a Platonic heaven, push the problem back a stage.  To appeal to such-and-such Platonic laws as an explanation of what happens in the world only raises the further problems of explaining why it is those laws rather than some others that govern the world, and what makes it the case that any laws at all come to be instantiated.  Humean accounts, meanwhile, interpret a law as a statement that such-and-such a regularity holds, or would have held under the right conditions.  But in that case an appeal to laws doesn’t really explain anything, but only re-describes it in a different jargon. 

Consider, in light of these points, what Carroll says about causation later on in the debate:

Why should we expect that there are causes or explanations or a reason why in the universe in which we live? It’s because the physical world inside of which we’re embedded has two important features. There are unbreakable patterns, laws of physics -- things don’t just happen, they obey the laws -- and there is an arrow of time stretching from the past to the future. The entropy was lower in the past and increases towards the future. Therefore, when you find some event or state of affairs B today, we can very often trace it back in time to one or a couple of possible predecessor events that we therefore call the cause of that, which leads to B according to the laws of physics.  But crucially, both of these features of the universe that allow us to speak the language of causes and effects are completely absent when we talk about the universe as a whole.  We don’t think that our universe is part of a bigger ensemble that obeys laws.  Even if it’s part of the multiverse, the multiverse is not part of a bigger ensemble that obeys laws.  Therefore, nothing gives us the right to demand some kind of external cause.

End quote.  Now in fact it is Carroll who has said absolutely nothing to establish his right to dismiss the demand for a cause as confidently as he does.  For he has simply begged all the important questions and completely missed the point of the main traditional classical theistic arguments (whether or not he has missed Craig’s point -- again, I’m not addressing that here).  One problem here is that, like so many physicists, Carroll has taken what is really just one species of causation (the sort which involves a causal relation between temporally separated events) and identified it with causation as such.  But in fact, the Aristotelian argues, event causation is not only not the only kind of causation but is parasitic on substance causation.

Carroll's remarks strike me as sensible.  They form, perhaps, the basis of an inchoate theory which sees (efficient) causation as just one way in which humans have attempted to conceptualise relations between objects and events.  A consequence, perhaps, of our evolved capacity to find pattern in the stream of sense data we each receive.  It is thus an element of the manifest but not the scientific image.  The case for seeing final causation this way is much more obviously made.  Material causation is readily assimilable to the scientific image, and so, as I suggest above, is formal causation.  Indeed, we can think of science as a vast elaboration of the notion of formal cause.  The diverse origins of material, efficient, formal and final cause under this view, suggest that they do not fall under a unified concept.

But put that aside, because the deeper problem is that Carroll supposes that causation is to be explained in terms of laws of nature, whereas the Aristotelian view is that this has things precisely backwards.  Since a “law of nature” is just a shorthand description of the ways a thing will operate -- that is to say, what sorts of effects it will tend to have -- given its nature or substantial form, in fact the notion of “laws of nature” metaphysically presupposes causation. 

Only within the Aristotelian metaphysical system.  Under the alternative system---which I guess counts as a metaphysics as it tries to explain why we think about the world along certain lines---the remaining three of the four causes are indeed subsumed under a version of formal causation.

Furthermore, what “allows us to speak the language of causes and effects” has nothing essentially to do with tracing series of events backwards in time.  Here again Carroll is just begging the question.  On the Aristotelian-Scholastic analysis, questions about causation are raised wherever we have potentialities that need actualization, or a thing’s being metaphysically composite and thus in need of a principle that accounts for the composition of its parts, or there being a distinction in a thing between its essence or nature on the one and its existence on the other, or a thing’s being contingent.  The universe, however physics and scientific cosmology end up describing it -- even if it turned out to be a universe without a temporal beginning, even if it is a four-dimensional block universe, even if Hawking’s closed universe model turned out to be correct, even if we should really think in terms of a multiverse rather than a single universe -- will, the Aristotelian argues, necessarily exhibit just these features (potentialities needing actualization, composition, contingency, etc.).  And thus it will, as a matter of metaphysical necessity, require a cause outside it.  And only that which is pure actuality devoid of potentiality, only what is utterly simple or non-composite, only something whose essence or nature just is existence itself, only what is therefore in no way contingent but utterly necessary -- only that, the classical theist maintains, could in principle be the ultimate terminus of explanation, whatever the specific scientific details turn out to be. 

We have come full circle.  From Ed's perspective Aristotelian metaphysical principles must indeed be presupposed by any possible natural science.  But we can equally well start from another point and see such principles as plausible modes of thought.  It is, perhaps, a matter of taste.

Carroll has not only not answered these sorts of arguments (which, again, I’ve only alluded to here -- see the various sources cited above for detailed defense).  He doesn’t even seem to be aware that this is where the issues really lie, and that they have nothing essentially to do with scientific cosmology.  But that’s not entirely his fault.  As I have indicated, in my view too many people (and not just Craig) put way too much emphasis on scientific cosmology where the debate between theism and atheism is concerned.  That just opens the door to objections like Carroll’s, since it makes it sound (wrongly, but understandably) like theism as such is essentially in competition with the sorts of models Carroll pits against Craig.

It seems to me that Ed can not or will not think outside the Aristotelian box.  He is just not open to the possibility of a new metaphysics that subsumes his own. 

That is not, by the way, to knock the kalām cosmological argument.  For (as Craig himself has emphasized) that argument need not appeal to scientific cosmology, but can be defended instead by way of appeal to more fundamental metaphysical premises.  (I have not had much to say about that argument myself because it is in my view less fundamental than the arguments I have focused on -- such as the Five Ways -- and there are, in any case, already many people writing about it.  If you’re looking for a Thomist’s defense of the kalām argument, you can’t do better than the relevant articles on the subject by David Oderberg.)


  1. "The chief arguments of natural theology (i.e. Aristotelian, Neo-Platonic, Thomistic and other Scholastic arguments) rest on premises derived from metaphysics rather than natural science, and in particular on metaphysical premises that any possible natural science must presuppose. For that reason, they are more certain than anything science itself could in principle ever either support or refute."

    This was the part of his post that absolutely had me falling out of my chair.

    Suppose you think science really does presuppose, logically or methodologically, some particular metaphysics. Ok. I find this claim highly dubious, but it's not something only an insane person would say, or someone utterly ignorant of modern science.

    Let us further suppose you think this metaphysics is a Scholastic one. I find it even more dubious that Scholastic metaphysics can even be reconciled (in the sense of being minimally logically consistent) with science as we now have it, much less that it is both true and a necessary precondition. So now I see a conjunction of improbable premises, which must perforce be even less likely than either conjunct.

    But he also claims that his particular interpretation of the Scholastics' particular interpretation of Aristotle's metaphysics is more certain than science itself?!? More certain than heliocentrism, more certain than the germ theory of disease, more certain than the atomic number of hydrogen? The hubris here is breathtaking.

    I can imagine how to engage in argument with people who hold the first two theses. But this last bit is just a "pray tell, Mr. Babbage" moment; I cannot construct a model of another human's thought process that would output this kind of upside-down dogmatism and still be amenable to alteration by means of me throwing evidence and logic at it.

  2. Hello SCG. I confess that it was the passage you quote that goaded me into writing this piece. When I was younger I would have no truck with metaphysics, which I associated with the Scholastic mode of thought that we both find, shall I say, problematic! I now see metaphysics as unavoidable, but my concept of metaphysics has altered. I think of it in a Kantian sense, if that's not too pretentious, as an account of why we think along the lines we do. This account has to be broadly in sympathy with what we know (or think we know) from science. One aspect of this is explaining why Aristotelianism is attractive, for some. Since it is an elaboration of common sense we have to account for the divergence of common sense and science. This is not an attempt to disprove Aristotelianism through logic and evidence. I certainly don't think you will much alter the mindset of an adherent this way. A very great deal, after all, is invested in Aristotelian causation. It's more an attempt to understand why anyone might think that way.

    Thanks for commenting.

  3. Thanks for your reply. I first found this place a few months ago from following your profile link during (if I recall correctly) your comments on Feser's series about Kripkean arguments against machines being able to "really"(TM) think. I've been dipping in periodically. Do you comment anywhere else besides Feser and Vallicella's blogs?

    Despite growing up Catholic, I never found Scholastic metaphysics even very interesting or relevant, so I didn't have any strong feelings about their content one way or the other. But on metaphysics in general, since grad school I've drifted further and further towards the Pragmatist idea that they're all just a series of strictly optional vocabularies we use to tie off the loose ends of whatever our best empirical models of the world happen to be at the time. They are aglets: they contribute nothing to the function of the shoe in the way the sole keeps rocks from pricking us and the tops keep rain from soaking us and the laces keep it all together. They are just a little afterthought we put on at the very tippity-tips because our thoughts might as well come to an end somewhere.

    Feser is trying to tell professional cobblers that they aren't "really" clothing our feet unless we buy his particular brand of aglets. I consider his methodology to be at best pseudo-philosophical. As this post shows, he neither begins nor ends in Doubt. He is interested simply in intoning dogmas at people, and teaching his followers the secret chants and invocations they should repeat by rote, as an exorcist should repeat his ritual to banish one possessed by a demon, or worse, by nonbelief.

  4. I have given up commenting at Ed Feser's blog. It's hard to engage the man himself and I'm too easily seduced into endless wrangling with the acolytes. You're right, Ed comes over as an apologist. Bill, on the other hand, is a seeker. There is a nice discussion going on at the moment---actually it's been running off and on for a number of years---about what might be described as the metaphysics of fiction, but is really about the intricacies of language and logic. Definitely worth a look in.