Evolution debunks metaphysics but not science?

This is the second entry about Ladyman and Ross' Every Thing Must Go (2007). As we saw, the authors accuse contemporary metaphysics of trying to preserve intuitive, common-sense beliefs about the world, despite the fact that these have no basis in science. They recommend a thoroughly naturalistic metaphysics, which ignores scientifically unanswerable questions and contents itself with drawing connections between different scientific fields.

One motivation for their criticism of contemporary metaphysics is the many examples of beliefs commonly held by metaphysicians (past and present) on the basis of intuition which are contradicted or at least made very implausible by current science. But this is of course not enough to indict the entire metaphysical enterprise. Metaphysicians of all stripes generally agree that their theorizing should be constrained by the findings of our best scientific theories, even if this makes their results more counterintuitive. All that follows from the fact that metaphysicians often believe things that contradict cutting edge science is that they should be more cautious and more scientifically literate. Admittedly, if the intuitions of metaphysicians almost always turned out to be false, this would give us good inductive reason to distrust them completely. But to show this would be very difficult, the authors' examples notwithstanding.

Fortunately, Ladyman and Ross put forward another, much stronger motivation for their anti-metaphysical stance. Our minds, according to standard evolutionary theory, are the products of a process that involved the random generation of traits and selective retaining of those traits that help the organisms that have them to survive and reproduce. We have the cognitive faculties that we have because they made our prehistoric ancestors reproductively successful. This, the authors tell us, gives us good reason to trust our minds where simple beliefs about the mid-sized objects of our immediate environment are concerned. After all, our ancestors wouldn't have survived very long if their beliefs about their immediate environments were mostly false, as they would fall into crevasses and get eaten by tigers. However, the authors go on to say that

proficiency in inferring the large-scale and smallscale structure of our immediate environment, or any features of parts of the universe distant from our ancestral stomping grounds, was of no relevance to our ancestors’ reproductive fitness (p.2)

That is, the ability to acquire truths of the sort sought by metaphysicians would have had no bearing on our ancestor's chances of survival and reproduction. Knowing the properties of the objects surrounding them would certainly have helped (is this berry poisonous?) but knowing what it means for an object to have properties certainly wouldn't have. To put it bluntly, evolution by natural selection 'doesn't care' about metaphysics. Ladyman and Ross infer from this that we should disregard our common-sense intuitions about the world where metaphysics is concerned (though they may be trustworthy where 'everyday' questions are concerned), since the evolution of our minds was not at all aimed at metaphysical ability.

Since this is not a new argument, the authors already have objections to deal with. E. J. Lowe (2002, p.6) argues that this line of thinking undermines science just as much as it undermines metaphysics, because

"It is equally mysterious how an evolved creature should have any capacity to acquire knowledge of such arcane matters as the formation of stars or the structure of DNA"

Of course, the natural answer here is that we have the capacity to overcome our natural limitations through instruments and mathematics. But this just pushes the problem further back: Lowe might as well have said that it is equally mysterious how an evolved creature should have any capacity to create instrumental and mathematical methods that result in scientific knowledge.

Similarly, Ted Sider (2001, xv.) concedes that metaphysics has no good epistemological foundation (i.e. there are no good answers to the question, 'why think that the methods used by metaphysicians are reliable or justified?'), but argues that in this respect, science and mathematics are 'companions in the guilt', since these are likewise foundationless.

Both Lowe and Sider's responses consist in saying that whatever evolution-inspired epistemic objections can be levelled at metaphysics can also be levelled at science. This would be bad news for the project laid out in Everything Must Go, which calls for a rejection of metaphysics as it is currently practiced in favor of a metaphysics that merely unifies the claims of science.

Ladyman and Ross respond that there is a deep asymmetry between science and metaphysics with respect to the evolutionary problem. Science, unlike metaphysics, is "vindicated" (p. 7) by its impressive predictive success. Science has "borne fruits of great value", whereas metaphysics "has achieved nothing remotely comparable" (p.16). This is a variant of the well-known 'no miracles' argument for the truth of our best scientific theories. We can trust the methods used by scientists, because their reliability best explains how successful their products are at predicting nature (how could we be so good at predicting the weather if our climatological theories were all dead wrong?). In contrast, metaphysical theories don't enjoy such predictive success, because their nature is such that they predict nothing. It is impossible to devise predictions that can be used to test, say, mind-body dualism or platonism about numbers. Hence, metaphysics, unlike science, is not vindicated.

This response to the evolutionary problem for science is, in my view, ineffective. Philosophers of science generally agree that scientific theories are formed using inference to the best explanation (or abduction). This mode of inference involves thinking up explanations for our evidence, identifying the one that best explains the evidence, and inferring that the best explanation is in fact true. For instance, scientists tell us that the molecular structure of water is H2O, because this best explains many of water's properties, like the fact that it boils and freezes at certain temperatures. To claim that scientific methodology is trustworthy, then, entails the claim that inference to the best explanation is trustworthy. Now, recall Ladyman and Ross' answer to the evolutionary problem: we have good reason to trust scientific methodology, because its reliability best explains the predictive success of our theories. In other words, Ladyman and Ross defend the reliability of inference to the best explanation using inference to the best explanation! This is clearly a circular response to the evolutionary problem, like arguing that the Bible is trustworthy because it says that it is trustworthy.

Ladyman and Ross could respond that circular arguments aren't always unsound. But in that case, the defender of contemporary metaphysics has an easy answer to the evolutionary problem as well: he can just say that common-sense intuitions about the world are reliable, because it is intuitively true that they are. This is obviously a circular response, but there is no reason to regard it as less legitimate than the previous one. In this regard, metaphysics and science stand or fall together.

Alternatively, Ladyman and Ross could object that the inference to the best explanation involved in their response to the evolutionary problem is somehow not the same kind as inference as the one used in science. They could do this by distinguishing philosophical from scientific inference to the best explanation. But even if there is such a distinction, I think this would entail abandoning the project of advancing a thoroughly naturalistic philosophy. Recall, the purpose of Everything Must Go is to argue for a conception of metaphysics -- and, more generally, of philosophy -- on which the latter does nothing more than show how different scientific subfields (e.g. physics and biology) relate to each other. But if philosophical inference to the best explanation shows that scientific methodology is reliable, then clearly philosophy's job description is much more extensive than that.

I conclude, in disagreement with Ladyman and Ross, that if evolution 'debunks' metaphysics, it must debunk science as well.

References:

Ladyman, J., & Ross, D. (2010). Every Thing must go. Oxford: Oxford Univ. Press.

Lowe, E. (2002). A survey of metaphysics. Oxford: Oxford University Press.

Sider, T. (2001). Four-dimensionalism. Oxford: Clarendon Press.

Is metaphysics just comforting fiction?

In their provocatively titled Every Thing Must Go, (2007), philosophers of science James Ladyman and Don Ross argue that metaphysics as it is practiced in much of contemporary analytic philosophy is ultimately a wrongheaded and worthless pursuit. Metaphysics takes itself to be the branch of philosophy that tackles the most fundamental questions -- 'what are the most basic constituents of reality?' 'To what extent do our concepts reflect the real structure of the world?' 'How does mind relate to matter?' and so on. Unfortunately, Ladyman and Ross argue that it will not be up to the task unless metaphysicians radically reconceive their methods. In particular, they should stop relying on common-sense intuitions, especially when these clash with our best physical theories. The authors provide numerous examples of careless metaphysicians implicitly or explicitly relying on scientifically outdated assumptions that have more to do with 'A-level chemistry' than cutting edge physics. Contemporary metaphysics, we are told, must be replaced by a thoroughly naturalistic metaphysics, which exclusively concerns itself with drawing connections between the sciences, and rejects scientifically unanswerable questions as pointless distractions.

I will not attempt anything like a full-scale review of the book (though I liked this one). I'd much rather examine some of the specific points and arguments made throughout the book. This entry will address the following excerpt, in which the authors quote a well-known introductory metaphysics textbook :

" Michael Loux and Dean Zimmerman explain the methodology of metaphysics as follows: ‘One metaphysical system is superior to another in scope in so far as it allows for the statement of satisfactory philosophical theories on more subjects—theories that preserve, in the face of puzzle and apparent contradiction, most of what we take ourselves to know’ (2003, 5). Here is a conception of metaphysics according to which its function is to reassure the metaphysician that what they already believe is true." (p. 12, emphasis mine)

The textbook states that a good metaphysical theory is one that sits sufficiently well with ('preserves') "what we take ourselves to know", that is, our basic, common-sense intuitions about the world. If a metaphysical theory has a consequence that violates such intuitions, leading us into "puzzle and apparent contradiction", this is regarded as a cost. For example, if a metaphysical theory tells us that everyday material objects like chairs and tables do not exist, there better be excellent reasons to believe said theory, to make up for the cost of violating our intuitive beliefs about such objects.

Ladyman and Ross retort that, on this conception, metaphysics serves the purpose of comforting us in the conviction that what we already believe is true. We begin with the assumptions that the material objects around us exist, that they have various properties, that they could or couldn't have existed with some of those properties, and so on, and construct theories that simply confirm those assumptions. In contrast, the sciences routinely (and sometimes dramatically) challenge our unwarranted prejudices about the world, like the belief that all facts have causes (rejected by current physics) and that 'human being' is a real category (thought to be incompatible with evolutionary theory).

Granted that metaphysicians typically try to make their theories 'fit' with their intuitive, common-sense beliefs about the world. Often, such beliefs are so intuitive because their truth is a matter of definition -- e.g. it just follows from of our concept of pain that it has to 'feel' a certain way. In other cases, common-sense beliefs are held because their truth is thought to be a necessary condition of any kind of objective inquiry. For instance, our investigation of the world wouldn't go very far without beliefs that perception, memory and basic logical reasoning are unreliable. One upshot of this is that scientific investigation, no less than its metaphysical counterpart, also relies on common-sense beliefs, but I will address that in another entry.

For now, what matters is that regardless of whether or not our intuitive common-sense beliefs about the world are warranted, the fact that metaphysicians try to devise theories that agree with them does not at all imply that metaphysical theories are just there to reassure us that what we already believe is true -- in other words, the textbook's characterization of metaphysics in no way implies Ladyman and Ross' characterization of it. What it does imply is that metaphysical theories aim for coherence, in particular, coherence with background beliefs. For example, if I believe that my wife is a human being, I will not try to explain her behavior in ways that would entail (or make very probable) her non-humanity. If she comes home later than usual on a Friday night, I will not explain this with the hypothesis that she was attending a party strictly reserved for intergalactic vampires, but rather with the hypothesis that she is at the pub with colleagues. It would be senseless here to accuse of me of trying to comfort myself in my belief that she is human.

Interestingly, much of ethics works like this as well. Ethicists will usually reject ethical theories that entail clearly false moral claims, e.g. the claim that torturing people for fun is morally permissible. But such ethicists aren't generally aren't denounced for trying to convince themselves that what they already believe is true.

Hence, it seems much more reasonable to say that on the textbook account, metaphysics serves to provide theories which cohere well with (many of) our intuitive common-sense beliefs about the world, not to reassure us that such beliefs are true. Again, this is true regardless of whether or not our intuitive common-sense beliefs are warranted (I will discuss that in another entry).

Defenders of Ladyman and Ross might respond that I am taking them too literally, and that the above quote is more of a tongue-in-cheek quip than a serious criticism of analytic metaphysics. After all, they do not really elaborate their point on metaphysics as comforting fiction, though they make the interesting claim that metaphysics' reluctance to violate intuition draws it closer to "story-writing" than to science (p.13). They spend much more time arguing that the intuitive common-sense beliefs that metaphysics tries to preserve are unwarranted.

That may well be true. But in any case, it seems useful to highlight the difference between the purpose of cohering well with our common-sense beliefs and that of reassuring us that such beliefs are true (many apologies to those for whom this distinction is obvious). At the very least, doing so defends metaphysical endeavors from a natural worry of the kind sometimes expressed by interested bystanders. Moreover, this is a good opportunity to underwrite the crucial importance of coherence to metaphysics, which, arguably, is fundamentally about the construction of a coherent world view.

Ladyman, J. and Ross, D. (2010). Every Thing must go. Oxford: Oxford Univ. Press.