[for a brief explanation of this ongoing series, as well as a full table of contents, go here]
Willard Van Orman Quine
“Belief in some fundamental cleavage between truths that are analytic, or grounded in meanings independently of matters of fact, and truths that are synthetic, or grounded in fact” and “reductionism: the belief that each meaningful statement is equivalent to some logical construction upon terms which refer to immediate experience.” These are the famous two “dogmas” that W.V.O. Quine imputed to positivism (Quine 1980: 20), and that he proceeded to dismantle in one of the best examples of progress in contemporary philosophy. As we shall see, the rejection of a sharp distinction between analytic and synthetic truths, as well as the abandonment of the strict logicism of the positivists, do not necessarily amount to the complete abandonment of “first philosophy” (i.e., philosophizing to be done independently of any empirically-driven scientific investigation). Nor does it follow that philosophy blurs into science to the point of subsiding into it, a position not exactly championed by Quine, but to which he came perilously close. Regardless, one cannot talk about progress in philosophy, and especially about the naturalistic turn, without taking Quine seriously.
I do not have the space to get into an in-depth analysis of the history of positivism and the reactions it engendered by the likes of Quine, Putnam and others (see, for instance, chapters 3-6 of Brown 2012). Nonetheless, the transition from positivism to Quineian post-positivism is a very good example of what I am arguing for in this book: philosophy makes progress by exploring, refining and sometimes rejecting positions in the vast conceptual space pertinent to whatever it is that certain philosophers are interested in, in this case the nature of epistemology and the foundations of knowledge. The positivists did indeed make far too much of the analytic/synthetic distinction, which in turn leads to the notion that certain types of knowledge are possible a priori, and that therefore there is ample room for an ambitious kind of “first philosophy.” They also went too far in their peculiar form of “reductionism,” an approach to meaning that excluded not just a lot of philosophy, but even a good chunk of science from consideration, on the ground that it is (allegedly) literally meaningless. But the positivists themselves were attempting to make progress by reacting, among other things, to the excesses of a metaphysics that sounded either Scholastic or obscure (e.g., their sharp criticisms of Heidegger). It seems to me that as a result of positivism we do indeed have a number of tools to question speculative metaphysics, and even excessively speculative science (e.g., string theory: Smolin 2007); and as a result of Quine’s criticism of positivism we have the outline of a naturalistic epistemology, metaphysics, and indeed philosophy. What we do not have, however, is the collapse of epistemology, metaphysics, and philosophy into science — pace the bold pronouncements of the sort of philosophically less than literate scientists we have encountered in the chapter on Philosophy’s PR Problem.
Quine left no doubts about what he was up to in a late reconstruction of the goals of his own work: “In Theories and Things I wrote that naturalism is ‘the recognition that it is within science itself, and not in some prior philosophy, that reality is to be identified and described’; again that it is ‘abandonment of the goal of a first philosophy prior to natural science’” (Quine 1995, 251). I will get back to what exactly it may mean to abandon the idea of a first philosophy prior to science, but it is crucial to point out that Quine immediately proceeded to “cheat” somewhat about the actual scope of his project, for instance by allowing mathematics to be treated as a science (p. 252). This is somewhat odd, or at the least controversial, considering that if there is anything that defines science it is its concern with empirical evidence about the nature of the world, and mathematics most certainly does not share this basic concern (the Pythagorean theorem is about abstract triangles, not about the clumsy variety we may actually trace on the ground or on a piece of paper). Realizing this, Quine shifted the focus to applied mathematics (p. 257), which, however, does not improve things very much, since applied mathematics is only a relatively small portion of what mathematicians actually do, and at any rate it still is not derived from (although it does apply to) the empirical realm. Mathematics thus presents a problem for Quine’s overarching denial of the existence of a priori truths. His response is ultimately to argue that mathematics is justified by the role it plays in science, and therefore by experience. But that justification seems to get things backwards, and it certainly would surprise the hell out of mathematicians and most philosophers of mathematics (e.g., Brown 2008).
In his analysis of Quine’s contribution to contemporary philosophy Hylton (2010) points out that the full force of Quine’s critique of analyticity is not understood if one focuses on standard examples of the latter, like the ubiquitous “All bachelors are unmarried.” Rather, one has to consider sentences like “Force equals mass times acceleration.” Indeed, referring to examples such as the one about bachelors, Quine himself (1991, 271) says “I recognize the notion of analyticity in its obvious and useful but epistemologically insignificant applications.” But rejecting standard examples of analytical truths such as definitions on the grounds that they are “epistemologically insignificant” begs the question of what, precisely, makes a given statement epistemologically “significant.”
And this brings us back to math: for Quine math is certainly not epistemologically insignificant, which is why his example of F = ma is interesting, particularly in light of his ongoing disagreement with logical positivists (and later the logical empiricists), especially Carnap (1937), who had a lot to say about both the analytic/synthetic distinction and the status of physical laws such as Newton’s. Now, F = ma can be interpreted in a variety of ways, including as a definition (of force, or, rearranging the equation, of either mass or acceleration), but it does not snugly fit the classic conception of analytic truth. That’s because one can argue that the equation is only true within a particular empirically-based theory of the natural world, from which it derives the meaning of its constituent terms (“F,” “m” and “a”). Its truth is not rooted in mathematical reasoning per se. Indeed, it could be argued that F = ma should not even be treated as a definition of force, but rather as an operational way to measure force. There is no explicit conceptual content in the equation, which in itself is compatible with different ideas of what force, mass and acceleration are, as long as they remain related in the way in which the equation connects them. It seems like Quine was focusing on examples like F = ma because they are the sort of statement that may have looked analytical (since it is expressed in mathematical language), but is actually closer to philosophers’ understanding of synthetic truths. This narrow focus, however, may exclude too much, somewhat undermining Quine’s bold claim that there is no such thing as a priori truths. But if the latter re-enter the game — however qualified and circumscribed — then some kind of first philosophy cannot be far behind.
A related issue here is that Quine does not admit the existence of necessary truths, a negation that would be yet another nail in the coffin of pretty much the entire enterprise of (first) philosophy, at least as classically conceived. Quine, of course, arrived at this view because he was what some have termed a “radical” empiricist, and if there is one thing that empiricists abhor is the very idea of necessary truth. Indeed, for Quine even logic was — at least potentially — on the chopping block of his version of a naturalized philosophy. But another major philosopher of the 20th century, Kripke (certainly not a naturalist ), argued shortly thereafter that there is a new way of conceiving of necessary truths: in modal logic, these become truths that hold in all possible worlds (Kripke, 1980). The caveat with Kripke’s reintroduction of necessary truths is that they turn out to be so a posteriori, as in the famous example of whether water is necessarily H2O, something that can be answered only by science, and therefore on empirical grounds. A posteriori necessary truths are controversial in philosophy, but for the purpose of our current discussion they count as a type of necessary truth, and of course they do not exclude the possibility of more standard, a priori necessary truths anyway. Indeed, Kripke insisted that he was making an ontological point about the existence of a priori truths; how we find out about them (scientific investigation = a posteriori, philosophical reasoning = a priori) is an epistemological issue. Considering again F = ma, Kripke’s point would be that the equation, if expressed as an identity statement, would be both necessarily true and known a posteriori.
The broader context of this discussion is provided by Quine’s views of metaphysics and epistemology, which are in turn related to his idea of knowledge as a “web of beliefs,” a metaphor that I very much like, with some caveats. Let us begin by tempering the web-of-beliefs metaphor the way the master himself did: “It is an uninteresting legalism … to think of our scientific system of the world as involved en bloc in every prediction. More modest chunks suffice, and so may be ascribed their independent empirical meaning, nearly enough, since some vagueness in meaning must be allowed for in any event” (Quine 1960, 71). This admission may be somewhat surprising, and indeed Fogelin (1997) uses it effectively as part of his argument that Quine’s naturalism was of a more limited scope than is commonly understood. The Quineian holistic thesis, it seems, is to be taken with a large grain of salt, as a logically extreme possibility (a “legalism,” albeit an “interesting” one), but in practice we need to limit ourselves to examine only local portions of the web of beliefs at any given time, taking much of the background for granted, at least until further notice.
The other pertinent caveat made explicit by the above quoted passage pertains to Quine’s critique of the logical positivists’ distinction between synthetic and analytic truths that we have just explored. That critique is based on the idea that the distinction (one of the two “dogmas”) deploys terms whose meaning is insufficiently clear. But as Hylton (2010) points out, critics have remarked that Quine’s standards for clarity and adequacy are themselves not clear and possibly artificially high. From the point of view of a web of knowledge, the meaning of the terms used by the logical positivists cannot be understood in isolation, but requires a holistic approach. The problem is that if one pushes holism too far one gives up on the idea of meaning altogether, as Quine himself realized.
From epistemology back to metaphysics. According to Fogelin (1997) Quine began by admitting a fairly broad ontology, but became increasingly committed to physicalism (by about 1953), which was “whole-hearted except for the grudging admission of a few, seemingly unavoidable, abstract entities” (Fogelin, p. 545). Quine did allow — in principle — things like the “positing [of] sensibilia, possibilia, spirits, a Creator,” as long as they carried the same sort of theoretical usefulness as quarks and black holes (Quine 1995, p. 252). Analogously, E. Nagel (1955) wrote that “naturalism does not dismiss every other differing conception of the scheme of things as logically impossible; and it does not rule out all alternatives to itself on a priori grounds” (Nagel 1955, 12). Early on Quine even entertained (and ultimately abandoned) an ontology that used only the set of space-time points, i.e. an ontology of entirely abstract entities, something that nowadays would be considered an extreme form of structural realism of the type defended by Ladyman and Ross (2009; we’ll get back to them later on). Quine went on to articulate what he called a “regimented” theory that contains no abstract objects other than sets (his famous “desert” ontology). As Hylton (2010) points out, however, sets can be used to define a wide range of abstracta, only some of which are acknowledged by Quine (e.g., numbers, functions, and mathematical objects in general). Quine excluded propositions and possible entities from his list of admitted abstracta, on the ground that the identity criteria in the latter cases are “unclear.” But as I mentioned earlier, Quine’s own criteria for including or excluding something from his ontology were themselves not very clear.
The bottom line is that for Quine metaphysics is metaphysics of science, because science is pretty much the only game in town when it comes to grounding our beliefs about reality. It then naturally follows from this position that epistemology is just psychology, as he famously stated, a conclusion that has seen some push back since, as also evidenced by the empirical fact that epistemologists have not migrated en masse into Psychology departments.
It is worth remembering that Quine did not understand scientific knowledge as different from ordinary knowledge (Hylton 2010), which means that his position can be construed as different from blatant scientism (Sorell 1994): the latter is about reducing everything worth investigating to science, so that philosophical questions become either irrelevant or scientific. It would be more accurate to say that for Quine there was little if any distinction between science and philosophy because both, when done correctly, were in turn indistinguishable from (sound) ordinary knowledge. Indeed, he wrote (Quine 1995, 256) “Is this sort of thing still philosophy? Naturalism brings a salutary blurring of such boundaries.” Blurring boundaries is not at all the same as collapsing philosophy into science, as some more aggressive contemporary naturalistic philosophers are either explicitly advocating or implicitly endorsing (e.g., Alex Rosenberg in the first group, and perhaps the more recent writings by Dan Dennett in the second).
However we want to reconstruct Quine’s project — something that as any Quine scholar will readily testify is certainly open to a variety of interpretations — it was supposed to retain the philosophically crucial normative aspect of epistemology: “Naturalistic epistemology … is viewed by Henri Lauener and others as purely descriptive. I disagree. Just as traditional epistemology on its speculative side gets naturalized into science, or next of kin, so on its normative side it gets naturalized into technology, the technology of scientizing” (Quine 1995, 258). But it is not at all clear on what scientific or technological grounds one can move from descriptive to prescriptive epistemology.
Let me bring up a simple example to make the point a bit more clearly. There is a lot of talk these days about how recent discoveries in cognitive science are rendering the study of philosophically based critical thinking and informal logic obsolete. For instance, experimental psychologists have now documented the existence of a number of ingrained cognitive biases, from the tendency to confuse correlation and causation to the confirmation bias (ignoring evidence contrary to one’s own beliefs and accepting evidence supporting them), and many others. Interestingly, cognitive biases tend to map with well known formal and informal logical fallacies, as they have been analyzed by philosophers and logicians for some time. The difference between the psychologist and the philosopher here is precisely that the first one describes the problem empirically, while the second one prescribes the solution logically. The discoveries made by cognitive science actually make it even more important that one study logic, not less. To argue that the psychology somehow supersedes the philosophy would be like suggesting that since many people are really bad at estimating probabilities (thanks to which phenomenon the gambling industry thrives), therefore we should stop teaching probability theory in statistics courses. On the contrary! It is precisely because, empirically speaking, human beings are so bad at reasoning that one needs to emphasize the prescriptive aspect of theoretical disciplines like probability theory and logic (besides, without the latter two fields, how would psychologists even know that people are getting things wrong?). Depending on how exactly one reads Quine, he may have been perfectly fine with the distinction I have just drawn, but I am worried by some authors being more Quineian than Quine these days, which easily leads not just to a “salutary” blurring of boundaries between science and philosophy, but comes close to an outright selling out or dismissal of the philosophical enterprise (e.g., Rosenberg 2011; some of the literature on experimental philosophy that we will take on in the last chapter).
According again to Hylton (2010), one of Quine’s revolutionary steps was to apply naturalism to naturalism, arguing that the reason to believe that natural science provides us with the best way to understand the world is natural science itself. This may sound like straightforwardly circular reasoning, but it would be so only if one were to look for a “foundation” to the edifice of knowledge. If instead one does away with foundational projects altogether and substitutes them with the concept of a web of belief, one does arrive at an intricate — and I would argue more realistic and useful — picture in which science, philosophy, mathematics, logic and “ordinary knowledge” all grade into each other, and all influence each other. Even so, we have seen earlier that Quine himself did not take the metaphor of a web of belief too far (cfr. his comment on “legalism”). What then emerges from a reasonably moderate reading of the Quineian critique of positivism is that the web of belief is underpinned by a number of partially distinct yet overlapping approaches, the resulting patchwork being reflected in the prima facie distinctions we do make among philosophy, science, mathematics, logic and common knowledge. The blurring of disciplinary boundaries is then salutary because it encourages dialogue and cooperation. But altogether ignoring the existence of such boundaries (blurry as they may be) throws the baby out with the bath water and encourages a rapid slide into scientism. In a sense, for me the best response to a strong reading of Quine is that a scientist (or a mathematician, let alone a common person) would most certainly not recognize Quine’s own writings as scientific (or mathematical, or as instances of common knowledge). But no philosopher — whether he disagrees with Quine or not — would have difficulty in recognizing them as philosophy.
 As Kripke himself put it: “I don’t have the prejudices many have today, I don’t believe in a naturalist world view. I don’t base my thinking on prejudices or a worldview and do not believe in materialism.” Quoted in “Saul Kripke, Genius Logician,” David Boles Blogs, 25 February 2001.
Brown, J.R. (2008) Philosophy of Mathematics: A Contemporary Introduction to the World of Proofs and Pictures. Routledge.
Brown, J.R. (ed.) (2012) Philosophy of Science: The Key Thinkers. Continuum.
Carnap, R. (1937) The Logical Syntax of Language. Kegan Paul, Trench, Trubner & Co.
Fogelin, R.J. (1997) Quine’s limited naturalism. The Journal of Philosophy 94:543-563.
Hylton, P. (2010) Willard van Orman Quine. Stanford Encyclopedia of Philosophy (accessed on 19 December 2012).
Kripke, S.A. (1980) Naming and Necessity. Blackwell.
Ladyman, J. and Ross, D. (2009) Every Thing Must Go: Metaphysics Naturalized. Oxford University Press.
Nagel, E. (1955) Naturalism reconsidered. Proceedings of the American Philosophical Association 28:5-17.
Quine, W.V.O. (1960) Word and Object. MIT Press.
Quine, W.V.O. (1980) From A Logical Point of View. Harvard University Press.
Quine, W.V.O. (1995) Naturalism; or, living within one’s means. Dialectica 49:251-261.
Rosenberg, A. (2011) The Atheist’s Guide to Reality: Enjoying Life without Illusions. W.W. Norton & Company.
Smolin, L. (2007) The Trouble With Physics: The Rise of String Theory, The Fall of a Science, and What Comes Next. Mariner Books.
Sorell, T. (1994) Scientism: Philosophy and the Infatuation with Science. Routledge.
A thought for the night;
Science requires a commitment of the mind, but objectivity of the heart.
While art requires a commitment of the heart, but objectivity of the mind.
From those whose professions are a necessary art, but impossible science.
To those who think all is science.
Robin, good point about not everything about logical positivism being wrong.
LikeLiked by 1 person
Were the positivists really so bad? As discussed above, Quine was very interested in their ideas, and takes many of their stances to be sensible. Personally, I would be closer to naive realism, but the agnostic view of Carnap (his defenders argue that he didn’t think metaphysics impossible, just not doable at the time) is that we don’t yet have a framework that can discriminate between instrumentalism and realism. And he personally progressed, discarding syntacticism after interacting with Tarski.
Hintikka (in Bonk 2003) comments:
Specifically, he thinks that “distributive normal forms in first order language” are
“an incontrovertible counter-example to Quine”, and that “Quine’s rejection of the analytic-synthetic distinction has been a major disaster in philosophical methodology.”
His other comment is on the “web of knowledge” w.r.t. scientific thinking.
“Scenario 1: mathematicians arrived at a conceptual framework of “natural numbers” because they matched real-world behaviour and allowed us to count and were useful.”
You keep doing the same thing, and missing the same point. Nobody has ever argued that mathematics didn’t start with real world problems. But it has gone so far beyond them that to keep saying that math is empirically grounded and therefore a science is, in my mind, bizarre. But I’m not going to convince you (or vice versa), so I’ll drop it, again.
“I’m not so sure about that. I mean, it depends what we mean by “water”. These days, H2O is baked into the definition of water”
Of course, but we have arrived at that definition empirically, since ancient people didn’t know chemistry, and yet knew water.
Socratic, Robin, david,
yes, people these days have a knee-jerk reaction against logical positivism, and that’s too bad, because those were smart people who made a long and sustained effort that certainly contributed to progress in philosophy of science.
> Of course, but we have arrived at that definition empirically, since ancient people didn’t know chemistry, and yet knew water.
Right. I think you miss my point, though. It’s not really an a posteriori necessity if it just depends on the definition. I could define “water” as “H50” and then “water” would be a compound of five hydrogens joined to an oxygen despite the fact that (like unicorns) there is no such compound.
> but my experience with many of the actors here has taught me that no amount of discussion on my part is going to do any good.
I really wish you didn’t always paint a failure to reach agreement as obstinacy on the part of those who disagree with you.
LikeLiked by 3 people
I don’t think I missed your point. If we are simply talking about definition (as opposed to metaphysical necessity) then of course water must be H2O, but that’s entirely uninteresting, philosophically.
> then of course water must be H2O, but that’s entirely uninteresting, philosophically.
And that’s my point. We are either insist on it being necessary, in which case we are just talking about definitions (which is philosophically uninteresting) or we are just talking about an empirical finding, in which case it is not necessary. In either case, we are not dealing with a separate kind of a posteriori necessary truth, as Kripke would have us believe.
That water is H2O is only necessary given the definition of water as H2O. We arrived at the definition empirically, OK, but our empirical findings were not necessary (we could have found otherwise in another possible world), so the original statement is not necessary unless one treats it as being about definitions.
Anyway, that’s how it seems to me. I’m probably missing something.
LikeLiked by 1 person
“We are either insist on it being necessary, in which case we are just talking about definitions (which is philosophically uninteresting) or we are just talking about an empirical finding, in which case it is not necessary. In either case, we are not dealing with a separate kind of a posteriori necessary truth, as Kripke would have us believe”
There is a third option, that of a metaphysical necessity, which is neither a simple matter of definition nor an empirical question. Now, I don’t believe that there is any such thing as a metaphysical necessity, as distinct from either logical or physical necessities, but that’s what at issue.
> There is a third option, that of a metaphysical necessity,
Yes, that is a third possibility. Good point. It could be metaphysically necessary that water is H2O, meaning that for some metaphysical reason there is no possible world where water (wet, transparent, drinkable stuff) could be anything else. But we don’t know that, even a posteriori. So this third possibility doesn’t seem to me to be what Kripke is talking about, but then I haven’t read the book so I could have the wrong end of the stick on that one.
LikeLiked by 1 person
I believe that’s exactly what Kripke is talking about: a necessary, a posteriori, truth.