Positivism and the failed attempt to bury metaphysics (The Return of Metaphysics)

Reading | Philosophy

Dr. Giuseppina D’Oro | 2022-11-20

It would seem fair to say that today metaphysics is thriving. In the philosophy of mind, panpsychists argue that the nature of reality is not quite how it is thought of by physicalists; in contemporary analytic metaphysics, debates concerning the nature of time are all the rage. This was not the case in the first half of the twentieth century, when logical positivism mounted one of the most scathing attacks on the very idea that the nature of reality could be known by reflection alone, a priori, from the so-called philosophical armchair. Logical positivism sought to put an end to what it regarded to be irresolvable metaphysical pseudo-disputes by arguing that genuine knowledge claims must be verifiable; that there must be, at least in principle, evidence that can be cited to determine whether a claim is true or false. Claims that cannot be found to be either true or false in this way—the argument goes—express meaningless propositions, and the treatises in which they are contained should be confined to the flames, just as Hume suggested.

Logical positivism, however, failed genuinely to leave metaphysics behind. Rather than doing away with the idea that knowledge of pure being is possible, it merely placed natural science in the privileged epistemic position once occupied by philosophical reflection as a presuppositionless form of knowledge capable of disclosing reality in itself. As Collingwood argued in An Essay on Metaphysics (1940)—a thinly disguised attack on Ayer’s Language Truth and Logic (1936)—the only way to go beyond the metaphysics that logical positivism critiqued is to acknowledge the presuppositions that govern our attempts to come to know reality in different domains of inquiry, and to make explicit the role that they play in giving rise to the kind of questions different forms of knowledge seek to answer. It is only once the role that presuppositions play in shaping the questions we ask (and seek answers to) in different forms of inquiry is acknowledged, that one can truly go beyond the uncritical dogmatic metaphysics that positivism tried—but arguably failed—to do away with.

A. J. Ayer (1934) illustrated the pointless nature of metaphysical debates by contrasting a dispute between two art critics discussing the authenticity of a canvas attributed to Goya to a dispute between an idealist and a realist metaphysician. There are, he claimed, facts of the matter that can be invoked to settle whether the canvas is or is not a genuine Goya: the nature and direction of the strokes could be compared to those of certified Goya paintings; the canvas could be carbon dated to establish whether the paint matches the relevant period of time, and historical records mentioning such and such commissioning the painting could be referred to. The issue may be difficult to solve, but it is resolvable at least in principle, because there is an understanding of what kind of facts could be adduced in evidence either for or against the claim that the canvas is a genuine Goya. This is not the case with metaphysical disputes, which are not just difficult to settle in practice, but unresolvable in principle.

Imagine—Ayer suggests—that the two art critics debating the attribution of the canvass belonged to two different metaphysical schools: idealism and realism, and that they started debating whether the paint on the canvas is real or ideal. The realist argues that the paint really exists, the idealist that it is an idea in the mind. There is no fact of the matter that could prove the realist to be right and the idealist to be wrong, or vice versa. While there are facts of the matter that can be consulted to establish whether the canvas is a genuine Goya or a fake, there is no fact of the matter that could be cited to establish whether the paint on the canvas is real or ideal. The paint would look exactly the same, whether it is real or ideal, as Berkeley pointed out to assuage fears that a commitment to an immaterialist metaphysics might require abandoning the belief in the existence of mountains and rivers (Berkeley: Principles §34). The dispute between the two art critics is like one between two persons debating whether or not it is raining outside; such a dispute can be settled by consulting the facts, but there are no facts that can be consulted to establish whether the rain is real or ideal, precisely because the rain—just as mountains and rivers for Berkeley—looks exactly the same, whether it is real, as the materialist argues, or ideal, as the immaterialist claims.

The demand that knowledge claims should be verifiable, that there must be evidence that can be provided to substantiate one’s views, seems to be reasonable enough; rejecting it would lead to a form of dogmatism. But what the logical positivists also assumed is that the criterion of verification that belongs to the empirical sciences is a universal criterion of meaning, not a domain-specific criterion that merely determines what does and does not count as a genuine scientific hypothesis. They uncritically extended the criterion of verification that governs empirical enquiry to all claims (bar tautologies), rather than acknowledging it as a heuristic principle of scientific enquiry. As a result, they also failed to satisfactorily address the question concerning the logical status of the verification principle that states, ‘propositions which are not empirically verifiable are meaningless, unless they are tautologies.’

Since the verification principle cannot easily be accommodated within the Humean fork [Editor’s note: a principle by 18th-century Scottish philosopher David Hume]—according to which all meaningful propositions must be either empirical propositions about matters of fact or analytic propositions concerning relations of ideas—it looks suspiciously like a foundational principle for a positivist metaphysics. Positivism, it seems, does not dispense with metaphysics; it merely proposes a different kind of (naturalist) metaphysics. Therefore, failure to reflect on the logical status of the verificationist principle—to acknowledge it as a heuristic principle that governs scientific knowledge of reality—not only encourages a form of methodological monism—one which denies the autonomy of other forms of knowing—it also betrays a commitment to an uncritical realism, which assumes the scientific method reveals the ultimate nature of reality and, in so doing, places science in the position once occupied by metaphysics as the science of pure being.

This is precisely the point that R. G. Collingwood makes in his An Essay on Metaphysics. Collingwood argued that the principles that govern the verification of knowledge claims in any form of inquiry have a different logical status from the propositions which are made possible through a commitment to those principles. The inductive principle, according to which nature is uniform and the future resembles the past, for example, does not have the same logical status as the empirical generalization “ice melts when the temperature rises above 0°C.” The claim about ice is a proposition that can be verified or found to be true or false. The inductive principle is a presupposition that is neither true nor false, but makes possible knowledge claims like the one about ice. The role of philosophy is to uncover those principles that govern the verification of knowledge in different domains of inquiry—what Collingwood calls “absolute presuppositions”—not to propound true propositions that provide factual knowledge from the philosophical armchair, as the kind of metaphysics that Ayer critiqued did. In treating the principle of verification as a true second-order philosophical proposition, rather than as a presupposition of scientific inquiry, positivism ends up advancing the very kind of synthetic a priori claim whose possibility it wants to deny.

The reason why Ayer is a metaphysician malgré lui, as Collingwood would argue, is that he failed to acknowledge the distinction between propositions and presuppositions, between the criteria for the verification of knowledge and the verifiable claims made possible by the endorsement of such criteria. Perhaps, in exempting tautologies from the requirement that they should be empirically verifiable on pain of being meaningless, positivism implicitly acknowledges that the verificationist principle is a local presupposition that is constitutive of a particular form of (empirical) inquiry; one that differs from the criterion of verification that is constitutive of the exact sciences, not a universal criterion of meaning. But in so far as the significance of making an exception for tautologies is not fleshed out, the verificationist principle plays, even if only by default, the role of a foundational principle for a different kind of (naturalist) metaphysics. Had the significance of making an exception for tautologies been fully appreciated, the verificationist principle may have been recognized as a constitutive principle or presupposition of natural science, rather than a true second-order philosophical proposition spelling out which propositions express genuine knowledge claims and which do not.

Some might conclude, from the inability of logical positivism to escape the very metaphysics it sought to oust, that metaphysics is unavoidable; that the choice is not between either being or not being a metaphysician, but between either the kind of rationalist metaphysics that logical positivism attacked or the kind of naturalist metaphysics to which Ayer and logical positivism are surreptitiously committed to. But the conclusion that one must choose between a rationalist or a naturalist metaphysics is a little hasty. To leave behind the kind of metaphysics that the logical positivists sought to overthrow without committing to a naturalist metaphysics requires doing precisely what the logical positivists, in their haste to dispose of rationalist metaphysics, failed to do: to recognize that all knowledge, including scientific knowledge, rests on presuppositions and that the principle of verification is, in fact, a structuring principle of a certain form of knowledge, not a true proposition.

One should recognize, as Collingwood pointed out, that scientific knowledge is a form of knowing, with its own distinctive presuppositions; that it is a science, in the Latin sense of the term Scientia, meaning a body of knowledge with a specific method and subject matter, not the science or form of knowledge, in the sense in which the term “science” has come to be used—i.e., as slang for natural science, just as the term drink is being used as slang for alcoholic drink (Collingwood An Essay on Metaphysics: 4). Failure to acknowledge the role that presuppositions play in the pursuit of scientific knowledge grants natural science the epistemically privileged status of the science of pure being once enjoyed by rationalist metaphysics; it does not get rid of dogmatism, but merely replaces one kind of uncritical dogmatic realism with another.

 

References

Ayer, A.J. (1934), “Demonstration of the Impossibility of Metaphysics”, Mind 43 (171): 335-445.

Ayer, A. J. [1936] (1990), Language, Truth and Logic, London: Penguin Books.

Berkeley, G. [1710] (2020) “A Treatise Concerning the Principles of Human Knowledge”, in Principles of Human Knowledge and Three Dialogues, Graphyco Editions.

Collingwood, R.G. (1940), An Essay on Metaphysics, Oxford: Clarendon Press, revised edition, with an introduction by Rex Martin, Oxford: Oxford University Press, 1998.