About Objectivism, part II: Epistemology

By Massimo Pigliucci

[Part I of this series was on Objectivist metaphysics]

For a philosophy named Objectivism, epistemology is foundational, since the possibility of objective knowledge is an inherently epistemological question. In this regard, Rand and her followers are surprisingly ambiguous. For instance, conscious of the obvious fact that human beings continually make errors of judgment, observation and reasoning, they agree that one cannot actually be certain of any proposition one utters. Oh? From when, then, comes any claim of objective knowledge? Well, as Leonard Peikoff put it in his Objectivism: The Philosophy of Ayn Rand, any proposition is “certain” if the available evidence supports it, i.e., it is certain within a particular context.

But this is pretty close to wanting your cake and eating it too. Any sensible epistemologist would simply say that said proposition is made more or less probable by the available evidence, and that this judgment may change if the evidence (the “context”) changes. No claim of certainty is warranted. The problem, of course, is that rephrased this way the Objectivist idea becomes unsurprising while at the same time undercutting any broad claim to objective knowledge.

Rand spends a lot of time rejecting religious and mystical claims, on grounds very similar to that adopted by the (failed) previous school of logical positivism in mainstream philosophy. While I certainly agree with Rand that faith is not a short-cut to, but rather a short-circuit of knowledge, the idea that one can reject claims about the supernatural on grounds that “nothing has been said” is, again, reminiscent of the logical positivists’ verifiability criterion — which famously and ironically undermined logical positivism itself. (It works like this: “nothing that cannot be verified makes sense.” “All right, then, how do we verify the latter sentence?”) In the case of Objectivism, it seems fair to raise the “nothing has been said” objection to Objectivist utterances such as “consciousness is identification,” which is a cardinal point of Objectivist epistemology.

Rand and her colleagues also rejected the long-lasting distinction between analytic and synthetic statements in philosophy. An analytic statement is true by virtue of its meaning, like “A bachelor is an unmarried man,” or the less obvious “The sum of the angles of a triangle is 180 degrees” (the latter is analytic because it turns out that having a total of 180 degrees in angles is part and parcel of being a triangle). The truth of a synthetic statement, on the other hand, generally depends on external, empirical evidence. For instance, “All bachelors are unhappy (presumably because they are unmarried),” or All mammals have a four-chambered heart.”

While things got complicated once Kant got hold of the analytic/synthetic distinction, such distinction is similar to that previously made by Hume between a priori and a posteriori propositions, which according to him are the only two types of propositions that make any sense. Here is how he expressed what became known as Hume’s fork, in his An Enquiry Concerning Human Understanding:

“If we take in our hand any volume; of divinity or school metaphysics, for instance; let us ask, Does it contain any abstract reasoning concerning quantity or number? No. Does it contain any experimental reasoning concerning matter of fact and existence? No. Commit it then to the flames: for it can contain nothing but sophistry and illusion.”

I must admit a lot of sympathy for this sentiment. But back to the analytic/synthetic distinction: as I said, Kant improved on it (no need to go into the details now, but you can get the run down here). Then, in a famous paper published in 1951, W.V. Quine claimed to have demonstrated that the notion of analytic propositions is untenable, and that therefore the distinction between the two types has to be rejected. This paper is much more widely cited than it is understood (again, a summary can be found here); still, while Quine’s basic argument is highly technical, it hinges on the idea that the concept of analytic statements is in turn based on the concept of synonymy (because, for instance, “bachelor” and “unmarried man” have to mean the same thing if the phrase “a bachelor is an unmarried man” is to be meaningful). Quine claims that we don’t have a sufficiently clear explanation of synonymy (according to his rather stringent and idiosyncratic criteria of “clear” and “sufficient”), so the whole house of cards crumbles. I have never been convinced by this argument, and like many philosophers I know, I acknowledge Quine and swiftly proceed to ignore him: to me “bachelor” and “unmarried man” are unproblematically synonymous, and the truthfulness of that sentence is known a priori because it depends only on meaning and not on facts about the world. Period.

Now, why do we care about any of the above in the context of Objectivism? Because it too rejects the analytic/synthetic distinction, since it wants to deny the possibility of a priori knowledge. I’m not sure why some people are so set against a priori knowledge, but logic and mathematics are darn good examples of it, and I for one wouldn’t want any epistemology that didn’t take logic and mathematics seriously and as distinct from the natural sciences.

Another positivistic sounding notion in Rand’s epistemology is the rejection of emotions as a type of cognition. Indeed, Peikoff went so far as to state that “emotionalism” is synonymous with irrationality (I wonder if he checked with Quine before claiming that something is synonymous with something else). This sort of attitude has been common in philosophy since Plato (read: nothing new in Objectivism), with the notable exception of Hume, who was one of the first philosophers to seriously play up emotions as both a source of certain kinds of knowledge (particularly moral judgment) and as actually guiding reason rather than being controlled by it. Modern philosophy of mind and cognitive science are producing a much more nuanced understanding of the necessary integration of emotional and cognitive functions in the brain, without which, human knowledge and in fact the human condition itself, would not be possible.

Perhaps the most curious part of Rand’s epistemology was her contention that sensory perception is valid in an axiomatic sense. Since perception is a physiological function, according to her it comes without error, which led Rand to bizarre statements such as that optical illusions are conceptual errors, not errors of sight — as if “sight” were somehow clearly distinct from the brain’s conceptualization of what we see. The reason perception had to be perfect is that Objectivism is a kind of empiricism, relying on the notion that all our knowledge is ultimately based on the senses, just like the classical British empiricists (Hume, Locke and Berkeley) had maintained (though Objectivists actually have a problem with Locke, and presumably Hume, since they acknowledge the imperfection of the senses).

But by the time Rand was writing, philosophers in general, and epistemologists in particular, had gone way past a simplistic distinction between empiricism and rationalism, acknowledging the existence of both empirical and non-empirical knowledge, as well as getting down to analyzing the many ways in which both sensorial perception and reasoning can go wrong (which means giving up the dream of purely objective knowledge). That work continues today, and bleeds into similar work in the cognitive sciences (see, for instance, this fascinating article on the philosophy and physiology of delusions), from what I can see with precisely zero contribution from Objectivism.

Next time: Objectivist ethics (oh boy).