Tackling the Trilemma

What Is to Be Done?

Disclaimer: In tackling the Münchhausen trilemma in an essay, I’m tackling a challenge for which each component has had, not just a book, but an entire body of books, written about it. I do not claim to have read every work every written on every aspect of the trilemma and it’s components, merely enough that I am attempting to do the hard work of assembling my own thoughts systematically. But it is a task admittedly akin to attempting to assemble a puzzle whose every piece has a fractal boundary. If under a microscope some of the puzzle pieces I have fit together actually have gaps between them, I hope that the overall picture I have assembled is nevertheless a systematically interesting one.

Faced with the horror of the Münchhausen trilemma, what is to be done? Here let us sketch out four of the possible approaches, which I shall name the Pragmatic, the Straussian, the Nietzschean, and the Randian. (There are other approaches that merit further investigation, but these four will suffice for my essay.)

Within the Pragmatic approach I include the views of philosophers such as Charles Sanders Pierce, John Dewey, W.V. Quine, Karl Popper, and Hans Albert. The latter developed the modern version of the trilemma we are wrestling with; and of course, the Pragmatists as a whole can trace their roots to the Pyrrhonist skeptics who developed the ancient version of the trilemma.

Not surprisingly the Pragmatist’s own response to the trilemma is their theory of Fallibism, which holds that we cannot know anything to be true with certainty, merely know that some things are false. Fallibists claim we do not need to abandon our knowledge, merely accept that knowledge is fallible. The late David Stove, however, argues in Scientific Irrationalism: Origins of a Postmodern Cult that Pragmatism ultimately collapses into irrationalism which leads to the very sort of post-truth society we seek to avoid. I am inclined to agree with Mr. Stove.

The Straussian approach is rooted in the work of the eponymous Leo Strauss. Strauss, in his various works, argues that Western civilization took a wrong turn at the Enlightenment, when it foolishly tried to found its culture on explicitly rational values. The Classical philosophers knew that this was impossible; philosophy and politics - man as rational and man as social - are inevitably in conflict. For this reason, Strauss claims, the philosophers wrote esoterically, hiding the corrosive skepticism of their findings behind aphorism, parable, and double talk. Strauss, like Plato, embraces the noble lie. Since society depends on truth, we must conceal the fact that the trilemma is unsolvable. The good philosopher teaches rational skepticism to the select, while mouthing the thoughts and prayers of the masses in public.

The Nietzschean approach also accepts the trilemma as unsolvable and argues that coherentism is the historical answer to the trilemma. In On Truth and Lies in a Nonmoral Sense, Friedrich Nietzsche writes that truth is actually “a sum of human relations which have been enhanced, transposed, and embellished poetically and rhetorically, and which after long use seem firm, canonical, and obligatory to a people: truths are illusions about which one has forgotten that this is what they are; metaphors which are worn out and without sensuous power; coins which have lost their pictures and now matter only as metal, no longer as coins.”

So far, Nietzsche and Strauss seem to be in alignment. Both accept the trilemma as a real problem (as compared to the pragmatics, who see it as merely a theoretical one); but where Strauss wants to keep this esoteric, Nietzsche strives to make it exoteric. When Nietzsche writes “God is dead,” he does not mean to imply that God was ever alive, but rather that the coherent theory of truth based on the Christian God has become overturned and is no longer believable or useful. What we must do, says Nietzsche, is create a new narrative, a new culture. It does not matter if this narrative is True; there’s no such thing. Indeed, Nietzsche argues that Man is little interested in truth per se; he has no “will to truth.” He has instead a will to power. And the new narrative will be created by a man of exceptional will, who creates new myths and new truths by deed.

Of course, once a Nietzschean superman has succeeded in the transvaluation of values, it will require Straussian-like esotericism to preserve the new values, until those too decay into worn out metaphor and coins which have lost their pictures, and are replaced by new, strong ideas.

One could argue that present-day Trumpism is effectively Straussian. Trump re-affirms the tenets of America’s civic religion, not with new arguments, but with calls to tradition and history. Similarly, present-day Progressivism is effectively Nietzschean. It seeks to tear down the culture of 19th century liberalism and replace it with a new culture based on narratives of race, gender, and social justice. In accepting the trilemma, both are functionally postmodern and post-truth.

Both Nietzsche and Strauss, looking at America’s competing narratives today, would likely conclude there will be war - cultural war certainly, civil war possibly. Strauss would see the conflict as a tragic result of corrosive skepticism applied to a culture which cannot stand up to it; Nietzsche would see this as a healthy overtoppling of worn-out worldviews, so that they can be replaced by new, stronger, healthier culture. But if we hope to resolve the culture war through reason rather than power, neither offers much hope. Trumpism and Progressivism seem far from finding peace with each other.

The Randian approach, named for Ayn Rand, seeks to solve the trilemma. Rand’s politics of right-wing laissez-faire minarchism are controversial, especially among left-leaning philosophers; but if philosophers are ignored based on their politics, then everyone from Aristotle to Heidegger must be discarded. We will not here be discussing Rand as a political theorist in any case, but rather Rand as an epistemologist. Rand actually named her philosophy Objectivism because she considered her foremost contribution to philosophy to be in the field of epistemology, and her most important work to be An Introduction to Objectivist Epistemology.

If you don’t know much about what Ayn Rand actually said, you are hardly to be blamed: Almost nobody does. Ayn Rand rejected on philosophical grounds both the analytic school of the English-speaking world and the continental school of Europe. Her writing style, intended to be approachable to the layman, ending up lacking both the measured and cautious tone of analytic philosophy and the poetic and aphoristic style of the continentals. Her scathing diatribes against both schools won her little love and few academic followers. As such, despite a voluminous and brilliant corpus of work, she remains an outsider. Meanwhile “pop culture” Objectivism is largely a caricature of her views and is better left outside the halls of thought.

The key to a proper understanding of Ayn Rand is to realize that, though she was neither an analytic or a continental, she did belong to a philosophical school: Aristotelianism. Rand considered Aristotle and his successor St. Thomas Aquinas the two greatest philosophers who had ever lived, and the more you understand Aristotle the more you understand Rand.

Aristotle was a foundationalist. As the inventor of logic, he was the foundationalist. At this point, we should clarify that foundationalism is rightly split into two sub-schools. Rational foundationalism asserts that all propositions fundamentally depend on certain unprovable axioms. Empirical foundationalism asserts that propositions are fundamentally justified non-propositionally by perception. Aristotle’s foundationalism incorporates elements of both rationalism and empiricism. His Prior Analytics explains the rules by which propositions can be justified through deduction. His Posterior Analytics handles perception, induction, and scientific knowledge. In the Posterior Analytics, Aristotle ultimately reveals himself to be an empiricist, asserting that our first principles are ultimately justified by our perceptions.

For empirical foundationalism to be safe against the trilemma, the following propositions must be defended: (1) High-order concepts can be justified by low-order concepts; (2) low-order concepts can be justified by perception; and (3) perceptions can be justified by their direct knowledge of reality. Does Aristotle succeed in defending these propositions?

We will start with proposition (3). Aristotle argued that our perceptions gave us direct knowledge of the world, a position today known as direct realism. Direct realism, not coincidentally, lends itself to the classic correspondence theory of truth, where a statement is true if it accurately reflects the state of affairs of the world.

Direct realism was defended throughout the Middle Ages. However, the version of direct realism that Aristotle defended was based on his hylomorphic metaphysics and physics; and as those came under sustained criticism, so too did his direct realism.

In the Enlightenment, direct realism was supplanted by the indirect realism of René Descartes and John Locke. According to indirect realism, our perception gives us only an indirect “representation” of the world; we perceive sense-data, not reality. Descartes remained a foundationalist, but his foundationalism became a strictly rationalist foundationalism based exclusively on a priori knowledge. In severing the direct connection between perception and reality, Descartes paved the way for Immanuel Kant, who distinguished the perceptual world from the real world and claimed we could only know the former and never the latter. This, in turn, lead to rise of phenomenalism, which argues that the external world and its things cannot even be said to exist except as bundles of sense-data. If these modern philosophers are right, then Aristotle is wrong about proposition (3).

Let us put aside proposition (3) and turn to proposition (2). How are low-order concepts justified by perception? In addition to holding to the direct realism of perception, Aristotle also held to the immanent realism of universals. What is a universal? To offer up a definition is to choose a school of thought, so it is better to proceed by example. Universals are things like beauty, loudness, redness, or triangularity. The problem of universals is how universals correspond to things existing in the world. There are four solutions to the problem, with transcendental realism and nominalism as extreme positions and immanent realism and conceptualism as moderate positions.

Plato was the originator of transcendental realism. He famously held that universals exist transcendentally in a “realm of forms”. Beautiful things partake of the universal of beauty; red things partake of the universal of redness.

Nominalism was introduced into Western philosophy by William of Ockham. Ockham held that universals do not exist; they are merely labels that human beings assign to actually existing things in order to form then into collections or series.

Conceptualism has been attributed in its earliest form to Zeno and the Stoics. Conceptualism admits the existence of universals, but only as concepts in the mind. Conceptualists claim that we do not know whether or not the mental objects have any foundation outside of our minds in nature.

Finally, immanent realism was developed by Aristotle. It proclaims the existence of universals immanently in things. Triangularity exists, but not in a transcendental way; instead it exists in triangles. Triangular things have a metaphysical essence which includes the attribute of triangularity.

Now, if immanent realism is correct, then point (2) is defended. When we perceive things, we perceive their essences and thereby gain knowledge of universals they instantiate. These universals become our low-order concepts.

Unfortunately, as was the case with Aristotle’s direct realism, Aristotle’s immanent realism was founded on his hylomorphic metaphysics and physics. Once Aristotle’s metaphysics and physics were discarded during the Enlightenment, his theory of immanent realism was discarded with it.

It is not coincidental that Aristotle’s theory of perception and his theory of universals were both discard at the same time. The theories of perception and the theories of universals described above actually correspond to each other in near perfect parallel. (I have never seen this point made in the literature, so it may be original to me.) Consider:

• Direct realism holds that we perceive the real world; immanent realism holds that the universals we abstract from our perceptions actually exist in the things we perceive.

• Indirect realism holds that we perceive sense impressions of the real world; conceptualism holds that the universals we abstract from our perceptions are true of our sense impressions.

• Phenomenalism holds that we perceive sense impressions but the real world cannot be said to exist; nominalism holds that the universals we label things with cannot be said to exist.

So Aristotle’s defenses of point (2) and point (3) collapsed simultaneously and for the same reason.

Now let us turn to proposition (1), the requirement that high-order concepts be justifiable by low-order concepts. Proposition (1) is the subject of Aristotle’s Prior Analytics and Metaphysics. There he lays out what would later be called the laws of thought: the law of identity, the law of non-contradiction, and the law of the excluded middle. A is A; A is not not-A; X is either A or not-A. On the basis of these laws, Aristotle was able to develop a system of deduction that allows us to justify high-order concepts systematically from low-order concepts.

Now, the laws of thought themselves can be directly assaulted, and that is a line of attack to which we will return later. But for now, let us grant the laws of thought as valid and pursue another attack against poor Aristotle. Let us instead show that if even proposition (1) holds, it doesn’t matter!

When direct realism and immanent realism were replaced by indirect realism/phenomenalism and conceptualism/nominalism, the cleaving of perception from reality created the analytic-synthetic distinction. Here, an analytic truth is one that is true by definition, while a synthetic truth is one that is true contingently. Put simply, an analytic truth is necessarily true (“all dogs are canines”), while a synthetic truth must be evaluated for truth through evidence (“all dogs are fuzzy”). But if we have no way of perceiving the real world, then we have no way of every identifying synthetic truths as truthful. As Rand explains, if the analytic-synthetic distinction holds, we are trapped: “if your statement is proved, it says nothing about that which exists; if it is about existents, it cannot be proved.”

Does the analytic-synthetic distinction of truth hold? Arguably it doesn’t just hold, it actually swallows up analytic truth. In 1951, W.V. Quine published the essay "Two Dogmas of Empiricism", where he argued that are no analytic truths at all; all truths are synthetic, all truths involve an empirical aspect. But if that is the case, and if direct realism is not true, then we cannot prove anything, because we lack access to facts about reality to offer as evidence. Quine’s arguments are hotly debated (and philosophers don’t even agree about what Quine was saying) but for our purposes let us conclude that if Quine is correct, and if direct realism is incorrect, then absolutely no higher-order concepts, even purely logical ones, can be ever proved in the manner that a foundationalist would want them to be. That demolishes proposition (1).

This lengthy digression now leads us back to Ayn Rand. Ayn Rand’s philosophical goal can best be understood as defending our three foundationalist propositions: (1) High-order concepts are justified by low-order concepts; (2) low-order concepts are justified by perception; and (3) perceptions are justified by their direct knowledge of reality. Her challenge is to do so without resorting to Aristotle’s hylomorphic metaphysics. (I’m aware that there is an active Thomist and Neo-Aristotelian school that defends Aristotle’s metaphysics, but for reasons of space I will not be addressing it here.)

Because she rejects Aristotle’s metaphysics, she must reject essences and with them Aristotle’s immanent realism. She does not, however, reject direct realism. Treating direct realism as an assumption, Rand marries it with a type of conceptualism to produce a new solution to the problem of universals. In the Introduction to Objectivist Epistemology, she explains her reasoning as follows:

Aristotle held that [universals] refer to metaphysical essences, which exist in concretes as a special element… and he held that the process of concept-formation depends on a kind of direct intuition by which man’s mind grasps these essences and forms concepts accordingly.

Aristotle regarded “essence” as metaphysical; Objectivism regards it as epistemological… produced by man’s consciousness in accordance with the facts of reality.

Rand says, in effect, that if direct realism is true, then conceptualists are wrong to assert that “we do not know whether or not our mental objects have any foundation outside of our minds in nature.” Universals can exist as epistemological concepts, yet be objectively in accordance with the facts of reality. If that is the case, then the analytic-synthetic distinction is overcome and we are on our way to defending foundationalism.

Is it the case? Does Rand successfully defend her theory? Because of its length and detail, an elaboration of Rand’s full epistemology is outside the scope of this essay. Here let us sketch out merely the briefest overview of it. According to Rand, universals are epistemological concepts that classify together concretes or abstracts whose measurements fall within the same category of measurements within commensurable characteristics. If what’s being classed together are concretes, it’s a low-order concept. If what’s being classified together are abstracts, it’s a high-order concept.

Let’s break that down:

• A concrete is an existents such as a car or a blueberry.

• An abstract is a concept about a concept.

• A measurement is a quantitative assessment of some value. Rand agrees with mainstream physics in arguing that everything in the real world can be measured mathematically. For instance, color is often deemed subjective, but Rand notes “there are differences in degree along three measurable axes: hue, saturation, and brightness... Modern computers use numbers from 0 to 255 to specify the setting of each of these three parameters.”

• A characteristic is an attribute of an existent that can be measured, such as height, weight, and color. Rand includes within characteristics what she calls relative characteristics, such as proximity, which are measured in a context.

• A set of commensurable characteristics are characteristics across two or more existents that have the same units of measurement. For instance, “redness” and “blueness” are commensurable because both are measured along the same three axes. “Redness” and “height” are not commensurable.

Low-order concepts are thus grounded in direct measurement through the evidence of the senses. High-order concepts are developed by a process of “abstraction from abstractions.” Here, the concept-forming processing is applied to lower-level concepts, classifying concepts together in the same manner as concretes were classified together. An example of a high-order concept would be “dimensionality,” which would class together the low-order concepts of height, weight, and depth because their measurements all fall within the commensurable characteristic of “extension in space.”

These points are all well-defended at length in Rand’s own book, as well as in Harry Binswanger’s book How We Know. For now, let us grant, for the sake of further investigation, that this epistemology is correct.

Even so, we have not yet defeated the trilemma. First, if Rand’s theory of universals is correct, but direct realism is wrong, then she has merely established a particularly well-grounded form of conceptualism. But that, we have already seen, is not enough.

Second, even if Rand’s theory of low-order concepts is correct, grounded as it is on the evidence of the senses, her theory of high-order concepts may be wrong because it relies on the laws of thought to make its case.

The skeptic can thus still assail the Randian solution on two accounts: (1) an attack on direct realism and (2) an attack on the laws of thought themselves. These attacks, and possible defenses, will be the subject of our next essay.