A Priori, Bee, and C.

A recent post on Bluesky, the social media site that is simply buzzing with high-quality discourse, got me thinking about bees, and then, naturally, Immanuel Kant.

Sometime around the turn of the last century, the Swiss psychiatrist and myrmecologist Auguste Forel took his breakfast on a terrace and noticed that he had company. Honeybees arrived at the jam. This is not remarkable. What was remarkable is that they kept arriving at the same hour even after the jam stopped coming out, prospecting an empty table with what looked like conviction, and then giving up until the following morning, when they would try again at precisely the same time.

Forel wondered, idly, in the manner of a man with a well-furnished mind and a free morning, whether bees had a sense of time.

The question sat idle for thirty years. In 1929, Ingeborg Beling, working under Karl von Frisch in Munich, published the experiment that should have settled it and instead opened it wider. She trained bees to a sugar-water station available only during a fixed window, then took away everything the bees might have been using as a cue: constant light, constant temperature, constant humidity, a sealed room. The bees kept their appointment. Later workers went further and took the hives underground, into a salt mine, to rule out whatever residual geophysical rhythm a skeptic might still be pointing at. The bees kept their appointment there too.

Then, in 1955, Max Renner did the thing that makes the story worth telling. He trained bees in Paris to feed at a set hour. He put them on an airplane. He unloaded them in New York, into a closed room with controlled light and temperature, and waited.

They fed on Paris time.

They had been carried across five time zones and had brought their schedule with them, intact, indifferent, arriving punctually for a morning that was no longer occurring. The bees had jet lag. Which is to say: the bees had a temporal order that did not come from the world, and that the world could not immediately correct.

I want to argue that this is the best empirical illustration we have of a claim Kant made in 1781, and then I want to argue that calling it an illustration rather than a proof is not a hedge but the whole point, because the moment you upgrade it to a proof you have destroyed the thing you were trying to defend.

The Orthodox Position

In the Transcendental Aesthetic, Kant asks what space and time are, and rejects both available answers. They are not, as Newton had it, self-subsisting containers that would go on existing if you emptied them of everything. Nor are they, as Leibniz had it, relations abstracted from the arrangement of substances. They are forms of intuition: the structures our sensibility contributes to whatever it receives, prior to receiving anything.

His arguments for time run roughly thus (A30–31/B46–47). Time is not an empirical concept abstracted from experience, because to notice that two events are simultaneous or successive you must already have the representation of time; you cannot derive it from what presupposes it. Time is necessary, in that you can think away every event but you cannot think away time itself. Time is not a general concept with instances under it but a single unlimited whole with parts marked off within it, which is what an intuition is, not what a concept is. And time is given as infinite in a way no concept could deliver.

The conclusion is the pair of theses that make Kant Kant: time is empirically real and transcendentally ideal. Every object you will ever encounter is in time, without exception, and this is not an illusion or an approximation. But time is not a property of things as they are in themselves. It is the form of inner sense, the shape under which the mind is affected by itself, and because every outer appearance must also be taken up into inner sense in order to be mine, time is the form of all appearances whatsoever, not merely the inner ones. Space governs half the manifold. Time governs all of it.

Now set Renner's bee beside that. A creature is placed in a novel world. The world offers it a set of cues: light, temperature, the whole ambient rhythm of Long Island in the 1950s. The creature ignores them and continues to order its experience according to a structure it brought with it. The structure was not derived from the environment; it survived the total replacement of the environment; and its persistence is exactly what makes the mismatch, the jet lag, possible. You cannot be jet-lagged unless you carry something that fails to update.

That is a very pretty picture of a form of intuition. Content changed completely. Form did not. And the phenomenology of the discrepancy—that queasy, ill-fitted, unwell feeling that any of us who has flown to Frankfurt-am-Main knows—is what it is like when the form and the world come apart.

Why This Is Not a Proof, and Why That Matters

Kant would not have taken the offer.

He draws the distinction himself, at the opening of the Deduction (A84–85/B116–17), between the quaestio facti and the quaestio juris: the question of how a representation was in fact acquired, and the question of its right, its entitlement, its validity. He introduces the distinction specifically to dispose of Locke's "physiology of the human understanding," which he regards as a perfectly respectable inquiry that answers a different question than the one he is asking. Renner's experiment is a quaestio facti result of exceptional elegance. It establishes an origin. It cannot establish a title.

There is a sharper way to put the problem. Renner's experiment is conducted in time, by experimenters, using clocks, comparing Paris time to New York time, timing the interval between feedings with instruments whose readings are themselves successive states. Every element of the apparatus presupposes the form under investigation. Nothing empirical could underwrite the claim that time is the condition of the possibility of experience, because the empirical is what that condition makes possible. This is not a limitation of Renner's design. It is a structural feature of transcendental claims, and it is why they are called transcendental.

And there is the falsifiability trap, which is the one that should actually worry you. Suppose Renner's bees had reset immediately on landing. Suppose the whole Zeitgedächtnis literature had come out the other way, and every apparent time sense in animals had turned out to be a sensitivity to some subtle environmental cue we had not yet controlled for. Would that have been evidence against the Transcendental Aesthetic? Obviously not. Kant's claim is about the conditions of experience for a discursive intellect, not about the physiology of Apis mellifera. A thesis that cannot be disconfirmed by an experiment is not confirmed by it either. If you cash the bees as validation you have quietly agreed to a bet you were never willing to lose.

There is also a second problem, less often noticed, which is that inner sense is not a time-detecting organ. It is the form under which the mind is affected by itself, and it comes bundled, in Kant's architecture, with apperception, with the unity that makes a manifold mine, that permits the "I think" to accompany my representations. Whatever the bee has, it has without that. An endogenous circadian oscillator is a mechanism that keeps time. Inner sense is a form in which a subject has time. The hostile reading writes itself: Renner did not discover the a priori form of inner sense in the bee. He discovered an organ for the empirical detection of a determinate magnitude, which makes time an object of sense, precisely what Kant spent the Aesthetic denying it is. You have found a clock. Kant's entire point is that time is not a clock.

Good. Now that the strong thesis is dead, we can do something more interesting with the corpse.

Uexküll: The Kantianism of the Tick

Jakob von Uexküll knew exactly what he was doing, and said so. The theory of the Umwelt—the surrounding-world, the perceptual and operational bubble specific to each organism—was advanced by him as a biological Kantianism, and he was explicit that space and time belong to the subject rather than to the world. Each animal inhabits its own space, structured by its own capacities, and, crucially, its own time.

His illustration is the Moment: the smallest indivisible unit of experienced time, the temporal grain below which nothing can be distinguished. He gives figures—roughly an eighteenth of a second for humans—and reports experiments in which animals with a coarser grain fail to resolve motion that we resolve easily, and animals with a finer grain resolve what to us is a blur. A snail whose moment runs at three or four per second will step onto a rod oscillating fast enough that we see it vibrating, because for the snail it is standing still. The rod is not vibrating in the snail's world. The world's temporal texture is a function of the subject.

This is Kant with the transcendental subject pluralized into a phylum. And it does something important to Renner's bee. Once you have Uexküll, the bee flown to New York is not merely a mechanism that failed to update. It is a subject carrying its Umwelt with it, encountering a world whose temporal structure no longer meshes with its own, and the jet lag becomes something more like an experience: the friction of two incompatible times, one constitutive and one merely encountered.

Heidegger: The Bee Does Not Have the Honey as Honey

Which is precisely where Heidegger reaches for the same insect data and concludes the opposite.

In the 1929–30 lecture course published as The Fundamental Concepts of Metaphysics, Heidegger works through the biology of his day—Uexküll very much included—toward the triad that the course is famous for: the stone is worldless, the animal is world-poor, man is world-forming. The animal's condition he calls Benommenheit, captivation, a taking-up or be-numbment in which the creature is absorbed by its drives without any distance from them. His example is a bee, from the experimental literature, whose abdomen has been severed and which goes on drinking honey it can no longer retain. The bee is related to the honey. But it does not have the honey as honey. It lacks the "as," and lacking the "as," it has no world, only an encircling ring of disinhibitors.

Follow that out and the consequence for time is severe. In the Kantbuch of the same year, Heidegger had made time the ground of everything: the transcendental imagination as the common root of sensibility and understanding, time as pure self-affection, the innermost structure of finite subjectivity. Having elevated time that far, he cannot then hand it to the bee. What the bee has is sequence, cycle, the periodicity of a mechanism. What it does not have is temporality in the sense of an ecstatic horizon within which anything could be encountered as something.

So: two twentieth-century thinkers, working from overlapping experimental literature, and one concludes that the animal's world is constituted by its own forms of intuition while the other concludes that the animal has no world at all. That disagreement is the actual philosophical content of the jet-lagged bee, and it is more interesting than any vindication of Kant could have been. The experiment does not adjudicate between them. It sharpens what they are disagreeing about.

Friedman, Reichenbach, and the Cost of Naturalizing

There is a way to save something here, and it has a distinguished pedigree.

Reichenbach, in 1920, pointed out that Kant's "a priori" had been doing two jobs at once. It meant necessary and unrevisable for all time, and it meant constitutive of the object of knowledge, the framework in the absence of which the objects of a science are not even well-defined. Relativity had made the first sense untenable for Euclidean geometry. Reichenbach's proposal was to drop the first and keep the second. Michael Friedman has spent a career developing what follows: framework principles that are genuinely constitutive within a paradigm, genuinely not empirical hypotheses under test, and yet historically mutable, replaced when the framework is replaced. A relativized a priori.

The biological version arrived earlier and cruder, from Konrad Lorenz in 1941: the Kantian a priori is the phylogenetically acquired. What is hardwired in advance of any individual's experience is a priori for that individual, having been shaped by the species' encounters rather than the organism's. This is the ancestral document of evolutionary epistemology, and it has real defenders. It is also, as Uexküll and Helmholtz before it, part of a much older fight. The "physiological Kantianism" of Helmholtz and F. A. Lange was contested by the Marburg neo-Kantians precisely on this ground; Hermann Cohen's objection was that turning the forms into facts about our nervous organization concedes the entire critical enterprise while keeping its vocabulary. That argument was fully joined by 1871, eighty-four years before anyone put a bee on an airplane.

And the cost is exactly what Cohen said it was. Lorenz buys the innateness and loses the necessity. Species-typical is not apodictic. A lineage with a different evolutionary history would have different forms; forms that could have been otherwise are contingent; and contingent forms are not Kant's. You may keep the word.

The Question

This turns to my question and the thesis, such as it is, of this short essay.

Is a form that turns out to have a phylogeny still a form in Kant's sense?

One answer: yes, in the sense that survives, and that is the only sense worth having. Constitutivity is load-bearing. The bee's temporality is not an item within its experience but the condition of its having any; that this condition was laid down by natural selection rather than deduced from the possibility of experience changes its genealogy, not its office. Reichenbach and Friedman have already shown that you can lose apodicticity and keep something recognizably Kantian doing recognizably Kantian work.

The other answer: no, and the naturalizers are running a shell game. What made the Aesthetic worth arguing about was not that we happen to organize experience temporally. It was that nothing could count as experience for any discursive intellect otherwise, a claim about the space of possibility, not about the contents of the actual world. Naturalize it and the universal quantifier collapses to a survey. You are no longer telling me what experience must be. You are telling me what bees and beetles and primates in fact do, which is a fine thing to be told, and is called comparative physiology.

I lean, on most days, toward thinking that the second answer is right about what is lost and the first is right about what remains, and that the honest position is to admit that these are two different projects wearing one word. Kant asked quid juris. Renner asked quid facti. The bee, sealed in her room in New York, arriving on time for a Paris morning that ended five hours ago, is not evidence for anything. She is a picture of what it would look like if a form of intuition could be seen from outside. Which it cannot, which is why we needed the bee. We showed that there is a heteropsychological (to borrow Carnap's term) sense of intuition even in bees, that subjecthood and biology remain intertwined, consciousness is embodied, and that there can be no unity of the sensory manifold without the a priori forms of intuition.