Seven Things Radical Empiricism Cannot Grant

Radical empiricism updates beliefs from experience. It does not update the experiencer. That single asymmetry holds the whole framework together, and it's the reason a serious defender of the position has to refuse the seven claims below. Each of them reaches beneath the belief layer to the aperture itself. Grant any one, and the framework cannot hold.

What follows is the seven claims, briefly, with the evidence that makes each one hard to escape.

1. Cognition is limited by formation

The knower is not a standardized receiver with correctable biases. The knower is a shape that was built before any reasoning begins, and the shape determines what can register as evidence in the first place.

The clean empirical case is Chase and Simon's 1973 chess experiment. Masters shown a real game position for five seconds reconstruct it almost perfectly; novices get a few pieces. Same input, same five seconds, radically different perception. Then the researchers switched to random piece arrangements, and the masters performed no better than the novices. The mastery wasn't memory. It was a trained aperture — the masters were seeing patterns, and when the pattern wasn't there, the advantage vanished.

The stronger version is the Mandarin-tones case. Infants raised hearing tonal languages retain the ability to distinguish tones; infants raised without them lose it by around twelve months, as the auditory cortex is pruned based on what the environment marked as meaningful. Adults who study Mandarin for decades often never recover native tonal perception. The window closed; the organ was shaped.

Formation is not noise on a signal. It's the shape of the receiver, built physically, before any belief-revision can run.

2. Literature gives privileged access by changing the knower

Literature is not a source of propositions to update beliefs from. Literature is a technology for reshaping the aperture. A reader emerges from Anna Karenina able to perceive distinctions in human motivation she could not perceive before — not because she now holds more beliefs but because she is a different perceiver.

This is testable. Kidd and Castano's 2013 study in Science found that reading literary fiction — specifically literary, not genre — produces measurable, if short-term, improvements in theory-of-mind tasks. Nussbaum's Poetic Justice argues the longer case: moral perception is trained by novels in ways direct instruction cannot achieve. George Eliot said it plainly in the preface to Middlemarch: the function of art is the extension of our sympathies. What she meant wasn't add a belief about sympathy. She meant become capable of sympathies you weren't previously capable of.

The empiricist move is to flatten this into literature is just more experience. The equivocation is on the word experience. The experience of reading is one thing; what the reading builds in the reader over time is another. If you use one word for both, the framework survives. If you distinguish them, the second one is the formation thesis and the game is over.

3. Some knowledge cannot be captured in language

Polanyi's line in The Tacit Dimension is decisive: we know more than we can tell. A cyclist can ride. He cannot produce the differential equations that describe how he balances. If he tried to ride by consulting them, he would fall.

Every craft works this way. A master carpenter reads grain. A radiologist sees tumors in scans that look like noise to untrained eyes. A nurse knows the patient is crashing before any monitor flags it. A jazz musician hears the next note. In every case, the knowledge is real — it predicts, it discriminates, it saves lives — and it cannot be fully rendered in propositions. Attempts to produce expert systems that encode expert knowledge have repeatedly failed at exactly this boundary. The knowledge lives in the body of the knower, not in a statable rule.

A framework that only counts knowledge as knowledge when it can be put into sentences is not rigorous. It's amputated.

4. There is knowledge beyond prediction, and prediction is itself a frame

The hidden assumption in empiricism is that knowledge is prediction — that what a belief does, ultimately, is pay out in predictive power. Anything that doesn't predict isn't really knowledge.

This is a choice, not a discovery. It came from Bacon and got codified by Mill and the logical positivists, and it treats Aristotle's much older distinction — theōria (contemplative knowing), phronēsis (practical wisdom), technē (craft), epistēmē (demonstrative knowledge) — as if only the last one counts. Theōria doesn't predict. It beholds. Aristotle thought it was the highest form of human activity because it was the only one done for its own sake.

Beholding isn't a failed prediction. A person contemplating a sunset, a theorem, a beloved face, the structure of an organism, is not running a deficient forecast. They are in a different relation to reality. If the only question you can ask about a piece of knowledge is what does it predict, you have ruled out most of what humans have meant by knowing for most of human history — and you've done so by fiat, not by argument.

Prediction is a frame. Mistaking the frame for the floor is the move that hides from itself.

5. What you desire determines what you attend to

This is the asymmetry that breaks the system if you look at it. The empiricist subjects beliefs to rigorous demand for evidence. The empiricist pockets desires as given. "Desired outcomes" is the terminal point of utility calculation. "What I value" is private, inviolable, not to be questioned. The whole belief-revision apparatus runs on top of an appetite-structure that the framework refuses to scrutinize.

But the empirical case is overwhelming: what you want determines what you notice. A hungry person walking past a street of shops sees restaurants. A shopper sees clothes. A thief sees unlocked doors. Same street, same photons, different perceptions — because attention is desire-directed before it is belief-directed. Simons and Chabris's invisible gorilla experiment demonstrates this at the bottom level: watchers counting basketball passes famously miss a person in a gorilla suit walking through the scene. Not because they couldn't see it. Because the task — i.e., the imposed desire — had shaped what their visual system admitted.

If desire controls attention and attention controls what can register, then the most important epistemic question is not what do I believe but what do I want, and who made me want it. The empiricist will not ask this. Asking it would crack the floor.

6. Virtue is a prerequisite for certain kinds of knowledge

Aristotle, Nicomachean Ethics Book I: the young are not fit to study ethics. Not because they're stupid. Because they lack the formed habits that make ethical phenomena visible. The coward cannot see courage clearly — it looks like foolishness. The proud cannot see humility clearly — it looks like weakness. The habitually cruel cannot see tenderness clearly — it looks performative.

This is not a religious claim, though religion agrees with it (blessed are the pure in heart, for they shall see God, Matthew 5:8). It is a structural claim about how moral perception works. The perceiver must have been formed into the kind of being to whom the object of perception is visible. Ethics is an empirical discipline with a hard prerequisite: the observer has to be a certain kind of person, or the data won't show up as the data it is.

Extend this beyond ethics. A person with no reverence cannot see the sacred. A person with no patience cannot see slow growth. A person who has never loved cannot see another couple's love correctly — they will classify every tender gesture as performance or manipulation, because those are the only categories their formation gave them. Vice is not just moral failure. It's epistemic blindness.

Any framework that treats the moral state of the knower as irrelevant to what the knower can know is denying something Aristotle noticed in 350 BC and every serious moral tradition has confirmed since.

7. Literary knowing is not just experiential knowing

The empiricist's last retreat, when pressed on points 1–6, is to flatten literature into experience. Sure, novels teach you things — but that's just experiential knowing. We already have a category for that.

The flattening is an equivocation on experience. There are two different things being called by the same name:

• The experience of reading — photons on the retina, the phenomenology of attention to the page, the felt sense of the prose. This is experience in the empiricist's sense: a stream of inputs, revisable beliefs on the other end.
• What the reading does to the reader over time — the widening of moral perception, the acquisition of tacit pattern-recognition, the reshaping of the aperture itself. This is not an experience. It's a formation effect. It's what the experience builds in the one who undergoes it.

If experience means only the first thing, then literature is one more data-stream and nothing distinguishes War and Peace from a Napoleonic tactics manual — which is absurd. If experience means the second thing, then "experiential knowing" already concedes the formation thesis, and we are no longer doing radical empiricism.

The empiricist wants the word to mean both at once. He uses the first meaning when defending the framework and the second meaning when explaining why literature matters. The switch is what keeps the framework alive. Name the switch and the framework dies.

---

The shape of the seven

Every one of these is a different face of the same claim: the knower is not fixed. The knower is made, by formation, before any reasoning can begin; continues to be remade, by literature and practice and virtue and desire; and cannot see past the shape he is currently in except by being reshaped.

Radical empiricism depends on a fixed knower who updates variable beliefs. The seven points each show, from a different angle, that the knower is the variable and the beliefs are the surface. Update the knower — through formation, through literature, through virtue, through contemplation — and beliefs reorganize as a second-order effect. You don't revise your way to a new self. You're given a new self, and your revisions run on top of what you've been given.

The reason a radical empiricist has to refuse all seven is not that any one of them is individually implausible. Each is individually obvious to anyone who has taught a child, read a great novel, mastered a craft, fallen in love, or sat with a dying friend. The reason he must refuse them is that granting any one of them destroys the framework's central claim — that the knower is a neutral platform for belief-revision. Grant formation, and the platform was built. Grant literature, and the platform reshapes. Grant tacit knowledge, and the platform exceeds its own vocabulary. Grant contemplation, and the platform's purpose isn't prediction. Grant desire, and the platform's foundation wasn't examined. Grant virtue, and the platform had moral prerequisites. Grant that literary knowing is formation, and the platform was never what it claimed to be.

There is no order in which to concede one without the others following. That's why the seven move together, and that's why honest engagement with any one of them is the whole argument.