Not Knowing What You Want Is a Skill That Was Never Trained.

The form has a text box and the question above it is friendly: what would you like. She has typed nothing for four minutes. She tries "to relax", deletes it because it is what anybody would write, tries "I'm not really sure", deletes that too because it looks like she has not thought about it, and ends up with the cursor exactly where it started.

What is happening in those four minutes is not indecision. She is looking inward for a preference, the way you would look in a drawer, and the drawer is empty — and because everybody talks about knowing what you want as though it were a matter of honesty with yourself, she draws the obvious conclusion. That there is something missing in her. That other women have this and she does not.

This is one of the quiet essays and it takes that conclusion apart. The Library has already published on what happens when a woman can only phrase her wants in the passive voice — to be wanted, to be chosen, to be held — and that essay is credited rather than repeated. This is the case one step further back: the woman who cannot produce an answer in any voice at all. And the account of it is not repression, and not damage.

Start by noticing that the question assumes a model, and the model is wrong.

The model everybody uses is retrieval: a preference exists somewhere inside her, fully formed, and knowing what she wants means locating it and reading it out. On that model, finding nothing means nothing is there — which is why the blank feels like a verdict.

But telling two things apart is not retrieval. It is discrimination, and discrimination is a perceptual skill that gets built. Consider somebody who tastes wine professionally. She was not born able to distinguish one from another; she tasted a great many, was told what she was tasting, noticed the difference between this and that, and over time the distinctions sharpened until they were obvious to her and invisible to everybody else. Nothing was retrieved. A capacity was trained.

The training has three parts and each one matters. Repeated exposure, so there is something to compare. Attention at the time, so the experience is encoded rather than merely undergone. And feedback — somebody naming it, or the simple fact of having chosen and then noticed the result — so the distinction has a label and can be recalled.

Now ask how much of that a woman has had, on this particular subject. Not the exposure, necessarily: she may have had years of it. The attention, often not, because she was managing what was happening rather than noticing it. And the feedback almost never, because nobody asked, and on the occasions when she did say something nothing visibly changed. Two of the three inputs are missing, and the third arrived unlabelled.

A capacity with no training data does not produce answers. That is not a flaw in her. That is what an untrained discrimination does.

It is worth being specific about the middle input, because a reader will reasonably object that she has had plenty of experience and still cannot answer.

Attention has one location, as this chamber has argued, and hers has usually been outward. She was monitoring whether it was going well for the other person, whether she was taking too long, how she looked from that angle, whether the noise was carrying, what time it was. All of that is real cognitive work and it happens in the same place the noticing would have to happen.

So the exposure occurred and produced almost no data, in the way that driving the same route for ten years produces very little memory of the buildings. She was present and she was occupied. What she has afterwards is not a record of what she liked; it is a record of how it went.

And there is a second reason, which is more particular. To register a preference you have to have had at least two things to compare. A woman whose experience has been with one person for many years, in a pattern that settled early, has had one sample repeated — which tells her about that, and cannot by itself tell her what she prefers, because preference is a comparison and she has not been given one.

None of this is anybody being blamed. It is an account of why the drawer is empty that does not require her to have anything wrong with her.

If the capacity is trainable, the question is where, and the answer is much further down than anybody expects. It does not start with sex and it does not start with anything erotic.

It starts with the temperature of a room. With whether she prefers the window open. With which of two chairs she would rather sit in, which hand she would rather be held by, whether she likes the pressure heavier or lighter, whether she wants the music on. These are absurdly small and that is exactly the point: they are the low-stakes repetitions where a discrimination can be built without anything riding on the answer.

The training method is the boring one that works for every perceptual skill. Notice, say, observe. Notice a mild preference. Say it out loud, to somebody or to nobody. Observe what happens when it is acted on — and that third step is the one that has been missing her whole life, because a preference that is never acted on produces no feedback and therefore trains nothing.

What she will find, if she does this for a few weeks, is not that she suddenly knows what she wants in some large sense. It is that the answers start arriving faster on small things, and that the machinery involved turns out to be the same machinery. The woman who can say within a second that she would rather have the window open has a working discrimination. The one who takes four minutes on a text box does not yet, and the difference between them is practice rather than depth.

One more thing about the method, because it is the part people skip. The noticing has to happen at the time, not afterwards. A woman who reviews an evening later and tries to work out what she liked is doing history rather than training, and history is unreliable: she will remember how it went rather than how it felt, because that is what she was attending to. The whole value of the small repetitions is that they are small enough to notice DURING.

And there is a real asymmetry worth naming: the easiest preferences to find are the negative ones. Most women can say what they do not want long before they can say what they do. That is not a failure either — a dislike is a sharper signal and it trains faster — and it is a legitimate place to start. A list of noes is data. It is the first data most women have ever collected about themselves.

The documentary seeded for this row consists substantially of women describing their own sensation, in specific language, on camera. The usual reason to praise it is that it puts an under-discussed subject on the record.

The reason it matters for this essay is different and more technical. Listen to how those women speak. They distinguish. They say this rather than that, here rather than there, this kind of pressure and not the other. That is what a trained discrimination sounds like from outside, and it is worth noticing that it sounds like ordinary competence rather than like unusual self-knowledge.

And the training that produced it is visible in the same accounts: these are people who have paid attention over time, who have had the language to label what they noticed, and who have been in situations where saying it made a difference. The fluency is downstream of those three things — the same three this essay named earlier — and not a personality trait some women happen to have.

The limitation, plainly: a documentary selects and edits, its participants are people who agreed to speak on camera and are not a random sample of anybody, and nothing in it is evidence about what is typical. It is read here for one thing only, which is what the output of a trained discrimination sounds like.

The book seeded for this row is a well-known guide to non-traditional relationships. Its lifestyle argument is not adopted here and is not this essay’s subject; what it is used for is one narrow structural point that its tradition worked out earlier than most.

That point is that a preference which stays inside a person cannot be negotiated, and therefore cannot be met except by accident. Any arrangement between people that departs from a default script has to make the preferences explicit, because there is no script to fall back on. That community discovered, out of necessity, that unspoken wants are functionally absent wants.

Take the structural point and leave the rest. Saying it out loud is not a communication nicety on top of knowing; it is part of how the knowing gets built, because speech is where a vague inclination becomes a specific claim, and a specific claim is the only thing that can be acted on and therefore the only thing that generates feedback.

Which explains something that puzzles women in this position: why thinking about it harder does not help. She has been trying to solve a training problem by introspecting, and introspection with no new data just re-reads an empty drawer. The way out is not inward. It is through several weeks of small said-out-loud preferences and what happens to them.

You do not owe anybody a filled-in box. If you have been staring at a friendly question and producing nothing, you are allowed to write that you do not know yet — and anyone who treats that as a failure to prepare has misunderstood what they were asking for. Not knowing is a legitimate starting state and it is far more common than the confident answers suggest.

You do not owe anybody the conclusion that the blank means something is missing in you. What it means is that a capacity was never exercised, in a life where nobody asked and nothing you said was acted on. Those are facts about the conditions, and a capacity that has had no training data behaves exactly like this in everybody.

And you do not owe yourself the search. You have probably been looking inward for years and finding nothing, and getting quietly worse about it each time. You are allowed to stop looking in there. The thing you are after is not in the drawer; it is built outside, in small, unglamorous, repeated noticings that have nothing to do with anything erotic and cost nothing to begin.

It should be said first that this house asks the question at the top of this essay. The map a woman fills in before anything is arranged asks what she wants, might want, and does not want — and a house that asks a question a great many women cannot answer has to be honest about what that costs them and what it does with a blank.

So the obligations are recorded. An unanswered box is a complete answer here and is treated as one; nobody is asked to produce something in order to proceed, and no evening is withheld because a woman could not name a preference. The question is asked because a preference that exists should be recorded, not as an entrance examination.

The second is about the negative column, which is the one this essay said trains fastest. The map asks for noes with the same weight as yeses, and a woman who can fill only that side has supplied something genuinely usable rather than half of something. That is stated at the point of asking rather than left for her to discover.

And the limit. An evening here is a low-stakes repetition with feedback — a place where a small preference can be said and acted on immediately, which is precisely the training loop this essay describes. That is real and it is narrow. It does not tell her what she wants, it is not a substitute for the weeks of small noticings that do the actual work, and a house that sold it as self-discovery would be selling one repetition as a training programme.

This essay is not clinical, psychological or sexual-health guidance and describes no treatment. Where an inability to identify wants is accompanied by persistent low mood, numbness, distress, or follows an experience the reader has not been able to process, the question is not the one this essay is asking and belongs with someone qualified to help; nothing here should be read as a substitute for that.

The account of preference-identification as a trainable perceptual discrimination built by exposure, attention and feedback is this essay’s own argument. It is offered as an analysis rather than as a research finding: the comparison to trained sensory expertise is an analogy, no claim is made about any specific learning mechanism, and no figure appears anywhere.

The book seeded for this row is a guide to consensually non-traditional relationships. Its lifestyle argument is explicitly not adopted or recommended, and it is used only for the structural point that unspoken preferences cannot be negotiated. The documentary is read for one feature — what a trained discrimination sounds like when spoken — with the explicit note that it selects and edits, that its participants are not a random sample, and that nothing in it is evidence about what is typical.

The passive-voice account of how a woman phrases her wants is credited to this Library’s earlier essay and deliberately not restated; this essay addresses the prior case of producing no answer at all. The observation that attention has one location is likewise credited rather than repeated.

The section on the house states that the house itself asks the question this essay is about, and records the obligations: a blank is a complete answer and nothing is withheld for it, noes are asked for with the same weight as yeses and that is said at the point of asking, and an evening is one low-stakes repetition rather than the weeks of practice that do the work. Nothing is promised.