What Is Curiosity

The picture almost everyone carries is a hunger. Curiosity is an appetite for information; some people are born with a bigger one; children have the most and lose it; and the cure, if you want to call it that, is an answer. The language is entirely alimentary. We speak of a thirst for knowledge, of a voracious reader, of devouring a subject, and above all of satisfying our curiosity — a verb we otherwise reserve for hunger, and use here without noticing. It is a decent picture. It gets one thing exactly right, which is that curiosity is felt rather than chosen. And it fails on two of the most ordinary experiences anyone has with it.

What the appetite picture commits you to

Take the hunger seriously for a moment and count what it requires. That curiosity is a property of the person, so it goes where you go and reads roughly the same across subjects. That information feeds it, so being informed should leave you more contented and less curious. And that you can point it — that deciding a subject is important is a way of becoming interested in it.

Two ordinary cases break the picture, and the interesting thing is that they break it from opposite sides.

Case one: the answer ends it instantly. Think of the last thing you looked up. The origin of a phrase, the outcome of a match, the name of the actor you half-recognised. In the seconds before, the wanting was sharp enough to interrupt what you were doing. In the second after, it was not diminished — it was gone, completely, and you could not manufacture it again about that item if you tried. An appetite does not do this. A hungry person given one bite is still hungry. Whatever ends totally on first contact with its object is not a hunger; it is a gap being closed.

Case two: the emptiest subjects produce the least. Open a page of category theory, or a page of Sanskrit, or the live scoring rules of a sport you have never watched. Your ignorance there is close to total, so on the hunger account this is where the appetite should be fiercest. It is not. It is flat — not hostile, not resistant, simply flat, in a way that a subject you half-know never is. The regions of your ignorance that generate the most curiosity are the ones you have already been into.

Put the two together and the appetite account has to say that information both ends curiosity and is a precondition of it. That is not a hunger with a complication. That is the wrong category.

Going to the foundation

I work from a small set of primitives, and the one that does the work here is the first shared one. S1: when the unconditioned is conditioned, a circle appears. Every point on the circumference is a condition. Every named condition is an idea. And the circumference is infinite — no enumeration closes it.

Two more, briefly. P6: a condition is a prerequisite. It does not cause, it does not make possible, and it does not happen. P7: an idea is a named condition; naming does not create it, naming makes the relationship conscious. And a locked distinction the vocabulary insists on, which turns out to be the whole hinge of this piece: named and nameable are not interchangeable. Conditions are inexhaustible because they are nameable without limit. Named conditions, at any given moment, are finite — a finite set, held by a particular person, on an infinite circumference.

That structure has a consequence nobody has to argue for. A finite named interior sitting inside an infinite circumference has a boundary. There is an edge between what you have named and what you have not, and that edge is not a shortage or a defect. It is what having named anything at all looks like from the inside. Everyone has one. What differs between people, and between one person’s subjects, is not whether the edge exists but how much of it is legible to them.

Because not all of the unnamed is equally near. Some of it has become readable — you can formulate the question even though you cannot answer it. Most of it has not; you could not put the question into words, and would not know you were missing anything. The doctrine already carries this distinction in a neighbouring place: when the foundation treats the pull of a temptation, it is precise that the precondition is readability rather than appetite, which is why a demonstration can produce a pull where none existed. Nobody is drawn to what they cannot read. The same holds here.

Read the claim carefully, because it relocates the whole thing. Curiosity is not something you have more or less of. It is a position you are standing in, and the position has a measurable feature: how much of your own edge you can currently read. On this account, naming does not consume the edge. It moves one point of it inside, and in doing so it makes points adjacent to that one readable that were not readable before. The interior grows, and the legible perimeter grows with it.

The reward for learning something is not a smaller frontier. It is a longer one, drawn in a place you previously could not see.

Figure 01

Three positions on one circumference

• The strangera subject you have never entered
• The inhabitanta subject you have worked in for years
• The finishera subject you believe you have completed

• Named — conditions you hold as ideas
• Readable and unnamed — the edge
• Unreadable — you could not form the question
• The believed boundary

Both cases, resolved

The answer ends it because a specific curiosity is one readable, unnamed point — and naming moves that point inside. Nothing was fed. One segment of edge closed. This also explains the thing everyone notices and nobody explains: answering a question characteristically produces two more. The closure exposes adjacencies. That is not a hunger returning after a meal; it is a perimeter being redrawn longer at the place where it was just shortened.

The unentered subject is flat because there is no edge there to stand at. Category theory is not withholding anything from you. You have named almost nothing in it, so almost nothing in it has become readable, so there is no question you can form — and a question you cannot form is not a question you are missing. What feels like indifference is illegibility. This distinction matters more than it sounds, because the two feel identical from the seat and they have opposite remedies.

The flat that is not boredom

In an earlier piece I argued that boredom is not an empty room but a saturated denominator. The Reality Equation, P4, gives what you experience as R = A / E — the actual over the expectation waiting to receive it — and surprise as the logarithm of the magnitude. A dull hour is not an hour in which nothing arrives; something always arrives. It is an hour in which everything that arrives lands where the model already pointed. A long run of readings near unity. That argument stands, and this piece needs it.

But it leaves a gap I can now close, and closing it does not overturn the earlier piece — it says where the earlier piece applies. Consider the page of unfamiliar notation. Your expectation about that page is certainly not covering its content; you have no model of its content at all. So why is it flat rather than thrilling?

Because E has terms only at the descriptions under which you have named something. P4 is explicit that E has two components, subconscious prediction and ideas, and it is the second that carries the naming. Facing that page, your prediction is well-formed at one description — marks on paper, arranged in lines — and the page conforms to it exactly. At the description that would matter, you have no terms at all, so the content cannot miss anything. A reading near unity comes back either way.

What the laboratory found, and where it disagrees with itself

The empirical literature on curiosity is worth reporting here, and worth reporting carefully, because three findings that are routinely quoted as one finding are in fact about three different quantities. I mark this whole section as report, not grounds. The argument above does not lean on it, and the reader should be able to see where it fits and where it does not.

So the record does not hand me a clean monotonic law, and I am not going to pretend otherwise. What it does hand me is a shape, and the shape is the one the first construction predicts: low at the unentered end, high in the middle, and — this is the part worth pausing on — low again at the end where a person believes he has covered the subject. The account also gets a specific detail right that the appetite picture has no way to accommodate: the quantity that predicts curiosity is estimated knowledge rather than actual knowledge. On the appetite account that is strange, since an appetite tracks a real deficit. On this account it is required, because what a host can read is the edge as it appears from where he stands, and that is an estimate by construction.

One more finding is worth the space because it is not a rival. A separate line of work argues that expertise relocates where curiosity lands rather than raising or lowering its level — experts choose harder material, novices easier. Relocation is precisely what the first construction describes: naming moves the edge, it does not shorten it.

The falling arm

Which leaves the third position in the figure, and it is the one I find most useful. Why should curiosity drop at the far end — among people who know a subject well and have begun to feel they have it?

S1 is unambiguous that the circumference is infinite and no enumeration closes it. Named conditions at any moment are finite. Together those forbid the state the falling arm reports: there is no position from which a subject has been exhausted, because exhaustion is not a shape the geometry offers. So the drop is not the edge shortening. It is a host reading his own edge as nearly closed — and since what he can act on is the reading and not the boundary, the drop is real in its effects while being false in what it reports.

Whose curiosity is it

Here is where the framework says something that costs. P8: a condition is prior to what it conditions, and where a person’s naming is itself conditioned, that person is not the origin. The formula the foundation uses is blunt — ideas have people; people do not have ideas.

Applied here, that removes curiosity from the list of things you can take credit for. And the evidence is not theoretical, it is daily. You cannot make yourself curious about a subject by deciding it matters; anyone who has tried to care about a required course knows the decision does no work. And you cannot stop being curious about the one that has you, which is why people are awake at two in the morning reading about a thing that is no use to them. The foundation is direct about the second case: persistence proves structure, not nobility. An interest that will not leave is telling you something about the conditions in play, and nothing whatever about your merit.

R12 completes the thought. Ownership is a category error; the manifestation is accountable. So the accountable object was never the interest. It is what you made while under it — the artifact, which is the only thing that carries a mark at all.

And R14 supplies the part that stings. Every idea is biased toward itself; orientation and prejudice are the same act. Being deeply curious about one thing therefore makes you specifically and structurally incurious about its opposite, and — this is the trap — you will not experience that as your own orientation. You will experience the other thing as boring. Not uninteresting to me: boring, as a property of it. Nearly everything anyone says about which fields are dull is that error being reported as an observation. This system is an idea and is biased toward itself in exactly the same way, which is worth stating rather than hoping nobody notices.

What this costs

It costs the finish. If curiosity is the length of an edge, and S1 puts that edge on an infinite circumference, then no amount of naming reduces the unnamed and no subject arrives at a state of having been covered. The picture where you read enough, finally understand the thing, and settle into a warm competence with the questions behind you — that state is not late, and it is not being withheld. It is not a place. R10 says the same thing from the other side: every artifact is a miss, so the making never catches the exemplar either.

Two things that is not saying. It is not saying stop. The foundation is careful to distinguish settling — which is ceasing to correct, and which it calls the only sin — from resting, and from finishing a piece of work and putting it down. You are allowed to close a book. What you are not offered is the state in which there is no next book because the subject has been used up.

It is also not licence to close an open problem in passing, and this piece sits on top of one. The foundation lists, as unresolved, the circularity in the relation between an idea and what makes it readable: the idea fixes the angle, and the angle determines what is readable. Everything I have written about naming lengthening the readable edge lives inside that circle, and I have not escaped it — I have described the loop more concretely without giving it an independent handle. Closing it would take showing readability as a trajectory with a parameter that is not defined by the thing it is supposed to explain. I do not have that. The construction above is usable and it is not a resolution.

On an ordinary Tuesday

1. Diagnose the flatness before treating it. When something bores you, ask which flat it is. Could you predict what came next, or could you not read what came next? Those are opposite conditions that feel identical, and treating one as the other is why people abandon subjects they would have loved.
2. Buy vocabulary, not motivation. Interest is not purchasable and you already know that, because you have tried. Legibility is: twenty terms and one worked example will make a field readable, and the curiosity either arrives or does not — but it had no chance at all before. This is the single actionable claim in the piece.
3. Stop auditing your interests for worthiness. By P8 they were not your selections, so they are not your credentials and not your indictment. The thing you are accountable for is what you make while holding one, per R12. The interest was never the credential. The artifact is.
4. Distrust the feeling of having finished a subject. If a field has gone quiet on you, the geometry says the field did not close — your reading of it did. The test is cheap: find someone standing at a different angle to the same material and see whether one of their questions is unplaceable to you. If it is, the quiet was yours.

And one line to keep, if the rest of it goes: