Stop Saying “Anything Is Fine” and Intimacy Gets Riskier, and More Real

Where would you like to eat. Anything is fine. What would you like to do on Saturday. I do not mind, whatever you want.

Said a thousand times, by almost everybody, and understood on both sides as a small act of accommodation — she is being easy, she is not imposing, she is leaving room.

The companion essay that opened this month argued that asking for something costs a measurement rather than a value, and that a request is what makes a refusal possible. This one takes the opposite utterance: not the request but its cancellation, and what a culture of cancellation produces at scale.

The month’s private sentence is: I would rather say what I want and let someone choose freely. The atlas title claims that dropping the phrase makes things riskier and more real, and both halves of that claim turn out to be specifiable rather than atmospheric.

Begin with what the phrase actually does, because it almost never means what it says.

Genuine indifference exists and is common — there are evenings where any of four restaurants would be equally good, and saying so is a true report. But the phrase is used enormously more often than genuine indifference occurs, which means most of its uses are doing something else.

What they are doing is transferring a decision. After the phrase, he must choose. And the important part is what travels with the decision: if the evening is mediocre, the choice was his. If the restaurant is wrong, the error has an author and it is not her.

So the phrase is a transfer of liability. It moves both the decision and the accountability for its outcome onto the other party, and it does so in a form that looks like a concession rather than an allocation.

Which produces an asymmetry worth stating plainly. After anything is fine, one person can be wrong and the other cannot. He can guess badly; she cannot, because she made no claim that could turn out to have been mistaken. The phrase manufactures a game in which only one player is exposed.

And none of that requires any bad intention. Almost nobody says it in order to shed liability. It is simply what the utterance does, the way a door closes whether or not anybody meant to shut somebody out.

A habit this universal is being reinforced by something, and the reinforcement is real rather than imagined.

Wanting nothing imposes no cost on anybody. A person with no preferences is cheap to accommodate, easy to plan around, and never the reason a group has to negotiate. That is a genuine social good, and the people who supply it are genuinely valued for it.

So the phrase is a politeness technology, and it works. It keeps surfaces frictionless, it spares everybody a negotiation, and it signals something the culture prizes: that she is not difficult. Being easy to be with is a virtue and not a failing, and an essay that treated the phrase as simple cowardice would be missing what it is for.

The difficulty is that it is priced as though it were free. The cost does not fall at the moment of speaking — it falls later, distributed thinly across years, in a form nobody attributes to the phrase. At the moment of speaking there is only the benefit: the other person relaxes, the exchange is smooth, and nobody has had to want anything in public.

Which is the profile of most habits that are hard to drop. Immediate reward, deferred and diffuse cost, and no moment at which the bill arrives with an itemised account of what produced it.

There is a property of the phrase that makes it unlike an ordinary withheld answer, and it is worth isolating because it explains why the situation gets worse on its own.

A single use withholds information on one occasion. Repeated use does something more thorough: it teaches the other person that asking produces nothing.

And a question that reliably produces nothing stops being asked. Not out of indifference — out of efficiency. Why would anybody keep posing a question whose answer is always the same non-answer? So after enough repetitions the question disappears from the relationship entirely, and what replaces it is him deciding, silently, because that is what has always happened.

Which means the phrase is self-eliminating in the worst possible way. It removes the very occasions on which it could have been retired. She cannot begin stating preferences on Tuesday, because nobody will ask on Tuesday, and unprompted preference-stating is a considerably larger act than answering a question.

So the mechanism ratchets. Each use makes the next question less likely, which makes the next opportunity rarer, which makes eventually stating something a bigger and more conspicuous event than it ever needed to be. A phrase used to avoid making a fuss has arranged matters so that any change requires making one.

Here is the part that makes this a cultural argument rather than a description of a personality, and it explains why individual resolve so reliably fails.

Consider a group deciding where to eat. If everybody defers, the decision falls to whoever is first willing to state a preference — and that person is then doing something visible that the others are not. They have introduced a want into a space where nobody else has one, which reads as assertion, or as taking over, or simply as being the person with opinions.

So the cost of stating a preference rises with the number of people deferring. In a group of two it is a minor act. In a group of six where five have said anything is fine, the sixth is conspicuous, and the conspicuousness is produced entirely by the others’ deference rather than by anything in the preference itself.

Which makes the equilibrium self-enforcing. Every additional deferral raises the price for the next person, and nobody can lower it unilaterally. The individual resolution to be more forthcoming runs directly into a structure that penalises exactly that, and the penalty is not imagined.

This is why the phrase cannot be treated as a personal failing to be corrected by courage. It is a coordination problem of an entirely ordinary kind: a state in which each person’s best individual move produces a collectively worse outcome, and from which no single participant can exit without paying for everybody.

And it is worth noting that this is not confined to any one society, although the particular phrase and its intensity vary. Anywhere that imposing on others is more heavily sanctioned than disappearing, the same equilibrium forms, and it forms most strongly among people who are most considerate.

Which produces a consequence that is easy to observe and almost never named correctly.

Groups in this equilibrium reliably arrive at outcomes that nobody wanted. The restaurant that was fine. The holiday that suited no one in particular. The Saturday that passed pleasantly and that nobody would have chosen if asked directly.

And the usual explanation is compromise: everybody gave a little, so nobody got exactly what they wanted. That explanation is flattering and it is wrong. A compromise requires positions to be stated and then traded against each other. It is an achievement, and it produces an outcome that is responsive to what people wanted, even if imperfectly.

What happens here is not that. No positions were entered, so nothing was traded. The outcome was produced by whatever was left once everybody had declined to contribute — a default, a convenience, the thing nearest to hand, or the preference of the one person who could not stand the silence any longer.

So the outcome is a residue rather than a compromise, and the difference matters because a residue is not responsive to anybody. A compromise at least contains traces of what people wanted. A residue contains traces of who was most uncomfortable with an unfilled silence, which is a completely different quantity and one nobody would have chosen to optimise for.

Which explains a familiar and slightly melancholy experience: a pleasant evening that leaves no residue of its own, attended by several people each of whom privately believed they were accommodating the others.

The atlas title says that dropping the phrase makes things riskier, and the word is accurate rather than rhetorical. There are three specific exposures, and they are worth separating because they are not equally acknowledged.

The first is that a stated preference reveals taste, and taste is judged. Naming a restaurant, a kind of evening or a sort of music places her somewhere on a map other people are reading, and the map has status on it. This is rarely admitted as a reason and is frequently the operative one, particularly where tastes are finely graded and everybody knows the gradations.

The second is that the preference can be refused. That exposure is the subject of this month’s companion essay, which argued that refusability is precisely where the value of an answer comes from, so it will not be re-argued here beyond noting that it is real.

The third is the one nobody names, and it is the strongest deterrent of the three. Stating a preference makes her accountable for the outcome. If she chooses the restaurant and the restaurant is bad, the bad evening is hers. She has become the author of a disappointment, in front of somebody whose evening she spoiled.

Anything is fine is insurance against that specifically. It is not primarily about avoiding conflict or about being nice; it is about never being the person who picked the place where nobody enjoyed themselves. Which is why the phrase is used most heavily by people with a strong sense of responsibility rather than by people who do not care.

And the insurance is real. Dropping the phrase genuinely does expose her to all three of those, and no amount of reframing makes that untrue. The argument of this essay is not that the risk is illusory. It is that the premium is being paid in something more valuable than the claim.

The atlas title also says more real, and that half is specifiable too. Two things become possible which cannot exist at all under the phrase.

The first is being chosen specifically. If she has never stated a preference, nothing the other person does can be responsive to her — his accommodation has no target. He may be generous, attentive and entirely willing, and all of that generosity has to be aimed at a generic recipient, because that is the only recipient he has information about.

Which means the particular pleasure of somebody having got it right is structurally unavailable to her. Getting it right requires a prior statement to be right about. Under the phrase, the best possible outcome is a lucky guess, and a lucky guess cannot be received as attentiveness because it was not attentive — it was luck, and she knows it.

So the phrase does not merely withhold information. It forecloses an entire category of experience, and it forecloses the one she most often reports wanting: the sense of being known well enough that somebody could aim at her.

And the loss is silent, which is why it persists. Nobody notices the absence of an experience that never occurred. She notices only a vague sense that his kindness, which is real, somehow does not land — and there is no obvious explanation available, because nothing went wrong.

The second thing that becomes available is less comfortable and is probably more important.

When two people both state preferences, those preferences will sometimes conflict. She wants the quiet place and he wants the loud one. That conflict is usually experienced as a problem and counted as the cost of candour.

But a conflict between two stated preferences is information about two actual people. It is the first point at which either of them encounters anything resistant in the other — something that does not simply accommodate, that has its own shape, and that has to be dealt with rather than absorbed.

Which suggests a description worth taking literally. A relationship in which nobody states a preference is frictionless. And frictionless, in physical terms, is indistinguishable from not being in contact: two objects produce no friction when they are not touching. The absence of resistance is not evidence of compatibility. It is evidence of no interface.

That is what the atlas means by more real, and it is not a sentimental claim. Stating preferences produces occasional resistance, resistance is what contact feels like, and a relationship reporting none of it over years is not unusually harmonious — it is unusually unengaged, and both parties may be describing the absence of contact as peace.

This should not be overstated in the other direction. Manufactured conflict is not contact either, and nobody needs to generate disagreement to prove a relationship is real. The point is narrower: the occasional conflict is not the cost of stating preferences. It is part of what stating them delivers.

The practical content is small and is mostly about granularity rather than courage.

The reason stating a preference feels heavy is usually that people reach for a complete answer — naming the restaurant, specifying the evening. A complete answer takes the whole decision, which is why it feels like an imposition and why it carries full accountability for the outcome.

A constraint does neither. Somewhere quiet. Not too far. Something outdoors. I would rather not be in a group tonight. Each of those enters real information about her while leaving the decision entirely with him — he still chooses, he still gets to be the one who found somewhere good, and she has not taken over anything.

That is a better instrument for most occasions than a full answer, and it dissolves most of the third exposure as well. A constraint cannot be blamed for a bad restaurant in the way a named restaurant can, while still making his choice responsive to her rather than aimed at a generic person.

And the phrase retains a legitimate use, which should be protected. Genuine indifference happens, and reporting it honestly is not a failure of nerve. The test is small: if a stranger with no stake in the outcome asked her the same question, would an answer appear? If one would, she has a preference and is withholding it. If nothing appears even then, the indifference is real and the phrase is accurate.

None of this requires a conversation about the relationship, and none of it requires announcing a new policy. It requires answering one question with a constraint rather than with the cancellation, which is roughly four words of difference.

An arrangement here cannot operate on anything is fine. Preferences have to be stated in advance, specifically, in writing or in conversation, because nothing can be prepared otherwise. So this house requires precisely the practice this essay has been recommending, which looks like a point in our favour and needs examining rather than claiming.

The examination produces an awkward result. Stating a preference here carries none of the three exposures. Nobody here judges her taste, because nobody here is in a position to and because being unjudgemental is part of what was paid for. Nobody will be disappointed by her choice, because the person receiving it has no competing preference. And she is not accountable for the outcome in any way that costs her, because if the evening is not what she hoped, that is our failure rather than hers.

So what happens here is the practice with the risk removed. And the essay has just argued, at some length, that the risk is where the realness comes from — that the exposure is not the obstacle to the skill but a constitutive part of the thing worth having.

Which raises the only interesting question in this section, and we should not pretend to know the answer. Does risk-free practice transfer? It does in some domains: a musician practises alone and performs in public, and the private hours plainly help.

But there is a specific reason to doubt it here. In the musician’s case the difficult part — the notes — is the same in both settings, and only the audience is added. In this case the difficult part is not the saying. Anybody can produce the sentence. The difficult part is the exposure, and the exposure is exactly what a purchased setting removes. So the hours spent here rehearse the easy component and omit the one that was hard, which is the profile of practice that does not transfer.

The honest position is therefore narrow and slightly against us. Stating preferences here is pleasant, it is genuinely useful for finding out what she actually prefers when nothing is at stake, and it should not be mistaken for practice at the thing this essay is about. We also have an evident interest in a world where preferences are expensive to state elsewhere, and a reader who puts four of these words to the person she lives with on Saturday has done something this house cannot do for her and cannot substitute for.

This essay is cultural and structural analysis rather than clinical guidance, counselling or relationship advice. It describes no treatment, promises no recovery, diagnoses nobody, and does not tell any reader what to say to anybody.

The central claim — that the phrase is a transfer of liability rather than a report of indifference, moving both the decision and accountability for its outcome to the other party, so that one person can be wrong and the other cannot — is this Library’s own formulation. It is reasoned rather than measured, and the essay states explicitly that genuine indifference exists and that almost nobody uses the phrase in order to shed liability.

The account of the phrase as a politeness technology that is genuinely valued, with immediate reward and deferred diffuse cost, is offered as a defence of its function before any criticism of it. Being easy to be with is described as a virtue rather than a failing.

The claim that repeated use teaches the other person that asking produces nothing, so the question disappears and the phrase removes the occasions on which it could have been retired, is this Library’s argument. Reasoned rather than measured.

The coordination argument — that the cost of stating a preference rises with the number of people deferring, making the equilibrium self-enforcing and individual resolve insufficient — is a structural claim of an ordinary kind. It is stated not to be confined to any one society, no country, religion or era is named, and it notes that the equilibrium forms most strongly among the most considerate.

The distinction between a residue and a compromise is this Library’s own. It is offered to correct a flattering explanation rather than to criticise anybody, and it makes no claim about frequency.

The three exposures are separated deliberately: the taste exposure and the accountability exposure are this Library’s formulations, while the refusability exposure is explicitly handed back to this month’s companion essay rather than re-argued. The essay states plainly that the risk is real and that no reframing makes it untrue.

The claim that accommodation requires a target, so that the pleasure of somebody having got it right is structurally unavailable without a prior statement, is this Library’s formulation and is reasoned rather than measured.

The observation that frictionless and not-in-contact are the same physical description is offered as an analogy rather than as a finding about relationships. The essay explicitly warns against the opposite error, stating that manufactured conflict is not contact and that nobody needs to generate disagreement.

The practical suggestion to state constraints rather than complete answers, and the test of whether an answer would appear for a stranger with no stake, are derived from the essay’s reasoning rather than from evidence about outcomes. They are not instructions, and the legitimate use of the phrase for genuine indifference is explicitly protected.

The section about this house declines to claim the apparent credit, observes that stating a preference here carries none of the three exposures, and raises the transfer question without asserting an answer — while giving a specific reason for doubt, namely that the difficult component is the exposure rather than the utterance. It discloses an interest in preferences being expensive to state elsewhere and ends by naming something this house cannot substitute for. No figure, proportion or statistic appears anywhere in this essay.

I would rather say what I want and let someone choose freely.

Read against this argument, the phrase that sentence is replacing turns out to have been doing the opposite of what it appeared to do. Anything is fine looks like the maximally generous utterance — it hands him the whole decision. What it actually hands him is the whole liability, with no information attached, and then requires him to be responsive to somebody who has declined to be specified.

Which is why the month’s sentence puts choose freely at the end rather than at the beginning. He cannot choose freely in the absence of information; he can only guess and hope. Saying what she wants is not the part that constrains him. It is the part that gives him something to be generous about.

And the risk does not disappear under this analysis. It relocates. She gives up the insurance — the guarantee of never having been the one who picked the wrong place — and receives, in exchange, the possibility of being accommodated specifically rather than generically, and the occasional experience of friction, which is the only available evidence that two people are in contact at all.

The smallest version is four words on a Saturday. Somewhere quiet, perhaps. Not a position, not a demand, not a conversation about the relationship — just enough information entered into the record that whatever happens next is aimed at her, and can therefore be received as having been aimed at her.