I Don't Know

Three of the hardest words in the language, and the ones that hold up nearly everything you are right to trust. A defense of admitted ignorance against the fluent, the certain, and the wrong.

Somebody asks you a question and you do not have the answer. Not a hard question, necessarily -- often an ordinary one, the kind you feel you ought to be able to field. And in the half-second before you speak, you feel the pull. Not toward the truth, which is unavailable, but toward the production of something: a plausible shape, a confident sentence, an answer-shaped object to set down in the silence where the answer should be.

Three words would end the pull honestly. I don't know. They are among the hardest in the language to say out loud, and nearly everything worth trusting is built on the backs of the people willing to say them.

One -- the wisest man in Athens

Wisdom began as an admission

The Western argument opens with a man who built an entire philosophy on not knowing things. The oracle at Delphi declared that no one in Athens was wiser than Socrates -- who was quite sure he knew nothing worth the name, and treated this as a riddle to be solved rather than a compliment to be enjoyed. He went looking for someone wiser. He found politicians, poets, craftsmen: each fluent and certain within his own patch, each mistaking a narrow competence for wisdom at large.

His conclusion was not that he was clever. It was that he alone was not carrying around a false inventory. The others believed they knew things they did not. He, at least, knew the shape of his own ignorance, and he thought that single advantage decisive.

The one who knows he doesn't

Socrates -- Plato's Apology

"I am wiser than this man; it is likely that neither of us knows anything worthwhile, but he thinks he knows something when he does not, whereas when I do not know, neither do I think I know." The advantage is not a larger store of answers. It is an accurate map of the blanks.

Notice what this makes wisdom. Not the most boxes filled in, but the truest sense of which boxes are filled. The wise person is not the one who knows the most; it is the one whose account of what he knows is not a fiction.

Two -- worse than a lie

The opposite of knowledge is not error

We tend to picture the enemy of truth as the liar -- someone who has the fact and states its opposite. But the liar is a strange sort of tribute. To lie well you must know the truth and keep it in view the whole time, steering carefully around it. The liar is oriented by the truth; he needs it, the way a smuggler needs the border.

There is a second figure who is worse, and far more common, and the philosopher Harry Frankfurt gave him his proper name.

On indifference

Harry Frankfurt -- On Bullshit, 1986

The liar cares about the truth and hopes to lead you away from it. The bullshitter does not care at all. He is neither for the truth nor against it; he is producing speech to serve some other end, and whether the words match reality is simply not his concern -- "he just picks them out, or makes them up, to suit his purpose."

This is the deeper danger, because it needs no villain -- only a person, or a system, under pressure to produce an answer and indifferent, for that moment, to whether one exists. We have lately taught machines to do exactly this at industrial scale: to generate the fluent shape of a reply with the same evenness whether or not anything at all stands behind it. But the machines only automated a very old human reflex. Long before any of them, the meeting that must reach a decision, the expert who must not appear stumped, the student who must fill the page -- each already knew how to manufacture the answer-shaped object. Frankfurt's warning was that this is no minor sin. A sincere indifference to the truth wears away the ground faster than the occasional deliberate lie, because at least the liar has conceded that the truth matters.

Three -- the even voice

A wrong answer should not sound like a right one

Here is the property that ties the two together. The liar and the bullshitter both speak in the same tone whether they happen to be right or not. Their confidence carries no information at all.

Good judgement has the opposite signature. It is not certainty; it is calibration -- the fit between how sure you sound and how likely you are to be right. The calibrated forecaster who says seventy per cent is wrong three times in ten, and means to be. The calibrated mind runs its confidence up and down with the evidence, so that the strength of a claim becomes, itself, a piece of information you can act on.

The first principle is that you must not fool yourself -- and you are the easiest person to fool.

Feynman's line names the failure exactly. The uncalibrated voice is smooth at every point. It delivers the guess and the established fact in a single register, and in doing so it destroys the one signal you needed most: the ability to tell, from the outside, which is which. A person -- or an institution, or an instrument -- that is always certain has not thereby become more reliable. It has merely stopped telling you when to check.

Four -- the tax on honesty

We punish the ones telling the truth

If admitted ignorance is so valuable, why is it so rare? Because it is taxed.

In any room where answers are the currency, the person who says I don't know pays a cost the confident bluffer does not. He looks less able in the moment. He loses the account, the argument, the vote, to a man with no more knowledge but fewer scruples about advertising it. Across a career, across an institution, this pressure sorts people quietly: the honest hedge and lose ground, the fluent assert and gain it. We are not, most of us, selecting for people who know. We are selecting for people who sound as though they do -- a different and much cheaper thing.

The remedy is not new either. It is only unwelcome.

On the possibility of error

Oliver Cromwell -- 1650

"I beseech you, in the bowels of Christ, think it possible that you may be mistaken." Written to men who were entirely certain. The sentence has outlived the quarrel that produced it because the request it makes -- hold your own certainty at arm's length long enough to weigh it -- is the one nobody wants to grant himself.

Five -- the load-bearing words

Doubt is what makes your certainty worth anything

None of this is a brief for endless hedging -- the person who will not commit to the time of day. I don't know is not a destination; it is an honest starting line. Its full form almost always continues: I don't know, but here is how we would find out, or I don't know that, but I know this next-door thing, and here is where the two part company. It opens the honest version of the conversation that the confident guess quietly forecloses.

And it is what lends your other words their weight. If you are known to say I don't know when you don't, then on the occasions you do speak with confidence, the confidence carries real information -- it is backed by the fact that you were willing, on other days, to withhold it. Trust is not built on the days you were sure. It is built on the reliability of your refusals. A yes is worth only as much as the no standing behind it was real.

There is a cost that usually goes unmentioned, and it runs the opposite way from the one everybody fears. We worry that admitting ignorance will make us look small. We rarely count what the alternative costs.

The person who can never say I don't know has not escaped his ignorance. He has only mislaid the instrument that measured it -- thrown away the gauge and kept the engine, and now runs blind at speed, unable even to locate the thing he does not understand, because he has trained himself to paper over the blank the instant it appears.

You do not, in the end, trust the voice that is always certain. You trust the voice that told you, once, that it wasn't.