Years ago — the place does not matter — someone had authority over you and used it, and got the decision wrong, and you knew they had got it wrong, and it changed nothing at all.
Consider the case that comes to your own mind. Mine is a manager, early in a job, telling me to do something his way — something I happened to know more about than he did. I told him so. I set out the facts, calmly, the way people do when they are sure they are right and expect being right to be enough. He heard me out. And then he said: I hear you, and we are doing it my way. That is my decision.
Here is the part worth pausing on. I did it his way. Not because he had convinced me — he did not try. Not because he was right — he was not, and it cost us, and we both knew that within the week. I did it because the decision was his to make. It was wrong, and it was binding, at the same time; both were true at once, and neither cancelled the other. Being correct, it turned out, was one thing. Being the person whose decision counted was another thing entirely. That day the two came apart.
You have a version of this. The teacher who marked it wrong and would not change the mark. The official at the counter who said no, when you could see they were mistaken, and the no held anyway. The parent — yours, or you as one — who said "because I said so," which is the most honest sentence in the whole business, because it admits out loud what the others conceal: that whether a decision binds you and whether it is correct are two different questions, and always were.
This is Philosophy for Us — philosophy for everyone, no degree required. This is the third of eight episodes on what it is to be governed by a machine — and tonight, for the whole hour, there is no machine. Last time we took the instinct "make it fairer" seriously and followed it to its honest end, and found it cut the other way: run the argument from bias all the way down and it argues for the machine, because the human it replaced was often worse. The easy answer failed. But something the fairness argument never touched was left standing — the part where no one had to answer to you. Tonight we find out what that is, and to do it we set the machine aside completely. We go back to the oldest question there is about power: what makes anyone's decision binding on you at all — and whether "because they get it right" is even the kind of answer the question is asking for.
Let us be slow and exact, because the rest of the series depends on getting this right, the same way last week depended on counting two wrongs and not one.
When a power makes a decision about you, there are three different questions you can ask, and almost everyone, almost all the time, runs them together into a single reaction: this is not right. We need to pull them apart and keep them apart.
The first question is whether the decision is correct — accurate, unbiased, the answer a careful and fair mind would reach. Call that accuracy. It is the question last week was about, and the machine, we found, can win it.
The second question is a different one: was this the body whose decision counts? Set aside whether the call was correct, and ask whether this was the one entitled to make it. Call that authority. Consider a referee who blows the whistle and calls a player offside, when the replay shows the player was a clear yard onside. The call was wrong. And the call stands. The goal does not count. The player has been wronged — and wronged legitimately, by someone whose job it was to decide and whose decision binds the game even when it is mistaken. There is the gap in its purest form: the call was wrong on accuracy and right on authority, at the same time. Now reverse it. A spectator three rows back saw the whole thing clearly and is shouting that the player was onside, and he is correct — more correct than the referee. His correctness changes nothing, because he holds no office. He has the accuracy and none of the authority. The referee has the authority and, this time, none of the accuracy. They are not the same property. They never were. They seem to be the same only because, on a good day, one person holds both.
Now the third question, the one easiest to miss, because it stands behind the other two. Grant that the referee has authority: where did he get it? By what right is he the one with the whistle? The answer is clean and worth holding onto. Before the game, both sides agreed to play under a rule-book that appoints him and binds them to his calls. They agreed in advance to be bound, on terms they could accept. That is why his wrong call still binds the players and the correct spectator binds no one. Call that third thing legitimacy: not whether the decision is correct, and not even whether this body has authority, but whether the authority itself was come by in a way those under it could accept. Change the story and the difference is easy to feel. Suppose the referee had appointed himself, was paid by the other team, and announced at kick-off that he would call the game however he liked. He still holds the whistle. His calls might even be accurate. But something has gone out of it: the wrong call no longer binds the players in the same way, and they know it, even if they are made to play on. The authority is there, the accuracy may be there, and the legitimacy is gone.
Three things, then. Getting the decision right. Being the one entitled to decide. And having come by that entitlement in a way those under it could accept. Put as plainly as they will go: is it correct, is it theirs to decide, and is their standing to decide one I could accept? On an ordinary day, about an ordinary authority, all three travel together and there is never any need to tell them apart. This series is not made of ordinary days. The reason we are here is that the machine forced these three apart and would not let us go on treating them as one thing. And once they have been seen apart, they cannot easily be run back together — about the machine, or about anyone.
Here is what we have really done. You came in last week still half-believing that a decision about you is legitimate when it is correct — that "is it fair?" and "did it get me right?" are the same question. Look at the referee and that belief no longer holds cleanly. Correctness is one of the three. It is not the other two. And the question the rest of this hour pursues is the one that opens as soon as you separate them: when a power decides about you, and you are asking not "was it right?" but "could I accept it as binding on me at all?" — what is that answer made of?
Let us take the first answer — a power binds you because it gets the decision right — and give it its most serious form, not a caricature. There is a serious version of it — a deep and well-worked account of why any authority binds anyone; and if you think "because it gets it right" is only the line technology companies use, you have not yet met the real thing, and you need to, because it stays in play for the rest of this series.
Start with the cases where obeying an authority is plainly the wise course and no surrender at all. Suppose one is ill and frightened, and a doctor recommends one of two treatments. One could, in principle, go and read the studies oneself, badly, at two in the morning. One does not; one does as she says. And this is neither cowardice nor mere deference. It is sensible, because she is far more likely to get this right, and the whole point of having a doctor is that following her reaches the right answer more reliably than working it out alone would. Hold onto that example, because there is a whole account of authority contained in it.
The clearest statement of it comes from the philosopher Joseph Raz, in his book The Morality of Freedom, published in 1986, and it is worth taking slowly, because it is deeper than it first sounds. Raz asked what would make an authority's command genuinely binding — not merely frightening, not merely backed by force, but something one actually has reason to obey. His answer was this. An authority is legitimate when doing what it says helps one do better at what one already had reason to do — when one ends up tracking the right reasons more reliably by following it than by working them out alone. He called it the service conception: the authority serves the person under it, and earns its standing over them precisely by bringing them to the answers their own reasons were already pointing at, more reliably than they could manage alone. The doctor binds the patient because following her tracks the truth about the patient's body better than the patient's own late-night reading would. The authority is justified by what it delivers.
Notice how much this explains, and how little of it needs the machine. Much of everyday obedience runs on exactly this. Consider stopping at a red light at three in the morning on an empty road. Why do it? Not because anyone is watching. Because a traffic system that everyone follows, even when it seems pointless in the single case, gets all of us where we are going alive; and the reason one already had — not to die and not to kill anyone — is served better by simply obeying the light than by judging each junction on its merits. The authority delivers the outcome one's own reasons wanted; that is why it binds. Multiply that by the thousand small obediences a day is built on — the expert, the rule, the standard, the person who simply knows more — and the size of the idea becomes clear. Much legitimate authority, perhaps most of it, is legitimate in exactly Raz's way: it earns its hold by getting things right on one's behalf.
Now turn this back on the loan decision, the machine we have set aside — because this is where the idea has real force. If an authority earns its standing by bringing people to the right answers more reliably than they could alone, then a system that is more accurate than any human, less biased, better calibrated — one that tracks the real reasons more reliably than the loan officer ever did — is not a sad second-best to be tolerated. On this account it is more legitimate than the human, not less. It earns its authority the way the good doctor earns hers: by being right when it matters, more often, for more people — including the people the old human discretion used to treat badly. The person who built that system and says it should decide is not dismissing anyone's dignity. They are standing on the deepest account of authority we have — the one that explains the doctor, the traffic light, and a thousand things we obey willingly — and they are saying: this is more of the same. Why want less of the thing that gets it right?
Let me be plain before we go anywhere near a reply, because this is the same discipline as last week. That is not a weak argument in disguise. It is, I think, the most serious single account on the table of why any power deserves your obedience, and it has the doctor, the red light, and most of an ordinary day on its side. Do not let me, or anyone, move you past it quickly. If it misses something, that something has to be real — real enough to survive the doctor, real enough to survive the red light. So hold the whole of it, at full strength. And then let us look, carefully, for the place where even this — even the best version of "because it gets it right" — does not reach.
Here is the question that does not fit. Take Raz's account at its strongest and grant it everything — the doctor, the red light, the system that genuinely tracks one's reasons better than one could alone. There is still a question it does not answer, and once you have heard it, it is hard to stop hearing: even if this body would decide well — even if it is the one most likely to get it right — what makes it the one entitled to decide at all?
Listen to the gap between those two. Someone can be the best authority in the world on what you ought to do and still not be the one entitled to do it. Consider a friend who genuinely knows, better than you do, what would be good for you — which job to take, whom to marry, when to stop. Suppose the friend is simply right, more right than you are. Does being right hand them the power to make the decision for you? It does not. You can grant their wisdom completely and still say: it is mine to decide. The philosopher David Estlund, in his book Democratic Authority, published in 2008, put this as sharply as it can be put. The expert on what should be done is not, by that fact, the one with the right to do it. Knowing the good does not confer the right to impose it. There is expertise, and there is authority, and the road from the first to the second is not a straight line — it is not even clear there is a road. And this cuts at Raz's account from the inside. Raz tied authority to getting one's reasons right; Estlund points to all the cases where something gets one's reasons right and still has no standing to impose them. The wise friend tracks your reasons well and may still not decide for you. Getting it right was never, on its own, enough to confer the right to decide.
So if accuracy is not the whole of it — if being the one most likely to get it right does not, by itself, make a power one is bound to obey — then what does? A second tradition answers, on a different axis entirely. Its sharpest modern voice is the philosopher Rainer Forst, in his book The Right to Justification, published in 2012, and his idea is almost austere in its simplicity. Power over a person is legitimate, Forst says, only when it can be justified to that person — when the grounds it acts on are ones they could reasonably accept, not as a favour but as their due. Not "is the decision correct?" but "could the reasons it stands on be ones you could accept as a free equal, rather than ones simply imposed?" He calls it the right to justification, and the word right is doing real work: you are owed an account you could in principle stand under, and a power that cannot give you one — however well it performs — has not earned the standing to bind you. It rules you; it has not yet justified itself to you. And those are different.
Notice how far that has moved from where we began. The first answer measured the decision — was it right? This one measures the relationship — was the person it falls on ever in a position to accept the terms? One can be governed by something that gets every decision correct and never once owed a justification one could stand under — the self-appointed referee who happens to call a clean game. And one can be governed by something that gets decisions wrong and yet stands on terms one genuinely agreed to — the real referee, under the rule-book both sides accepted. On this second account, the second case has a legitimacy the first lacks, whatever the score. Authority is not earned by the quality of the output. It is earned by the standing of the relationship — by the governed having been, in some real sense, party to it.
Now I have to be careful, because the easy move here is to let the second voice win and send you away feeling you have found the answer. You have not, and it would be a cheat to pretend otherwise. Hold both accounts at once and watch them refuse to settle. Raz can push straight back at Forst, and hard. This right to justification, he can say, is a luxury. When you are ill, you do not want a power that justifies itself to you — you want the one that gets it right, and a doctor who stopped to win your free assent to every decision while the clock ran would be no kind of doctor. Much of the authority that keeps us alive never asked our acceptance and never could, and was legitimate anyway, by getting it right. And Forst can push straight back. A power that decides you cannot be trusted to be party to your own affairs, for your own good, is the oldest form of domination there is; and "it gets it right" is what every unaccountable power has always said. The argument goes back and forth, and does not close. This is not a gap that one more fact would fill. Two serious accounts of the one thing — what makes a power's decision binding on you — and they do not converge, any more than the long arguments about how to live converge. We have been here before on this show: some questions stay open not because no one has thought hard enough, but because the thing itself has more than one plausible answer and no neutral ground from which to judge between them.
I am not going to tell you which is right. I am going to make sure you leave holding both, because you are going to need both, and soon. What I will say is the small, hard thing the hour has actually bought you. You came in able to ask only one question of a power that decided about you — did it get me right? You leave able to ask a second that does not reduce to the first: whatever it decided, was its standing over me one I could accept? And those two questions can come apart. A decision can be right on the first and empty on the second. That gap — between correct and one I could stand under — is real, it is old, and you have just watched it open with no machine anywhere near it.
I have been running this on cases I chose, which proves little, because I chose them. So take one I did not pick, from your own life, and run the three questions on it yourself — a tool only proves out where I could not have arranged the ground.
Pick a real decision some authority made about you, or about someone you love, where it stung and you could not simply walk away. Not the machine — keep the machine out of it tonight. A human decision. A panel at work that ruled on a grievance, yours or one made against you. A landlord or a committee that said no. A school that disciplined your child, or turned them away. A council, a board, a tribunal, a manager. Hold one that cost you something, and that you came away from still feeling, somewhere, was not right — and let us find out exactly what that feeling was made of. Run the three questions in order, and do not let them blur together.
First: was it correct? Only that. Set everything else aside and ask whether they got the facts right and the decision right — whether a careful, fair mind with the same information would have landed where they landed. Be honest, even when honesty stings: sometimes the answer is no, they got it wrong; and sometimes the answer you would rather not give is yes — it was correct, and it still felt like a wrong. Note which.
Second, and this is the one people skip: was it theirs to decide? Set aside for a moment whether they were right. Were they the body whose decision actually counted here — the one with the standing to make this call about you at all? Sometimes plainly yes: the panel was the panel, like it or not. But sometimes, on inspection, the honest answer is no — the person who decided had taken a power that was never theirs, or was deciding something outside their remit, a spectator with no whistle acting as though he were the referee. Notice that this is a different complaint from "they got it wrong." You can be angry on this count at someone who happened to decide correctly.
Third, the deep one: granting they did have the authority — was their standing over you one you could accept? Was the power they held come by in a way you could stand under as a fair equal — appointed under rules you were genuinely party to, bound by a process you could in principle have contested — or was it a rigged game: a panel decided before you walked in, a hearing where the verdict was written first and the proceeding was for show, a "because I said so" with nothing behind the I but that they held the cards? Notice that this one can move independently of the other two. Ask it honestly of your own case and see what you actually find — whether a decision that was correct, and the rightful body's to make, still leaves a complaint standing, or whether, once you grant those first two, the complaint runs out and there was nothing underneath it after all. Look honestly. Do not decide in advance how it must come out.
Now hold your three answers up together and look at the shape they make, because the shape is the point. Sometimes all three come out clean — correct, theirs to decide, and come by in a way you could accept — and then there is something humbling to notice: the decision went against you and it may not have wronged you at all. That happens, and a serious person has to be able to say so. Other times the first answer is yes — it was right enough — and a complaint is still standing, attached to one of the other two: that this was never theirs to decide, or that their standing over you was not one you were ever party to. If you find that, take note — because you have found, with no machine anywhere near it, that "was it correct?" and "could I accept it?" can come apart in your own hands. Not that one beats the other. That they are two questions, and not one.
And then — this is the part I will not do for you — ask what that standing complaint actually is, because there are two honest answers and I am not going to choose between them tonight. It could be a real wrong: something you are genuinely owed, that getting the answer right was never going to settle. Or it could be, as the most serious version of the accuracy view would tell you to your face, the ordinary sound a person makes when a verdict goes against them — a wish to be the one who decides, offered as a principle, a kind of control a grown-up sometimes has to give up so that other people are spared worse. Both of those fit the thing you just felt. Which one it is, you do not get to settle tonight, and neither do I. What you walk away with is not the answer. It is the question, sharp and your own, and the tool to keep asking it — and that is the whole of what this hour was for.
Let me close with what tonight changed, because something did — even though, and I want to be straight about this, we did not settle a single thing. We took something apart and built a tool. That is all, and it is enough.
You came in carrying a quiet assumption you had never examined: that a decision about you is legitimate when it is correct, that "is this fair?" simply means "did they get me right?" Tonight that assumption came apart into three pieces. Getting it right is one thing. Being the body entitled to decide is a second. And having come by that entitlement in a way you could stand under is a third. And you saw — on the referee, on the wise friend who is not your boss, on your own case at the end — that they need not travel together. A decision can be wrong and bind you. A power can be right and have no standing to decide. And a standing can be real and still leave you only subject to the power, never a party to it.
And you saw that the question "by what right does any power decide about me?" has at least two serious answers that will not fold into one. One says it earns the right by getting things right for you, the way the doctor does — and on that account a power more accurate, more reliable, and less cruel than what it replaced is more legitimate, not less. The other says no amount of getting it right confers the standing to decide unless the power could justify itself to you, on terms you could accept as a free equal — and a power that cannot, however well it performs, is ruling you, not answering to you. I have not chosen between them, and I will not, because the field has not, and pretending otherwise would be the one dishonesty this show does not allow. Both are in play. Both are serious. You are going to need both.
So here is where we go, and you can probably see the shape of it. For two weeks I have kept the machine aside. Next time I bring it back, and do the hardest thing this series asks of you. I grant it everything. Not "less biased" — last week's fight, already conceded. Everything. Suppose the system is more accurate than any human who ever did the job, audited until not a trace of bias remains, governed and overseen and correct. Suppose it wins the first of our three questions completely and for good. And then we ask the question you now have the tools to ask, the one you could not ask two weeks ago: with the machine having won on correct, is that the end of it, or is it not? I am not going to tell you here. That is next week's work, and it has to be found with you, not announced to you, or it is worth nothing.
So carry the tool. Between now and next time, when some human authority decides about you — a boss, a board, an official, a call you cannot appeal — run the three questions. Was it right. Was it theirs to decide. Was their standing one I could accept. And watch, especially, for the case where the first answer is yes — it was right — and you still cannot say you could accept it. I am not telling you that case proves anything; it might be a real wrong you are owed an answer for, or it might be the wish to be the one who decides, offered as a matter of principle, and which of the two it is, is exactly what we have not settled. But once you have watched that question open in your own life, you will not stop seeing it — and that is enough to be going on with.
Thanks for listening. I'll see you next time.