Consider again the two figures this series began with. There is the person who, at dinner, decides not to do the thing tonight, and means it. And there is the same person, at two in the morning, still doing it. For four episodes we have assumed, without examining the assumption, that these are one person — one self, who knew or failed to know, who held a line or let it go. We have asked what happened to that self. We have not asked whether there was a single self there at all.
Tonight's philosopher refuses the assumption. He asks the question the series has been circling, the one promised at the end of last week: when the will divides in that way, which side is the person — the one who drew the line at dinner, or the one who crossed it at two? And he warns, at once, against the quick answer. To say "the one who drew the line" is the natural reply, and he means to show that it is not available for nothing.
His name is Harry Frankfurt. He does not ask whether you knew what you were doing; he grants that, as Davidson did. He does not ask whether you were holding a resolution; that was last week's question, Holton's. He asks something different. At any moment a person wants many things at once. Part of you wants to keep scrolling; part wants to sleep; part wants to be someone who wants to sleep. Frankfurt's question is which of all that wanting is the person, and which is merely something passing through.
He offers a way to answer it. A person can want things, as any animal can — a dog wants the food. But a person can also want something about their own wanting: you can want not to want the scroll. That second kind of wanting, Frankfurt argues, is where the person is located. He then describes a creature that has none of it — one that simply wants, and acts, and never asks which of its wantings to be moved by — and the uncomfortable suggestion is that at two in the morning you resembled that creature more than you would like.
Then a second philosopher, present in the argument from the start, objects that the whole method is a trick: it never explains why the higher wanting should count as the real person rather than as one more want among the rest. So there is genuine movement tonight — but onto contested ground, which is the only honest place for it to be.
This is Philosophy for Us — philosophy for everyone, no degree required. This is the fifth of eight episodes on the oldest and most ordinary failure there is: knowing better, and doing otherwise.
Last week, Davidson, Mele and Holton took the word "willpower" apart and could not agree on what to put in its place. Davidson showed that the episode at two in the morning was a real event and not a contradiction, then located a step at the centre of it that he held nobody can explain. Holton suggested you were not judging at all — that you made your decision at dinner and simply let the resolution lapse. Five accounts of the same night, and none of them explained the mechanism. Tonight, a philosopher who stops asking what failed in you and asks instead which part of you was you — and the division he examines is mine as much as anyone's.
Frankfurt begins from a question no earlier account had asked.
Harry Frankfurt was an American philosopher of the twentieth century. In 1971 he published a short paper with a title that appears to have nothing to do with a phone at midnight: "Freedom of the Will and the Concept of a Person." It begins by noticing something ordinary enough to be overlooked.
A dog wants the food on the counter. Suppose it also half-wants to obey, because it has been trained and something in it flinches at a reprimand. The dog then has two desires pulling in opposite directions, and the stronger one prevails, and the food is either taken or left. But the dog is not weak-willed, and it is not struggling with its own character at two in the morning. It has competing desires, and that is all it has. So competing desires, by themselves, cannot be what makes the human case the difficulty it is. Animals have competing desires too. A person has something further, and Frankfurt wants to identify it precisely.
The further thing is this. A person does not only want things; a person can step back from their own wanting and take a view of it. You do not only want the scroll — you can want not to be someone who wants it. You do not only crave the drink — part of you can wish the craving were not there at all. This second wanting is not directed at the world, at the phone or the glass. It is directed at your own first wanting. It is a desire about a desire. Frankfurt calls it a second-order desire, and, as far as we can tell, the dog never forms one. The dog wants the food; it never wishes it were a dog that did not.
The distinction carries the whole episode, so I will state it plainly once. A first-order desire is wanting some thing. A second-order desire is having a view about what you want — wanting this desire to prevail and that one to lose. Frankfurt's central claim is that the person is located at the second level: not in the wanting itself, but in the position taken towards it.
He then narrows it. Among your second-order wishes, one kind matters most. It is not the idle wish that you wanted different things. It is wanting a particular desire to be the one that actually moves you — the one that prevails, that moves your hand, that you act on. Frankfurt has a name for the desire that actually moves you to act. He calls it your will.
The word needs care, because the series has spent four episodes on a word that sounds identical and means something else. "Willpower" — the thing you arrived blaming — was treated as a faculty, a finite resource, a quantity you have too little of. Frankfurt's "will" is not a faculty at all. It is simply whichever desire is actually moving you at a given moment. So when Frankfurt asks whether your will is free, he is not asking whether your reserve of willpower is full. He is asking a different and better question: is the desire now moving you the desire you wanted to be moved by? Did you endorse the thing that is driving you, or did it simply take over?
That distinction — between a desire that merely moves you and a desire you endorse — runs through the rest of the episode. Frankfurt's own term for endorsing a desire, for making it your own as the one you want to be moved by, is identifying with it. A desire you identify with is yours in a way the others are not, even when the others are stronger.
Now set the previous night beside this. At dinner you identified with a desire: you wanted the desire to sleep to be the one that moved you; that was the one you endorsed. It was a real act, and you meant it. At two in the morning the desire that actually moved your hand was the other one — the desire to scroll — the one you had refused to endorse. So Frankfurt's account of the failure is unlike the four before it. You did not fail to know. You did not fail to weigh the options. You did not, exactly, fail to hold a resolution. What failed is that the desire you endorsed lost to the desire you did not. Your will — the desire actually moving you — was one you had disowned. You were moved by something you had refused to back.
This changes where the failure lives. For four episodes, what failed was somewhere in the head — in knowing, in seeing, in weighing, in holding. Frankfurt takes it out of the head and places it in the structure of a person's wanting. What broke was not an intellect that misfired. It was a self divided — split between the desire endorsed and the desire that won. A self that can divide in that way is not a resource that runs low. It is a structure. It has parts. It has a place where it can come apart.
That is a large step, and not yet a sufficient one, as Frankfurt knew. There is an easy reply available: very well, the desire I endorsed lost, but at least I endorsed the right one; the real me was on the side of sleep, and was simply overpowered. That is the answer Frankfurt warned, at the outset, against giving too quickly. For his next move is to ask whether you endorsed anything at all, or only told yourself you had.
Here is the figure Frankfurt draws, and it is the one that stays with people.
Imagine a creature with desires — many of them, and strong — that never once takes the step back. It never forms a second-order desire. It never asks which of its desires it wants to be moved by. It simply wants, and acts on whichever want is strongest at the moment. It has no view about its own will. It never endorses one desire against another, because it never takes a position at all. Frankfurt calls this creature a wanton.
Do not picture something stupid; that is the trap in the word. A wanton can be highly intelligent. It can plan, delay, and calculate the most efficient route to what it wants. What makes it a wanton is not dullness. It is that it never turns the question on itself — never asks whether to be the one who wants this. It has capacities, and takes no position on how it uses them.
Now the harder question. At two in the morning, were you a person or a wanton? The quick answer is: a person, plainly — I was present, I had the second-order wish, I was thinking "I don't want this, put it down." Frankfurt grants the thought. It was there. But he asks whether it was doing any work, or whether it was merely commentary — a running description of the event that changed nothing, while the strongest desire carried you where it was always going to. A wanton can produce that commentary too. Having the thought "I don't want to want this" is easy. The question is whether you endorsed it with any weight — whether it was you — or whether it was one more thing occurring in you while the desire you had disowned did the moving.
Frankfurt makes the point exact with a case; I give it as his, with one detail of my own added. Consider two people addicted to the same drug, with the identical physical craving. Both of them use it. Observe only their behaviour and you cannot tell them apart. But their inner situations, Frankfurt says, are opposite. The first has come to accept the craving. He wants to be moved by it; offered a cure that removed the desire, he would decline it; the craving is a desire he endorses without reservation. Frankfurt calls him the willing addict. The second is in conflict with himself. He wants, with everything in him, not to want the drug, and uses it anyway, hating it as he does. Frankfurt calls him the unwilling addict.
Same craving, same act. And Frankfurt says these are not the same case, and the difference is precisely what we care about. The willing addict acts as himself: the craving is his, he endorses it, and his will is free in the only sense that matters, addicted though he is. The unwilling addict is moved by a force he disowns; it is inside him, but it is not him; he is a person watching a desire he refused to endorse move his body regardless. The difference between them is not in the drug, and not in the strength of the craving. It lies entirely in which desire each one endorses.
And here is the conclusion the series has been moving towards. Consider what has become of "willpower." It has gone. It was never a finite resource, never a quantity you have more or less of. What is at stake at two in the morning is not an amount of anything. It is whether the desire that actually moves you is the desire you endorse. Which means the self is not a single quantity of anything that runs high or low. It is a structure with levels, and a division between them, and the failure occurs at the division. The thing you arrived blaming does not exist. No faculty came up short. A self divided, and the part that did the moving was the part you had disowned.
I will mark one further point and then leave it alone, because it belongs two episodes ahead. A structure with a division in it is a different kind of thing from a faculty you can only strengthen. A division is something that can be acted on from outside — widened, or narrowed. Keep hold of that; we will need it two episodes on.
Because tonight there is something to practise, and it is yours to do, not mine to demonstrate. Take the smallest version of it — not the phone, but the shopping cart. It is late; you open the app; the item is sitting in the basket; and before you have done anything, a sentence arrives, complete and unbidden: I'm just looking. I'm not going to buy anything. Consider what that sentence actually is. It is not the desire. The desire is simple, and you know exactly what it is: you want the thing. "I'm just looking" is the second-order story. It is what you say so that you can stand at the edge of the desire while feeling in control of it, as your hand moves towards the purchase. Then you buy it — and the story has already excused the disowning in advance: I wasn't even going to; it just happened. What has been arranged is this: you kept the desire and were spared being the one who wanted it. You had the thing and stayed innocent of wanting it.
That is not a resource running out. It is a self deceiving itself, right at the division. It has a name, and the series will spend real time on it, because this is where you learn to catch it: the small untruth you tell yourself about your own motives — not afterwards, to excuse the act, but beforehand, so that the desire never has to be examined at all. So what to practise is this: not the buying, but the saying. The next time the pre-emptive story appears — just this once, just checking, I've earned it, I'm just looking — recognise it for what it is. Not a fact about your wanting, but the sentence you produce so that the wanting is never owned.
We appear to be getting somewhere. What breaks is a divided self; the failure is a desire you did not endorse moving you regardless; and the surest way to feel like a person while behaving like a wanton is to keep the story and disown the desire. But Frankfurt's account remains abstract — a schematic description, not the experience itself. So before anyone presses him, it is worth going back sixteen hundred years, to a man who wrote the two-in-the-morning down from the inside, before any of these terms existed.
His name is Augustine. Around the year 397 he wrote a book called the Confessions, and in its eighth book he set down the most exact description of the two-in-the-morning that anyone has produced — thirteen centuries before there was a term like "second-order desire" to tidy it.
His situation was this. Before he is the famous saint, he is a man who cannot bring himself to do the thing he has completely and finally decided to do. He wants to change his life entirely; he is convinced; the arguments are all in; the matter is settled. And still he cannot take the step. He stands at the edge of it, fully persuaded, and does not move. And he notices something that arrests him — and it is the same thing that arrests you. The mind, he observes, gives the body an order and is obeyed at once: I decide to raise my hand, and my hand rises, with no gap between. But then the mind gives an order to itself, and is not obeyed. In his words: the mind commands itself, and is resisted.
Stay with how strange that is, for Augustine is the first to examine it closely. The hand is, in a sense, external to you, and yet it obeys at once. But your own will — which is supposed to be nearest to you, not separated from you by so much as a nerve — that you command, and it declines. You tell your hand to move, and it moves. You tell yourself to want what you have decided to want — to put the phone down and mean it all the way through — and the order goes out into the one place where you would expect to be fully in control, and nothing happens. The place where your command should be absolute is the one place it fails.
Augustine's explanation is the felt version of the division Frankfurt described in the abstract. It is not, he says, that he half-gives the command. It is that there are two wills in him, neither of them whole, pulling against each other. One wills the new life; one wills the old. And then he says what is, I think, the hardest and most honest thing in the episode. He refuses to call either will a stranger. It was I who willed it, he says, and I who was unwilling. It was I. Both of them. The will reaching for the new life and the will clinging to the old are both his, each carrying his whole name. He is not a good man afflicted by a bad element separate from himself. He is not an observer standing safely outside the conflict. He is the conflict. That is why it is painful: one cannot suffer a struggle one is merely observing.
Now note what he does next, because it is the load-bearing move, and the reason Augustine belongs here and not only Frankfurt. He had an easy and respectable way out, and he rejected it. In his day there was a popular teaching, that of the Manichees, which said exactly the consoling thing. When you are divided in this way, it held, it is because there are two natures in you: a good soul, which is the real you, and a separate, alien, evil element that has entered you and is not truly you. On that view your divided will is simply the good, true self held prisoner by a foreign bad thing. You can hear how comforting that is. The desire that drags you towards the worse course is not you; it is an intruder, a parasite; you are the good one, the captive.
Augustine looks directly at that consolation and refuses it. Both wills are mine. When I willed and when I un-willed, it was one and the same I — not I and an invader, but I and I. The desire pulling him towards the old life is not a demon that has broken in. It is his. He does not get to call it a chain; he has to call it by his own name.
That is the same refusal you practised earlier, with the cart. "I'm just looking" is a small version of the Manichean move: it tries to make the desire an alien thing, something that happened to you rather than something you are, so that when you buy the item, the real you was merely looking. Augustine, sixteen hundred years early, will not allow himself the move. Both wills are his. Which means the division in you is not an escape from responsibility. It shows you what you actually are. You cannot be the good will and disown the bad one; both answer to your name.
I will name one thing that is coming, so that it does not arrive without warning, because the series reaches it in two episodes. The oldest consolation there is — older than the phone, the gym, the diet — is that the bad wanting is not really me. Augustine refused it when the alien on offer was a demon. But in two episodes the same consolation returns in a modern form: the bad wanting is not really me — it is the environment built to defeat me; it is the brain chemistry I was handed. And here is the genuinely hard part, which I will not cheapen by answering tonight: that version may be partly true in a way the demon never was. The environment is real. The chemistry is real. So we will have to decide whether Augustine's refusal still holds when the alien is not superstition but fact. That is the work of the second half of the series. Tonight the point is only named, and then set aside.
We seem to have arrived at something solid. Frankfurt supplied the structure — a self with levels and a division. Augustine supplied the experience — two wills, both one's own, and no alien to blame. What failed is a divided self, and the division is fully and honestly yours. You might expect to rest there. But there was, from the start, a second philosopher present, and he holds that this entire picture is an evasion. He has one question. And if it cannot be answered — I will say now that I cannot answer it — it undermines everything just built. That is next.
His name is Gary Watson. In 1975 he published a paper called "Free Agency," and it is, in large part, a careful refutation of what has just been argued, conducted with great respect.
Watson reads Frankfurt and says: this is admirable, and it does not work, and here is the exact point at which it fails. Frankfurt tells you that the real you is the higher desire — the second-order desire, the one you endorse. Grant Frankfurt that. Watson asks the simplest possible question: why? Why is the second-order desire any more you than the desire beneath it? What is special about a desire merely because it is aimed at another desire, other than that it sits one level up?
Look closely. You want the scroll — call that the first level. You want not to want the scroll — the level above. Frankfurt says the upper level is the real you, the one that counts. But, Watson replies, that second-order desire is just another desire. It is a desire like any other. So what prevents me from asking the same question of it? Suppose I want not to want to want the scroll — a third level. Which is the real me now? The problem is plain: the levels can be stacked without end, and "the real you" keeps moving one level higher, and Frankfurt never says which level to stop at, or why any level is the true one. Philosophers call this a regress. Its point is simple: placing a desire at a higher level does not make it more yours. It only makes it higher. Frankfurt undertook to tell us which desire is the person, and at the decisive moment he pointed upward and could not say why.
Watson has his own answer, and it deserves full statement, because he is not present only to object. Frankfurt went wrong, he argues, because he was working with the wrong distinction from the start. What divides you at two in the morning is not desire-about-desire, not levels of wanting stacked up. It is something different in kind from desire altogether: the difference between what you value and what you merely want. What you value is what you actually judge to be good — the course you would defend in daylight, with reasons, before people you respect. What you merely want is what pulls at you, whether or not you would ever defend it. The two-in-the-morning division, Watson says, is not first-level desire against second-level desire. It is your values on one side — your genuine judgement of what is worth doing — and a bare pull on the other. What counts as you is not the desire that sits highest, but the part of you that holds what is good.
You may well find that more convincing than Frankfurt; it can seem so. But it should not comfort you, because of where it leaves us. What you judge good against what merely pulls at you — we have met that distinction before. It is Socrates, in modern dress: knowledge against appetite, the very thing the first three episodes found was not enough to explain the night. Watson may have caught Frankfurt in a regress only to return us to the oldest argument in the series.
And there is a case that catches both men, one we already hold: the willing addict, who has made his peace with the drug. Ask Frankfurt whether he is free. Frankfurt must say yes — the man endorses the craving without reservation, it is the desire he backs, his will is his own. And something in us objects, because a man owned by a drug does not seem free merely because he has stopped resisting it. Now ask Watson: does the willing addict value the using, judge it genuinely good — or has the craving reached so deep that it has shaped his judgement itself, so that he now takes the using to be genuinely good? Here Watson cannot give a clean answer, and neither can you. Both men who tried to locate the real self inside the divided person are stopped by the same human being. Neither can say cleanly where the man ends and his craving begins.
So consider where this leaves us; let me be precise about what you now have and what you do not.
You arrived, four episodes ago, with a single easy word — willpower — and a tidy account: a resource I am short of, and that is on me. Tonight something has been put in that word's place. Frankfurt: what failed is not a faculty but a structure — a self with levels, which failed at the division, where the desire you endorsed lost to the desire you did not. Augustine: and both of those desires are yours; the division is real, and you may not disown the side that cost you the night. And Watson: do not believe anyone — Frankfurt, or me, or the host — who tells you which side is the real you. The higher level, the part that values — the moment it is named, the mystery has merely been renamed and called an answer.
So what do you actually hold tonight that you did not an hour ago? Not an answer. A relocation. "What failed?" is no longer "a resource ran low." It is: "I divided, and the part I would have sworn was me lost control — and I cannot even prove which part was me." That is a worse position than the one you came in with. It is also, for the first time, a true one. The question has stopped being what is wrong with my willpower and become which of these desires was me.
Before any of that, there is what tonight has already done, and it is the sharpest and least welcome part, and I will not soften it. You watched yourself, at the cart, reach for the small story — I'm just looking — that keeps the desire and disowns the one who wants it. You cannot now return to not knowing that about yourself. It is not a gift; it is closer to an irritation you cannot remove. The next time you reach for the story, part of you will recognise what it is. That will not stop you. It will only cost you the innocence. That is what tonight leaves you: a worse question and a lost excuse. And the question beneath it is the one nobody in that room — not Frankfurt, not Watson, not I — can answer on your behalf. Which one was you? And what did you have to tell yourself, before you ever picked up the phone, so that you would never quite have to find out?
And here is where the series goes next. We have established tonight that the self has a division — a place where it comes apart. Next week we turn to who has been studying that division, and profiting from it. The environment you are in at two in the morning — the feed, the endless scroll, the thing built to be this difficult to leave — was not a natural growth. It was designed, by people who know where the division is and make their living widening it. So next week the question turns outward, from you to the environment: if the two-in-the-morning was built to defeat you, how much of what failed there was ever yours? That question will, at first, sound like relief — which is exactly where I would be careful.
Thanks for listening. I'll see you next time.