Why did the same resolution hold you on one night and fail you on the next? Same rule, same person, the same want pulling at you. One night you put the phone down; the next night you did not.
We ended without an answer. Aristotle had taken us further than anyone before him. He granted that you did know — the rule was real, and it was yours — and then he located the failure precisely: not in having the knowledge, but in using it. On the night you failed, the knowledge was present but not active, and a different set of considerations moved your hand. When we pressed him on the last question — why, on that night, the knowledge went inactive — he had no answer. Why it happened at all, he could not say.
This week we turn to a philosopher who approaches that question from a direction no one before him had tried. He will not argue about whether you knew; he grants it. His training is in logic, and he has noticed something that, if it is right, should have stopped this conversation two episodes ago. What we have been calmly describing — that you judged one course better and deliberately took the other, watching yourself do it — may not merely be unfortunate. It may be impossible. Not difficult, and not rare: logically impossible, in the way a four-sided triangle is impossible, a thing that cannot exist because to describe it is to contradict yourself.
If that is right, then everything said in the first three episodes rests on a sentence that cannot be true, Socrates was correct all along, and there is nothing left to discuss. So before this philosopher can offer anything, he has a stranger task first: to show that the failure you committed is even allowed to exist. He states the case for its impossibility as strongly as it can be stated, and then dissolves it with a single distinction. And at the point where you expect him to explain why the failure happens, he stops, points to a gap, and tells you the truth — that he can say precisely where the failure occurs and cannot say why it occurs there. He leaves the cause unexplained. But he locates it exactly.
This is Philosophy for Us — philosophy for everyone, no degree required. This is the fourth of eight episodes about the oldest and most ordinary failure there is: knowing better and doing otherwise.
The philosopher is Donald Davidson, who worked as closely with formal logic as anyone in this story. His question is not "did you know?" but "can what you did happen at all?"
One thing first, as every week. I am not here to grade anyone's reasoning. The phone was in my hand at two in the morning too; I am in this alongside you, not above it. So this is the two of us letting the most careful and formal mind in the series state our shared failure more precisely than we have managed — and then declining to stop at the elegant part.
Start with the philosopher himself.
Davidson is after something different from the two we have met. Socrates was trying to save the soul; Aristotle was trying to map the character. Davidson wants to know whether the description we keep giving is even coherent. He was a twentieth-century American philosopher, and he spent his career on a question that sounds dry until you see how much rides on it: what is it to do something for a reason, and what has to be true of a mind for that to make sense at all? Faced with the two-in-the-morning case, he does not reach for willpower or character. He writes the situation down as a set of claims and asks whether they can all be true together.
And once it is written down, the difficulty appears, and it is sharper than anything Socrates raised.
Here is the difficulty. Follow it as it assembles, because once you see it you will understand why a serious philosopher could spend a career insisting that your last night never happened.
Start with a claim about action that is very hard to deny: when you act on purpose — really act, not twitch, not stumble, but do a thing because you are choosing to — your action reveals what you take to be the thing to do. That is close to what acting for a reason means. If I reach for the coffee, then in reaching I show that I take reaching for the coffee to be the thing to do right now. The action displays a choice: of the options available, this one. You cannot act on purpose and at the same time hold that the thing you are doing is, right now, all things considered, not the thing to do — because then why is your hand doing it? The action itself settles the matter: this one.
Now set last night beside that. At two in the morning your action is to keep scrolling. So, by the claim just stated, some part of you, in doing it, takes scrolling to be the thing to do. So far, consistent. Except — and this is the whole night — you were also judging, the entire time and clearly, that putting the phone down was the better thing to do. That was the one you had sworn off; sleep was the thing to do. That was a real judgment. You would swear to it. You were there.
Here is the position, then: two judgments, at the same moment, about the same situation, both of them yours — sleep is the thing to do, and scrolling is the thing to do — and they cannot both be true, any more than you can judge one cup full and, at the same instant, judge it empty. One mind, one moment, two judgments that directly cancel each other. That is not an unhappy state of affairs. It is a contradiction. And a mind cannot hold a flat contradiction; being unable to is close to what we mean by a mind. So either you did not really judge sleep better, or you did not really choose to scroll. One of the two has to give.
That is the argument, complete. And it is what was supporting Socrates the whole time. We spent an episode arguing against him without quite seeing what held him up. This held him up. Socrates could say "nobody knowingly does the worse" with that steady calm because, on this logic, nobody can. What you are calling weakness of will requires you to judge two contradictory things at once, and that is not a weakness; it is a square circle. It cannot exist. When you first brought your case — certain that you had knowingly done the worse thing while you watched — Davidson's first response is to say, in effect: no, you did not, because what you have described cannot be done. Tell the story again, because the version you keep giving is logically impossible.
And here is why Davidson matters more than the discomfort he has just caused. He does not stop there. Aristotle handled this contradiction by weakening one side of it — the knowledge was not fully there; it was inactive — which, as we saw, conceded Socrates' point after all. Davidson refuses that. He holds that you really did know, fully and while awake; he will not take your knowledge away to save the logic. Which means he has to do something harder. He has to keep both of your judgments entirely real and entirely yours — that sleep is better, and to scroll — and still show that holding both is not a contradiction. He has to find the fault not in you but in the way we wrote the claims down. His view is that we stated the two judgments too crudely, and that they only look like a contradiction because of how they were phrased.
If he can do that — keep both judgments and dissolve the contradiction — he will have done what neither Socrates nor Aristotle could. He will have shown that weakness of will is real, that you did know, that you did choose, and that you contradicted nothing. He will have shown that your apparently impossible night is possible after all. It turns on a single distinction.
Here is the distinction. It is a small one, and it is the difference between an impossible night and a real one. Davidson sets it out in a 1969 essay with a title that states the whole problem: "How Is Weakness of the Will Possible?"
Go back to your two judgments. Putting the phone down is better. And scrolling is the thing to do. A moment ago these cancelled each other, like "this cup is full" and "this cup is empty." Davidson's reply is: look again, because the two judgments are not the same kind of judgment, and once you see that, they no longer collide.
Take the first — putting the phone down is better. Ask what that judgment is relative to. It is a conclusion reached by weighing. You take everything into account: you are tired, you have a commitment in the morning, scrolling will make you feel worse, sleep will make tomorrow manageable, and so on — every reason you can muster — and the conclusion comes out: all of that considered, sleep is better. That is the phrase Davidson relies on: all of that considered. The judgment is conditional. It is a judgment about how things stand in light of everything you have weighed. It is the reasoned conclusion: given the whole body of reasons, the better course is to sleep. Davidson calls this an all-things-considered judgment.
Now the second — to scroll. This is the judgment your action expresses. It is not relative to anything. It does not say "in light of my reasons"; it does not weigh. It is the flat, final judgment that actually moves the body: this one, now. Davidson calls it an all-out judgment — unconditional, with no "in light of," and it is the one that issues directly in action. It is not the reasoned conclusion about what is best. It is the settling of what to do.
And here is where the dissolving happens. There is no contradiction between a reasoned conclusion that, given all the reasons, sleep is better, and a flat final judgment to scroll — because a conditional conclusion and an unconditional one are not the same kind of judgment, and two different kinds of judgment cannot contradict each other. They are not two answers to one question. They are answers to two different questions. The first answers: what do the reasons, on balance, favour? Sleep. The second answers: what am I in fact going to do? Scroll. Consider an analogy. A judge can write thirty pages establishing that the weight of the evidence favours the defendant, and then rule against him. Those two acts do not contradict each other; they are not even the same kind of act. One is a finding about where the evidence points. The other is a decision about what to do. And a court, like a person, can perform the second without honouring the first.
Pause here, because we have been over this ground before and said the opposite. Earlier in the series I said that at two in the morning it did not feel like a decision at all — there was no moment of laying out the reasons and coming down on one side, no verdict reached. I was trying to get the strangeness right, and I mostly did. But Davidson corrects it, and the correction is precise. There were two judgments after all: the reasoned conclusion and the one that acted. What was missing was not a judgment. It was the connection that is supposed to run from the first to the second. The reasoned conclusion said sleep. The judgment that acted said scroll. And nothing carried the first through into the second.
That missing connection has a name, and it is the most important thing Davidson gives us. He says a rational person lives by a standing rule — after Davidson, philosophers call it the principle of continence, though it is plainer than the name suggests. The rule is simply: make your final judgment match your best weighing; do the thing your all-things-considered judgment came down for. It sounds too obvious to state. Of course you should do what, all things considered, you judge best — what else would you do? But state it anyway, because the whole of weakness of will is located exactly where this one rule is broken. The weak-willed person — you, last night — is the person who has the reasoned conclusion, and forms the opposite final judgment, and so breaks the one rule that is supposed to tie the acting judgment to the weighing. You did not fail to weigh. You weighed carefully. You failed to carry your own weighing through into the act. The conclusion and the deed came apart.
And this is more than Aristotle gave us, not less. Aristotle had to make part of your knowledge inactive to make the night make sense; the operative piece, he conceded, was not fully there. Davidson makes nothing inactive. On his account both of your judgments are fully awake and fully yours — the weighing and the acting judgment alike. The knowledge that sleep was better was not reciting in a stupor. It was a fully real, fully conscious judgment that sleep was better, standing right beside a fully real, fully conscious judgment to scroll, with nothing illicit passing between them, no contradiction anywhere, and the phone still in your hand. Davidson has done what was promised: he has shown that your apparently impossible night is possible. Knowing better and doing otherwise is not a square circle. It is a person whose acting judgment did not follow his own reasoned one. It can happen. It happened to you. Davidson has shown you were not talking nonsense.
So you might think we had finally reached safe ground. We have not. The relief of seeing that it is possible, that you are not confused, lasts about a second — and then comes the obvious question, the one Davidson has very carefully not answered. Granted, the acting judgment can defy the reasoned one without contradiction. But why did it? What made you go against your own thirty pages? That is the unanswered question again, the same one Aristotle left us. And what Davidson does with it is the most honest, and the most frustrating, thing anyone does in this whole story.
So that question is still standing. Why did the acting judgment defy the reasoned one? Here is what Davidson does with it. He looks at it and tells the truth, which is that he does not know — and, more than that, that his whole account predicts he will not be able to tell you.
That is rare. Most thinkers, cornered, reach for a flourish. Davidson reaches instead for precision about the size of his own ignorance. He says: I can show you the failure is real, I can show you it breaks no law of logic, I can put my finger on the exact place it happens — in the gap between the all-things-considered weighing and the acting judgment, where the principle of continence is supposed to hold and does not. I can name that place exactly. What I cannot give you is a reason that explains the crossing. And here is the deep point. If there were a reason that explained it, that reason would simply be one more consideration on the scales. It would enter the weighing. And then the weighing would come out differently, you would have judged scrolling better all things considered, and this would not be weakness of will at all — it would just be you changing your mind, which we set aside in the first episode. So the very thing that makes it weakness, rather than a change of mind, is that the crossing happens without a reason that justifies it. By its nature it is the step your reasons do not take for you. To explain it would be to explain it away.
That is a rare kind of honesty in philosophy. Davidson has not left the question unanswered by failing at it. He has shown that this particular question cannot be answered, by design — that the absence of an answer is the very thing we were pointing at the whole time. "Weakness of will" is precisely the name for the step you take that your own best reasoning did not license. Davidson drew the cleanest possible boundary around the problem and then told you, honestly, that there is nothing inside the boundary that reasons can reach. He did not solve it. He located it, proved it was real, and handed it back unexplained.
So take stock, three thinkers in. Socrates: there was no knowledge — you were ignorant. Aristotle: the knowledge was there but inactive — and he could not say why it went inactive. Davidson: both of your judgments were fully awake, you contradicted nothing, and you crossed the gap anyway — and he cannot say why you crossed it, and tells you no one can, because a crossing with a reason behind it would not be this crossing at all. Three accounts, three different places to locate the failure: in the seeing, in the using, in the crossing. And something has happened to the word you started with. "Willpower" — a single faculty, a tank, a muscle. It cannot survive this discussion. You cannot have one quantity that runs "low" when three careful thinkers cannot even agree which part of you is at issue. The thing that failed is coming apart, and we have not reached the strangest account yet.
Before we leave Davidson, one more figure has to come in, because the conversation did not stop in 1969. The philosopher Alfred Mele, in his book Backsliding, published in 2012, picks up where Davidson stopped, at the unexplained gap, and proposes that we can get one real step further into it. Mele agrees the night is real and agrees there is no contradiction. But he thinks "the crossing is simply inexplicable" stops too soon — that we can say something about what crosses the gap, even if the answer is not tidy. What crosses the gap, he says, is motivational strength. You have your all-things-considered judgment — sleep — and you have a desire — to scroll — and at two in the morning that desire is, as a matter of sheer force, pulling harder than your evaluation is. The weighing favoured sleep, but the wanting to scroll was outpulling the wanting to act on your own conclusion.
And Mele is careful about that word "strength"; this is the part to take seriously before we test it. He says motivational strength has sources you can actually name: how near and vivid the thing is, how much it seizes your attention, how tired or stressed or wound up you are — all of which raise or lower a desire's pull without changing one word of your considered judgment. Against that pull he sets self-control — not a mood but a genuine capacity, a set of things a person actually does: looking away, putting the phone in another room, refusing to picture the thing, talking oneself down, waiting out the peak until it passes. On Mele's account, two in the morning is a real contest between a desire whose strength has been raised by the hour and the tiredness, and a capacity for self-control that, tonight, was not used in time or was not strong enough to hold. That is a serious account. It says where to look, and it even says what you might do.
But we do not stop at the part that works, and there is a fault in it — the same fault we found in Aristotle. We asked Aristotle: "the knowledge was inactive" — is that an explanation, or a more elaborate way of saying "I was overcome"? Ask Mele the same question about the bottom layer of his account. The vividness, the tiredness, the attention — those raise the pull, granted. But why, at the final instant, did the raised pull move the body against a judgment you genuinely held? Because it was the stronger motivation. And how do we know it was stronger? Because it won. And why did it win? Because it was stronger. He has named real forces and real counter-moves, and pushed the explanation a long way further — but at the very centre, where the action is actually settled, "stronger" only ever means "the one that won." The most careful version of Mele bottoms out where Aristotle did, in newer and better-furnished terms.
And there is a deeper problem in what Mele has done, and it is the one that carries us into the rest of the series. To make his account work, he has separated your wanting from your judging. He is saying that the thing which moves you can run at a different strength from the thing you judge best — that your motivation does not automatically track your evaluation. That may well be true. But the moment you say it, you have stopped describing one unified person who weighs and acts, and started describing a person with at least two sources of motive force that can pull against each other. Keep that in mind — someone next episode builds a whole account on it.
First, though, the method is yours to use, and it works on more than the phone. Take a resolution you set in advance and broke — not a craving, a plan. The budget you set on Sunday and overspent on Wednesday. The alarm you set for the gym and turned off in the dark. Now run Davidson's question on it, honestly. Find the two judgments. There was a reasoned one — all things considered, stick to the budget, money is tight this month, this is the thing to do — and you really made it, on Sunday, in daylight, with all the reasons in front of you. And there was the acting judgment, on Wednesday: buy it. Now be honest: can you explain the crossing from the first to the second? Not justify it — explain it. Can you find the reason that carried you from your careful Sunday conclusion to the Wednesday purchase? Try. Because if you find one — money was not really that tight, it had been a hard week, you had earned it — what you have actually done is re-run the weighing and reached a different conclusion. Which means one of two things: either you genuinely changed your mind on Wednesday, in which case this was never weakness — or that reason is a story you told afterwards, to present a crossing that had no reason as though it had one. Keep track of which it was. We will need it later.
Because there is one more figure to hear tonight, and he says something that makes all three thinkers so far look as though they were answering the wrong question. Every one of them — Socrates, Aristotle, Davidson, and Mele too — took it for granted that what failed at two in the morning was a judgment: that the whole matter is about knowing the better and not doing it. What if it is not? What if, at two in the morning, you were not judging anything at all?
What if, at two in the morning, you were not weighing anything at all? This is Richard Holton, a philosopher working in our own time, in his book Willing, Wanting, Waiting, published in 2009. He thinks the three men before him made the same mistake before they said a word: they mis-described what actually happened. And once he re-describes it, you may recognise something that has not quite fit since the first episode.
Here is what he does. Go back — really back — to your actual two in the morning, and ask a plain question no one has asked yet: at two in the morning, were you weighing anything? Were you laying the reasons out — sleep on one side, scrolling on the other, all things considered, reaching a conclusion? Because that is what every account so far has assumed. Socrates, Aristotle, Davidson, Mele — every one of them pictures you at two in the morning as someone actively deliberating, the reasons coming in, a judgment being formed, with the only dispute being what became of that judgment. But Holton says: that is not what two in the morning is like, and you know it is not. At two in the morning there is no deliberation. You are not weighing. You are not reaching an all-things-considered anything. The scrolling simply starts; the thumb is already moving. There was no deliberation at all.
So when did the weighing happen? Earlier. At dinner. That is when you deliberated. That is when you took it all into account — tired, an early morning, scrolling makes it worse — and reached your conclusion, and then, crucially, did something with it: you made a resolution. No screen in bed. And here is what Holton points to in that resolution — the thing that sets it apart from a mere judgment. You did not make it to record what you had decided. You made it to stop yourself from re-deciding. That is the whole point of a resolution. It is a line you draw while you can think clearly, precisely because you know that at two in the morning you will be in no condition to deliberate fairly — tired, worn down, the scrolling looking large and sleep looking small. The resolution is there to take the question off the table, so that the two-in-the-morning version of you does not get to re-open it. You settled the matter at dinner on purpose, so it could not be re-opened under worse conditions.
So what actually failed at two in the morning, on Holton's account? Not a judgment — there was no judgment; nothing was being deliberated. What failed is that you re-opened the question you had deliberately settled and set aside. You revised the resolution. And weakness of will, Holton says, is exactly that and nothing grander: not acting against your better judgment, but abandoning a resolution too readily — giving way to the very pressure the resolution was built to withstand, for reasons you would have dismissed at once at dinner.
This fits the experience more closely than anything yet — more closely, I will grant, than the four accounts before it, though "closer to how it feels" is not the same as "shown to be correct," and in a moment we will see that it has not been shown correct. When I go back into my own two in the morning, honestly, Holton's description fits the night better than the others did. I did not judge the scrolling better. I did not re-examine the case and reach a contrary conclusion. There was no conclusion, because there was no deliberation. I simply let it go. The line I had drawn at dinner, I quietly let lapse, as if it had never mattered, as if I had never meant it. That is the actual texture of the thing — not a reasoned conclusion overturned, but a resolution left unenforced, a decision I had already made and then declined to keep.
And see what that does to the thing that failed — this is why Holton matters, not just that he fits the experience. Every account before him located the failure somewhere in the intellect — in your knowing, your seeing, your weighing, your judging. Holton moves it out of the intellect entirely. The failure is not in the deliberation at all. It is in the keeping — in whether the will can hold a resolution shut against pressure. That is a different capacity, if it is a capacity at all. It is not about seeing the better; it is about staying behind a line you drew when you could see clearly. The whole question has shifted sideways, off "did you know" and onto "did you hold."
But we do not stop at the part that favours him, and Holton has a problem, a real one, and he knows it. If weakness is revising a resolution — wait. Sometimes revising a resolution is the wisest thing you can do. You resolve to run every morning, then you catch the flu and skip the run: that is not weakness, that is sense. You resolve never to contact an ex, then there is a genuine emergency, and contacting them is the right thing. Resolutions should be revisable; a person who never reconsidered a single resolution, whatever the world threw at them, would not be strong but fanatical. So Holton cannot say that all revision is weak. He has to draw a line between reasonable revision — the flu, the emergency — and weak revision — two in the morning, the scrolling. And where is that line? His answer is roughly this: revision is weak when you revise for reasons you would not have accepted when you made the resolution — when nothing has genuinely changed except that you are now in the grip of the very thing you were trying to resist. That is honest, and it points in the right direction. But the line is carrying an enormous load, and it blurs on close inspection. "Reasons you would not have accepted at dinner" — but at two in the morning the reasons feel entirely different from how they felt at dinner; that is the whole problem; the scrolling has reasons now that it did not have then. Who is to decide, and when, whether tonight's reasons are genuine new information or merely the pressure speaking? Holton has relocated the problem precisely — out of the intellect and into the will's holding — but he has not dissolved it. He has handed us a new question exactly as hard as the old one: when is letting go wisdom, and when is it giving in? He is the latest to carry the problem into a new place — Aristotle into the using, Davidson into the crossing, Mele into the pull — and to find it just as heavy there.
And next time it stops being about your knowing or your holding at all, because someone raises the question that has been hiding under every one of these. Every account tonight assumed there is a single you at two in the morning — one person who knows, or holds, or gives in. But you said it yourself, the very first night: it felt like two of you. The one at dinner who meant it, and the one at two in the morning who is still going. Next episode a philosopher takes that seriously, and stops asking did you know or did you hold, and asks instead: when your will divides like that, which of them is actually you — the one who drew the line, or the one who crossed it?
Thanks for listening. I'll see you next time.