Here is a sentence you have said, or had said to you, and probably never stopped to examine.
I believe in you.
A parent says it to a child on the morning of an exam. A coach says it to a team before they go out. You might say it to someone you love who is about to do something difficult. I believe in you. And no one who hears it takes it as a report on the evidence. No one thinks: she has reviewed my abilities and concluded that I will probably pass. If your father said "I believe in you" and you answered "on what evidence, exactly?", you would have misunderstood him completely. He was not making a prediction. He was telling you where he stands: I am with you, I am for you, whatever happens I have thrown in my lot with you. It is closer to a promise than to a forecast.
This is the point I kept returning to this week, after five episodes of using the word "believe" as though its meaning were obvious. We do not only use it in the juror's sense — the sense of weighing the evidence and returning a verdict. We use it in that other sense constantly. I believe in you. I believe in this marriage. I believe in this country. There is even the make-believe of a child's game. The wedding vow is not a statement of probabilities about the next forty years; it is a pledge. That meaning is alive; it is in ordinary speech every week. We simply stop hearing it the moment the subject turns to God — at which point "believe" narrows to a single meaning: a verdict, returned on the evidence, the way a juror announces guilty or not guilty.
So I want to take up the question I left open at the end of last week, and this time answer it. Which of those is what "believe" really means — the juror, or the vow? And, just as important: was it always the juror, or did the meaning change?
This is Philosophy for Us — philosophy for everyone, no degree required.
This is Is There Anyone There?, episode six of eight. Last week we examined the last position that still felt secure — the idea that wanting evidence is simply being reasonable — and found that it has a name, an author, and a serious weakness. It is a position called classical foundationalism; it cannot pass its own test; and when it is turned on the question of God as though it were the neutral default, it quietly assumes the very thing in dispute. It is not a neutral starting point. It is a standpoint, and it is one you hold without ever having read it. I ended by handing back a word you were suddenly less sure of — belief — and one question: this picture of believing as a verdict returned once the evidence is in, where did it come from? Is that simply what the word has always meant? I said I would not answer it then. Tonight I will. The word has a history and an older meaning, and once you have heard it, the question we have been asking for five weeks stops looking like the only question there ever was.
Begin with the natural objection. Those are two different uses of one word, someone will say, and they are not equal. "I believe the keys are in the drawer" is the real, literal use: believing that something is the case. "I believe in you" is a warmer, looser relative of it — a figure of speech, the same word softened to mean something like I am rooting for you. On this view the evidence-weighing sense is the original, and the vow is a later, sentimental offshoot of it.
That is the natural assumption, and I held it for years. Notice it is an assumption, and not something anyone has checked — because we are about to find that it has the order of descent exactly backwards.
Here is why it cannot be settled by reflection alone. When a word has two meanings and you want to know which is the original — which came first, and which grew out of it later — no amount of hard thinking will tell you. You have to look. You have to trace the word back through the centuries of people using it and see which meaning was there first and which arrived late. At that point it is not a philosophical question but a historical one. It has an answer; the answer is in the record; and most of us have simply never read it. We received the word carrying only its most recent meaning, with no sense of how old the earlier ones are.
Someone did read the record. His name was Wilfred Cantwell Smith, a historian of religion and one of the most distinguished of the last century, who spent a career on exactly this: not arguing for God, and not against, but tracing the history of religious words and watching what they meant to the people who used them, century by century, before the meaning we take for granted was fixed. In two books, Faith and Belief and Belief and History, published in the late nineteen-seventies, he set out what he found. And be clear about what he was doing. Smith was not an apologist smuggling God in; he was a historian describing what a word did, and when it changed. You can be the most committed atheist in the room and still want to know the true history of the word you are using — indeed you should want it most of all, because we have spent five weeks uncovering rules we were relying on without ever inspecting them, and "believe" is the largest uninspected thing left. We have been running an entire trial — the question of God, weighed on the evidence — on a word no one at the trial ever paused to define.
So consider what Smith found. It does not leave the word where it started.
Here is what the record shows.
Our word "believe" comes down through the old Germanic languages, and at its root — the reconstructed form scholars write as galaubjan — it does not mean "to judge probable." It means to hold dear, to prize, to set one's heart on. It is the same root that gives us belove, as in beloved, and it stands directly behind the ordinary word love. You can still see the two side by side in German, where the word for "believe" is glauben and the word for "love" is lieben — close enough to rhyme, because they came from the same place. To believe something, at the root, was to belove it: to hold it dear, to be loyal to it.
The other great source of the word is Latin, the one the church carried. The creed — the thing a Christian stood and said — is the Credo, "I believe." And credo comes from credere, which historians gloss as "to set the heart on, to entrust" — a word built out of the old root for heart. When a man stood up in the year four hundred and said Credo — "I believe in God the Father almighty" — he was not stating the conclusion of an investigation. He was not saying he had surveyed the evidence for theism and found it more likely than not. He was doing something closer to what a person does at a wedding, or what a soldier did in swearing loyalty. He was pledging himself. Credo was a vow of allegiance: I give you my heart, I throw in my lot with this. The creed was not a list of claims he judged probably true; it was a list of things he bound himself to.
This is the central point. For most of the life of this word, to believe was not to estimate but to commit. It named a relationship and a stance of the will — trust extended, loyalty pledged — and it added something to bare knowledge: the difference between knowing a fact, the way you know the boiling point of water, and knowing a person, the way you know someone you have thrown your life in with. Notice, too, what counts as going wrong in that older sense. If belief is a verdict on the evidence, then to believe wrongly is to miscalculate — to get the probabilities off, to reason badly. But if belief is allegiance, if it is credere, then to believe wrongly is something else entirely: it is betrayal, faithlessness, the breaking of a vow. Those are not two grades of the same error. They are two different accounts of what believing even is.
And here the point returns to ordinary speech. Say it again, silently, to whoever it is true of — I believe in you. When you mean it, are you offering that person a probability estimate? Are you saying you have run the numbers on their character and they have cleared the bar? Of course not; they would be right to be insulted if they heard it that way. You are doing the older thing. You are setting your heart on them — saying I am with you, I am loyal to you. The credere meaning did not die and leave only a trace behind. It is alive. It is the meaning we reach for the moment a belief actually matters to us — a person, a marriage, a cause worth suffering for — and we switch to the juror's meaning, the cold verdict on the evidence, only for things we hold at a distance. Which, given the subject of these episodes, is already worth pausing on.
We have met this shape before. A few series ago we spent an episode on trust — on the wish to be a self-checking auditor who verifies everything before extending any confidence, and on why that figure is a fantasy, since nearly everything we know we took on someone's word, on trust, without ever checking it. What we recovered there has the same structure as what we are recovering here. Trust is not a verdict reached after the evidence is in; it is a stance we take toward people that lets the evidence begin to exist at all. Credere is that stance, extended to the deepest things. Belief, in the old sense, was a kind of trust — which is exactly what a trial cannot register, because a trial trusts no one. It demands proof. That is its function.
So that is the older settlement: belief as the heart committed. Now — what became of it?
How did belief, in the sense of the heart committed, become the juror's verdict? Not overnight, and not because anyone proved the older meaning wrong. It happened the way such things usually happen — through a slow change in what educated people took the word to be for.
You can watch part of the change in John Locke, at the end of the seventeenth century. Locke is as responsible as anyone for the picture we have inherited. In his Essay Concerning Human Understanding, published in 1689, he defines faith, carefully and influentially, as assent to a proposition. Faith, he says, is assent to something "upon the credit of the proposer, as coming from God, in some extraordinary way of communication." Notice what has already happened in that sentence. Belief has become assent — a yes said to a proposition, a claim that could be true or false. And the only remaining question is whether one is entitled to say yes: is the proposer credible, is this genuinely revelation, does the warrant hold? Reason now sits over the whole matter as the judge of what may be admitted. The relationship is gone. What remains is a mind examining a claim and deciding whether its credentials are good enough. The juror has appeared.
The Enlightenment then makes the picture stricter. Belief becomes something owed to the evidence and to nothing else: proportion your assent to the proof, give a strong claim strong belief and a weak claim weak belief, and never grant more than the evidence licenses. It reaches its hardest form two centuries after Locke, in the figure we met a fortnight ago — William Clifford, in his essay "The Ethics of Belief," of 1877, taking the case of his negligent shipowner and turning the rule into a moral command: it is wrong, always and everywhere and for anyone, to believe anything upon insufficient evidence. That is the end of the line. Belief has gone from I give you my heart to I owe the evidence a verdict, and I do wrong if I grant more than it warrants. The relationship has drained away entirely. Where credere once stood, there is now a believer who is essentially a juror, and a belief that is essentially a probability estimate — a private judgement as to whether a proposition is true.
And that late, specific, hardened meaning is the one we inherited. It is the one we absorbed without ever noticing. Which answers the question that opened this whole arc: why does the question of God feel like a trial? It feels like a trial because, somewhere in the last three hundred years, "believe" quietly stopped meaning set your heart on and came to mean return a verdict on the evidence — and if believing simply is returning a verdict on the evidence, then of course the only honest way to face the question of God is to convene a court, weigh the proof, and rule. The trial is not the natural shape of the question. It is the shape the word takes once it has been narrowed this far. We did not choose the courtroom; we inherited it, already built, along with the word.
Now I have to be exact. I have said that "believe means weigh the evidence" is a recent narrowing of something older and larger. There is a cheap conclusion here, and it is tempting: that the modern meaning is therefore a mistake, a shrinkage, a fall — that the older world had it right, believing is really a matter of the heart, and the evidence-weighing sense is a modern corruption to be thrown off. That is not what has been shown, and anyone who leaves with that has been misled in exactly the way I have spent five weeks refusing to.
Showing that something has a history is not the same as showing it is false. I have shown when the demand for evidence arose; I have not shown that it is mistaken. Consider what the evidentialist is entitled to say in reply, at full strength, because his position has not been touched. He says: a fine history lesson, and granted — "believe" once meant to set the heart, to pledge, to throw in one's lot. And do you know what we learned, slowly and painfully, over exactly those three hundred years? That setting your heart on a claim has nothing whatever to do with whether it is true. People set their hearts on all manner of things that are not there. The narrowing you regret was the single greatest discipline the human mind ever imposed on itself: the rule that, when you want to know what is actually the case, you keep the heart out of it and follow the evidence — precisely because the heart will deceive you every time it wants something to be so. So do not offer credere as a lost paradise. We did not lose it. We outgrew it, for the best reason there is: it kept telling us comfortable things that were not true. That is a serious reply. Nothing said tonight has knocked it down, and I will not pretend otherwise. The history does not refute him. Both meanings are real and both are alive, and which of them belongs in the room when the question is God — that is precisely the dispute, still open, with the evidentialist still in it.
What the history does do — and it is enough — is show the modern meaning for what it is: one option, not the only one. Before tonight, "believe means weigh the evidence" was not a position one held; it was simply the obvious, the only thing the word could mean. Now it is visible as one meaning of two: the later one, the narrowed one, with a history, an author, and a rival it displaced. It may still be the right meaning to use on the question of God. But it is now something one would have to choose, knowing there is another — not something one may call the neutral default merely because one never knew there was an alternative.
We have done this same work before. A few series ago we took Stoicism, which had been worn down into a productivity technique — stay calm, control what you can, optimise your morning — and recovered the serious old philosophy beneath it, of how to live in a universe one does not control. We took worth, which the modern world had quietly reduced to price, and separated the two again. This is the same procedure: take a rich old thing that has been narrowed to a thin modern one that everyone now mistakes for the whole, and recover what was narrowed away — not to crown the old version, but so that the modern one becomes visible as a version, a choice with edges, rather than the shape of the world itself.
So here is what you actually leave with tonight. Not an answer about God — I have not moved you on that, and I am not trying to. What you leave with is the word restored: both of its meanings, side by side, where for five weeks you had only one.
And that matters for a reason we have been building toward from the start. Recall the first episode — the uncle at the kitchen table who, when you laid out the case against God, did not argue and did not counter your evidence, but only looked at you. I asked you then to hold on to the question of what that look was. You now have two readings of it, and you have to choose between them yourself. On the first reading, the uncle had nothing: the look was an evasion, an old man's way of not admitting he could not meet your standard, your demand for evidence was exactly right, and his silence was defeat presented as calm. That reading is still fully available. On the second reading — the one you could not have seen a week ago — the uncle was registering a mismatch of kinds. You had asked him to produce proof for something that, in his life, had never been a matter of proof. You were asking for a verdict on the evidence about something he had set his heart on. On this reading the look was not "I cannot meet your standard" but something nearer to "you are asking for the wrong kind of thing" — a demand for the juror's sort of warrant applied to what was, for him, the vow's sort of commitment.
And I am not going to tell you which reading is correct. That is the honest place this ends, and I am leaving you in it. Perhaps the uncle was evading and your demand was clean. Perhaps you were the one asking in the wrong terms, without knowing there was another. What you can no longer do — what tonight removes for good — is pretend there is only one thing his look could have meant.
As every week, the point is not yours until you have done it on ground where nothing is at stake. So here is the exercise, and it has nothing to do with religion. Take the most obvious, most timeless-seeming truth about marriage that people around you live by: you marry the person you are in love with. You marry for love. It hardly seems a claim at all — it seems simply what marriage is, so plain that it feels as old as the species. Now do tonight's work on it. Go and look. For most of human history, in most places, that is plainly not what marriage was for. People married for land, for alliances between families, for a working household, for heirs, for survival. Love, if it came, came afterwards, or alongside, or not at all — and no one thought the marriage a failure for it. The idea that the reason one marries, the thing that justifies the whole institution, is being in love with the person — that idea is recent. Recent enough to have a rough date of origin: a few centuries old, not a few millennia. It is one picture of marriage, a narrowing of something older and broader, and it prevailed so completely that we mistake it for the nature of the thing. That is the whole exercise, done on something with no bearing on belief in God. You have caught yourself holding a late, particular idea as though it were timeless. And, as before, dating it does not make it wrong. Marrying for love may be a genuine advance, the best thing that ever happened to marriage. The aim is not to knock it down; it is that you can now see it as a choice with a history, rather than mistake it for the way things simply are. Do that this week, aloud, with one thing "everyone knows" that you live by. When you can date a thing you took to be eternal, it is yours.
One caution before I finish, because I can see where some of you will have gone with all this. You have heard that the whole trial is a recent thing the word does, and it is tempting to conclude that the evidence-court was a modern invention, and that you may therefore set the whole demand aside and relax. Do not relax yet. There is one more position standing in this arc, and it is the hardest of them, and it does not approach from the side of God at all. It comes through something you are more certain of than you have ever been about God — more certain than of almost anything. You are certain that some things are really, bindingly wrong: not wrong-for-you, not merely disapproved of locally, but wrong, full stop, the way cruelty to a child is wrong whether or not anyone agrees. You would never think to question that. And next week I am going to ask the question you have never put to it: where does that certainty stand? What is underneath it? I will not answer tonight; I am only marking the place. But of everything we have examined in this arc, the one belief you were most sure of — the one you would never think to doubt — is the one we test next. Hold on to that.
So — is there anyone there? Six weeks in, the answer is no nearer than on the night you began. And tonight, instead of an answer, I have given you something stranger: a word you have used your whole life and took yourself to own. Believe. You came in certain it meant one thing — the juror's verdict, weighed and returned — and you leave knowing it once meant another, the heart committed, the self pledged, and that both meanings are still alive, one of them in your own mouth every time you say it to someone you love. I have not told you which belongs in the room when the room is about God. I cannot. That is now yours to decide, knowingly, with the choice at last visible — and deciding it with your eyes open is an entirely different act from never having known there was a choice. The trial may be exactly the right setting for the question of God. Or you may have been demanding proof for a vow. Both remain standing, and I am not going to settle it for you. You wanted to know whether there is anyone there. I have only made sure that when you ask it, you know what the word in the question means — and that you, not the word, are doing the choosing.
Thanks for listening. I'll see you next time.