Where We Left Off
Parts One to Six covered Books One to Nine. Here is what you need. This part stands on its own.
The man. Born 354 in Roman North Africa, educated at his family’s expense towards a career as a professional public speaker. By 384 he held the public chair of rhetoric in Milan. He resigned it in 386, was baptised in 387, went home to Africa, and was made a priest at Hippo in 391 and a bishop about four years after that. He wrote this book in his early forties, around 397 to 400.
The book. Confessio means three things at once: confessing sin, confessing faith, and confessing praise. It is a prayer, addressed to God throughout, and you are overhearing it.
The story so far. Nine years with the Manichees and their two eternal powers of Light and Darkness. A collapse into scepticism. Milan, and a bishop who showed him scripture could be read figuratively. Then books of the Platonists, which broke the assumption that anything real must be made of stuff, and let him say that evil is not a substance but a privation — a missing good in something that ought to have had it.
Books Eight and Nine. Having the answer turned out not to be the same as being able to live in it. Book Eight is a man who is certain and cannot move, and it ends in a garden in the late summer of 386. Book Nine is the year after: a resignation timed to the school holidays, a winter of writing in a borrowed house, baptism at Easter 387, and then his mother’s death at Ostia — where he fought his own tears all day, lost that night to a hymn, was ashamed of losing, and defended the tears anyway.
And that is the end of the story. Book Nine finishes the life. There are four books left and none of them is about what happened next. This part is about the first of them, and about why a man who has just brought his reader up to the present decides that the present is where the difficulty starts.
A Reminder About the Boxes
Six kinds of box run through this series. Word Box explains a hard term the moment it first appears. In Real Terms turns something abstract into something you can picture. How We Actually Know This names the evidence and says what it cannot show. The Argument is for places where experts genuinely disagree, and gives each side its strongest case. The Hidden Assumption digs out the premise everybody is quietly taking for granted. Remember This closes every chapter in the plainest words available.
Three Things to Watch For
This is the hardest part in the series. Book Ten is abstract, it has no plot, and it is the most influential thing in the whole work after Book Eleven. Three threads will hold it together.
One. The subject has changed and the method has not. Books One to Nine investigated a man who no longer exists, using evidence he could check. Book Ten turns the same equipment on the man doing the investigating, who cannot be interviewed, cannot be checked, and does not stay still. Every difficulty in the book comes from that.
Two. Memory is not what you think it is. Augustine’s memoria holds the past, but it also holds mathematics, skills, emotions you are not currently having, the fact that you have forgotten things, itself, and — he will argue — God. Chapters Three and Four are about a room much stranger than the word “memory” prepares you for.
Three. There is a bomb in the middle of this book. Somewhere in Chapter Nine of this part sits a prayer four words long in Latin. Around 405 a bishop read it out in Rome, a British monk in the audience could not stand it, and the argument that followed ran for twenty years, took in half a dozen church councils and an imperial decree, and settled what the western church would say about the human will for the next thousand years. This series has owed you that story since Part One.
Contents
1The Book That Should Not Be Here
Book Nine ended the story. Book Ten opens with the story finished and the author still on the page, and the first thing it does is explain why that is a problem rather than a conclusion.
1.1The break
Nine books have brought the reader from an infant in Thagaste to a grave at Ostia. The obvious next move is either to stop or to carry on with the years after 387 — the return to Africa, the priesthood, the bishopric, the controversies. Augustine does neither.
Book Ten skips eleven years without mentioning them and begins asking what a person is.
It is the sharpest structural break in the work, and it catches every first-time reader. People who have loved Books One to Nine put the book down here more often than anywhere else, and the reason is that they were reading a life and the life has stopped.
What follows is not a change of subject. It is the same subject with the tense changed, and Augustine says so almost at once.
1.2“Not what I was, but what I am now”
Early in Book Ten he states what he is doing, and it is worth having exactly, because it explains the whole rest of the work.
The point of a confession, he says, is not the record of what he used to be. It is what he is at the moment of writing.
Consider how much that gives away. Everything in Books One to Nine — the pears, the Manichees, the dead friend, the woman sent away, the garden — is, on his own account, the easy part. Those are settled facts about somebody who no longer exists, told by a man who is safe from them. The hard part is the sentence he is writing now, by a man who does not know how it will turn out.
A biography ends when it catches up. An autobiography ends when the author runs out of past. That is so ordinary that the shape of Book Ten reads as a malfunction.
The assumption is that the present is the resolution — the place from which the story is told and therefore not part of the story. The narrator is the vantage point, not one of the characters. Augustine will not have it. On his account the man at the desk is the least settled thing in the book: he does not know his own motives, he cannot predict what he will want tomorrow, and he is the only person in the work whose ending has not happened. Reaching the present has not resolved anything. It has just brought the investigation up to the one subject that cannot hold still to be investigated.
You can test this on any autobiography you have read. The narrator is always the calmest person in it, and always the least examined. That is not because writers are the calmest people. It is because the form puts them behind the camera.
The general form: treating the place you are standing as the answer rather than as another place.
That is why Book Ten exists, and it is also why the three books after it are about memory, time and creation rather than about Hippo. Once the present self has become the question, the useful next questions are what a self is made of and what time is.
1.3The second audience
Something else changes here, and it is easy to miss because it is a change in who is being spoken to.
For nine books the Confessions has been a prayer. God is the addressee; the reader is eavesdropping. In Book Ten, for the first time, Augustine turns and addresses the eavesdroppers directly. He asks what good it does them, whether they will believe him, and why they should want to.
The two addressees: from Book Ten onwards the work has a double audience and says so. God, who already knows everything in it, and a human readership, who do not and cannot verify most of it.
Augustine puts the obvious objection himself: what is the use of confessing to somebody who already knows? His answer is that confession is not the transfer of information. It is an act, done in the presence of the one who already knows, and the readers are there as witnesses to the act rather than recipients of the news.
Why it matters: it changes what kind of book this is. From here on it is no longer only a prayer overheard. It is also a public statement by a bishop who had enemies, and Chapter Eight has to reckon with that.
His remark about why people read this sort of thing is one of the coldest lines he ever wrote. The human race, he says, is endlessly curious about other people’s lives and lazy about its own. People want to hear what he is. They do not want to hear what they are.
And then the sentence that ought to be pinned above every memoir. He asks: how do they know I am telling the truth? They cannot get inside me. They will believe me because charity believes.
He is telling his readers, in the middle of the most famous work of self-disclosure ever written, that they have no way of checking it.
1.4Whether this book was part of the plan
The break at Book Ten is abrupt enough that readers have wondered for a long time whether it was always meant to be there.
What can actually be established about how the Confessions was put together is thinner than most readers assume, but it is not nothing.
Augustine wrote a book near the end of his life going back over everything he had published, correcting it and saying when and why each work was written. In that review he describes the Confessions as thirteen books, and says the first ten are about himself and the last three about scripture. That is his own division, made thirty years later, and it groups Book Ten with the life rather than with the commentary.
A biography written by a colleague who lived in his household for decades also treats the work as one book.
What that supports: that the thirteen books were regarded as a single work by the author and by the people around him, and that the boundary he recognised falls after Book Ten, not before it.
What it cannot support: that they were composed in one continuous stretch. Nothing establishes that, and a work of this length written by a working bishop over three or four years almost certainly was not.
Which leaves the question open in a precise way. Nobody doubts that the thirteen books are one work. Nobody can show that they were written as one. Readers have drawn very different conclusions from that gap, and all three of the following are held by people who have read the text more carefully than either of us.
Part One declared a position on the unity of the thirteen books and this is where that position has to be defended or revised in the open. The dispute is old, and none of the three positions is silly.
Look at the seams. Book Nine ends with a request that readers remember his mother and father at the altar — a closing formula if ever there was one. Book Ten opens by re-introducing the project and re-arguing why it is being undertaken, which is what a writer does at a beginning, not in the middle.
The declaration that the fruit of confession is “what I am now” reads like a man justifying a new instalment to readers who thought the work was finished. And the eleven missing years are missing because they were never planned for: a work conceived as thirteen books would not leave a hole where the author’s public career should be.
The author’s own late review divides the work after Book Ten, not before it, and calls the first ten “about myself”. Whatever the order of composition, the man who wrote it did not regard Book Ten as an appendix.
And structurally it is load-bearing. Book Eleven asks what time is, and answers it with a theory of the mind holding past, present and future together. That answer is unintelligible without the account of memory in Book Ten. Remove this book and the most influential passage in the whole work loses its foundation.
A third position, and the one that has quietly won ground. “Was it part of the plan” assumes a plan of the kind a modern author has: an outline, then execution.
A long work in late antiquity was dictated to secretaries in sections, over months or years, and circulated in parts. Books were physical rolls or quires that travelled separately. Composition in instalments is the normal case, not evidence of an afterthought.
On this reading the seams at the start of Book Ten are exactly what a new instalment of a continuing work looks like, and they tell you nothing at all about whether it belongs.
Where things stand: Nobody disputes the abruptness of the break or the author’s own later division of the work. What is disputed is whether the break is evidence of a change of plan.
What would settle it: A dated letter mentioning the work in progress would settle a great deal. None survives that pins the sequence down. Failing that, the argument runs on internal evidence, and internal evidence about composition order is notoriously easy to read both ways.
Why people care so much: Because it decides whether the Confessions is one argument or two good books stapled together — and Part One told you it was one argument, which is a debt this chapter is paying.
A position, declared, and it is a revision of Part One rather than a defence of it. Part One argued for the unity of the thirteen books and treated the question as settled. It is not settled, and the strongest case is the third one. The work is unified in argument and almost certainly was not composed in one run, and Part One should have separated those two claims instead of carrying them together.
Book Nine finished the life story. Book Ten skips eleven years and asks what a person is instead. It is the sharpest structural break in the work and it is where most first-time readers stop.
Augustine states the reason: the point of confessing is not what he was but what he is now. The past is the easy part, because it is settled and belongs to somebody who no longer exists.
The book acquires a second audience. For nine books God was the only addressee; here Augustine turns and speaks to his readers — and tells them they have no way of checking a word of it.
His verdict on why people read this sort of thing: the human race is curious about other people’s lives and lazy about its own.
Whether Book Ten was in the original plan is an old dispute. The best answer is that it is unified in argument and composed in instalments, and that those are two different claims which Part One wrongly ran together.
2What Do I Love When I Love My God?
A bishop, in a book addressed to God, stops and asks what he is talking about. The question is genuine, the answer takes the rest of the book, and the method he uses to get at it is the reason philosophers still read this.
2.1The question
Augustine begins the search by saying that he loves God, and then immediately asking what it is that he loves when he says that. It is not a rhetorical question and he does not answer it quickly.
Notice who is asking. This is not a doubter and not an enquirer. It is a bishop, writing a prayer, eleven years past the garden and several years into running a diocese. He has been preaching on this subject weekly. And he stops in the middle of his own book to ask what the word refers to.
That is the single most characteristic move in Augustine, and it is why this book outlived its religion’s cultural monopoly. The question is not “does God exist”. He settled that in Book Seven. The question is what the object of an attachment actually is when you cannot point at it.
2.2Everything it is not
His first move is a list of denials, and the list is unusually physical.
It is not the beauty of a body. Not the order of the seasons, not the brightness of light — which he loved — not the sweetness of melodies, not the smell of flowers or spices, not manna or honey, not anything you would hold. None of that.
And then he takes it all back, in a way. He says that when he loves God he does love a light, a voice, a smell, a food and an embrace — but of the inner person: a light that no place contains, a sound that time does not carry off, a fragrance no breeze scatters, a taste no eating diminishes, a holding that no satisfaction ends.
That double move — deny all five senses, then use all five to describe what is left — looks like poetry and is doing work. It is worth seeing the work.
Suppose you are asked what you love when you love a person. You can start with the body: the face, the voice, the way they move. But if the face changed in an accident you would still love them, so it is not that. You can go to their character, but people change and the attachment survives some of that too.
Every honest answer to the question has this shape. You get there by ruling out each thing you could point to, and you are left describing something you can only reach for in the vocabulary of the things you just ruled out.
Augustine is not being evasive. He is reporting that the only available language for what he means is borrowed, and saying so out loud, which is more than most writing on the subject manages.
2.3The interrogation
Next he goes outside and asks things.
He questions the earth, and it says it is not what he is looking for. The sea and its living things say the same. The moving air, the sky, the sun, the moon and the stars all say the same. He tells them: you have all told me what you are not, so tell me something about what I am looking for. And they answer, all of them together, that he made them.
Then the line that pulls the passage together. He asks how they said it, and answers: their beauty is the answer. What they say, they say by being what they are.
And then a distinction that is easy to skate over and worth stopping on. The same beauty, he says, is in front of everybody. It does not speak to everybody. It speaks to those who put the question and compare what they hear inside with what they are looking at. The difference between the person who hears an answer and the person who sees only scenery is not in the world; it is in whether a question was being asked.
This series does not rule on whether the world answers. What is worth reporting plainly is that Augustine builds an escape hatch into the argument in the same breath as the argument. If you look at the sea and hear nothing, his account already predicts that, and predicts it in a way that no observation could disturb. That is a weakness, and he does not flag it as one.
2.4The turn inward
Having got nothing usable from the outside, he turns round.
He goes through his own faculties in order, ranking them. There is the power that keeps the body alive, which he shares with animals. There is the power that takes in the world through the five senses, which he shares with animals too — and he is explicit that a horse and a dog have this and that it is not the thing he is after. So he goes up: past the body, past the senses, into the mind.
And there he finds the room this part is named after.
It is worth registering the shape of the journey before Chapter Three, because it is the argument of the whole book in miniature. He looked outside and was told to look elsewhere. He looked at his body and found an animal. He went into his mind and found something he could not get to the end of. The Confessions is a book about a man who kept going inwards expecting to find the bottom.
In the middle of a prayer, a bishop stops and asks what he is actually talking about when he says he loves God. The question is genuine and it takes the rest of the book to answer.
He rules out everything the senses can reach — light, music, scent, taste, touch — and then describes what is left using all five of them anyway, because there is no other vocabulary and he says so.
He goes outside and interrogates the world: earth, sea, air, sky, stars. Each says it is not what he is looking for; together they say they were made. What they say, they say by being beautiful.
He then notes that the same world is silent to people who are not asking — which protects the argument from every possible counter-example, and is a weakness he does not flag.
Getting nothing usable, he turns inward, ranks his own faculties, finds that the body and the senses are things animals have too, and goes up into the mind. That is where Chapter Three starts.
3The Palace of Memory
This is the passage philosophers read the Confessions for. It cannot be summarised without wrecking it, so this chapter takes it in order, one move at a time.
3.1Arriving
Having gone past his body and past his senses, Augustine says he comes to the fields and the wide palaces of memory. The plural matters and so does the scale: not a drawer, not a shelf, a landscape with buildings on it.
What is stored there, first of all, are the images the senses brought in. Everything ever seen, heard, smelled, tasted or touched, filed by the door it came through: colours by the eyes, sounds by the ears, and so on, each kept in its own place and each still recognisably from the sense that carried it.
Then the part that makes it a description rather than a diagram. He says that when he goes in and asks for something, some things come at once. Some take longer, as though they had to be fetched from further back. Some come pouring out in a heap when he wanted something else entirely, and have to be pushed aside to get at what he was after. And when he is telling a story in order, the things come out in order, the first standing aside for the next and going back into storage to be brought out again when he wants them.
Anybody who has ever tried to remember a name knows all four of those behaviours. What is remarkable is that nobody had written them down.
Memoria: the Latin word is wider than the English “memory”. It covers recollection of the past, but also the whole standing content of the mind — what you know, what you can do, what you are able to bring to mind — whether or not it came from an experience.
In the training Augustine had, memoria was also one of the five formal parts of rhetoric, alongside invention, arrangement, style and delivery. It was a subject with a technique attached.
Why it matters: reading memoria as “memory” makes the next twenty pages sound like a wildly overreaching theory of recollection. Read as “everything the mind has in it”, they are something else entirely.
And a small thing that is not small: he says memory is where he meets himself. When he wants to think about something he did, or is going to do, he goes to the same place.
3.2Why a palace
The image is so familiar now that it takes an effort to see it as a choice. It was one, and it came out of Augustine’s day job.
Ancient orators memorised speeches lasting hours without notes, and they had a documented technique for it.
You took a building you knew well — a house, a colonnade, a bath complex — and walked through it in your mind, fixing a series of distinct places along the route. Then you converted each part of your speech into a vivid image and put one image in each place. To deliver the speech you walked the building again and collected them in order.
The technique is described in three surviving handbooks of the sort Augustine was trained on and later taught from. It is the origin of the English phrase “in the first place”.
So when a professor of rhetoric says memory is a building with rooms and that he goes in and fetches things, he is not reaching for a poetic figure. He is using the technical vocabulary of his own trade, and his readers with the same education would have heard it that way.
What is original is the next step: he takes a filing system built to hold a speech and asks what else is in the building — and finds that most of what is in it was never filed by anybody.
3.3The size of it
Then the passage turns, and this is the turn everything afterwards depends on.
He calls it a great power, a vast inner chamber without limit, and asks who has ever got to the bottom of it. And then: this is a power of my own mind, and it belongs to what I am, and I cannot take in the whole of what I am.
Read that twice. The mind, he says, is too narrow to hold itself. Whatever of itself it does not hold is somewhere — but where? Outside it? Then how is it the mind’s? Inside it? Then why can it not reach it?
He does not solve this. He says a great astonishment comes over him, and moves on. It is one of the few places in the work where a problem is raised at full strength and simply left standing, and it is the reason this chapter is worth the effort.
The English word does almost all its work in one direction. Memory is what happened; remembering is retrieval; a good memory is an accurate archive. Every metaphor in current use — a tape, a hard disk, a file — carries the same picture.
Now look at what Augustine finds in there over the next few pages: the multiplication tables, the rules of argument, how to do things, emotions he is not currently feeling, the fact that he has forgotten things, the memory of remembering, and eventually God. None of those is a past event. Most of them never came in through the senses at all.
If the archive picture were right, most of the contents of your own mind would have no address. You did not experience the number seven. Nothing happened to you that was “how to ride a bicycle”. You did not undergo the rule that a thing cannot both be and not be.
The archive picture is not stupid; it is a good description of one drawer in the building. The mistake is treating one drawer as the floor plan, and it is why modern readers find Book Ten sprawling. It is not sprawling. The word has shrunk.
The general form: mistaking the best-known contents of a category for the boundaries of it.
3.4The sentence about mountains
In the middle of all this comes the line that escaped the book.
People go out, he says, and stand amazed at high mountains, at the great waves of the sea, at wide rivers, at the reach of the ocean, at the turning of the stars — and they walk straight past themselves without a second look.
It is a good sentence and it has had a long life. The most famous moment in that life is a letter written about nine hundred and fifty years later by the poet Petrarch, describing a climb up a mountain in Provence. At the summit, he says, he opened a pocket copy of the Confessions at random and his eye fell on this passage, and he shut the book and came down in silence.
It is worth adding, since this series does not let a good story stand unexamined, that the letter is dated 1336 and was probably written much later and shaped for effect, and that scholars have doubted for a long time whether the climb happened as described. The parallel with Book Eight is almost too neat: a man opens a book at random on a height and reads the line that changes the day. Petrarch had read Book Eight.
Past the body and past the senses, Augustine arrives at what he calls the fields and wide palaces of memory — a landscape, not a drawer.
He describes how retrieval actually behaves: some things come at once, some come slowly, some come pouring out when you wanted something else, and in a told story they queue up in order. Everybody knows this. Nobody had written it down.
The image is not poetry. The memory palace was a real technique taught in the rhetorical training he had and later taught, for memorising speeches hours long.
The turn that matters: memory is a power of his own mind, and he cannot take in the whole of what he is. The mind is too narrow to hold itself. He raises this, is astonished by it, and leaves it standing.
And the line that escaped the book: people cross the world to stare at mountains and oceans and walk past themselves without looking.
4What Is Actually In There
Having got into the building, Augustine goes through it. What he finds is stranger than the reader is expecting, and by the end of the tour he says plainly that he is out of his depth.
4.1Pictures, and things that are not pictures
The first floor is the obvious one. When he thinks of a stone, or the sun, or a city he has been to, what is in his mind is an image left behind by the senses. He was not there; a picture of there was.
But then he notices a second kind of content and the distinction is the hinge of the whole chapter. When he thinks about grammar, or about the rules for telling a good argument from a bad one, he is not handling pictures of those things. He is handling the things themselves.
There is no image of “a valid inference”. Nobody has seen one. It did not come in through a door, because it has no colour, no sound and no smell. And yet it is unmistakably in there, available on request, and he can be wrong about it — which means it is something he is consulting rather than something he is inventing.
4.2Where it came from, and why thinking is collecting
So he asks the obvious next question. If those contents did not come in through the senses, how did they get in?
Not, he says, out of somebody else’s head: when a teacher tells you something like this and you agree, what you do is recognise it. You say yes, that is right. But you can only say that if you already had something to check it against.
His conclusion is that the material was already there, lying about unsorted and unattended, and that learning consists in gathering it up and putting it where it can be reached.
Then he makes a point about the Latin that is worth keeping. The verb for thinking, cogitare, is built on the verb for driving things together into one place, cogere — the same root as the word for a herd being rounded up. On his reading, the language is telling you that thinking is not producing something new. It is collecting.
The etymology is, as it happens, correct, which is not something one can often say about ancient word-derivations. It does not prove the theory. It does show a man reading his own working vocabulary for evidence, which is a habit worth stealing.
4.3Feelings you are not having
The next room is the odd one, and it is the part of Book Ten that modern philosophy of mind keeps coming back to.
Memory, he says, also holds the states of the mind — wanting, being glad, being afraid, being sad. And it holds them in a way that has a peculiar property: he can produce them without undergoing them.
He can remember a fear and not be afraid. He can remember having been happy while he is miserable, and remember being miserable while he is happy. Being sad about a past joy is an ordinary experience. Nothing else in the building works this way — remembering a colour does not produce a colourless version of it.
His comparison for this is deliberately undignified and he knows it. Memory, he says, is a sort of stomach of the mind, and joy and sadness are like sweet and bitter food: once swallowed into that stomach they can be stored, but they cannot be tasted. He adds that the comparison is probably ridiculous and that there is something in it anyway.
Test it on yourself. Bring to mind something that frightened you badly, ten years ago. You can report it in detail. You can say how afraid you were, and rank it against other frights.
Your pulse has not changed.
Now bring to mind a shade of blue. You are not seeing blue either — but nobody expected you to be. With the fear, something that was the whole point of it is missing, and you can still handle the thing accurately without it.
That gap is what Augustine is pointing at, and no theory of memory has ever quite disposed of it.
4.4Remembering that you have forgotten
Then he walks into the wall.
Memory also contains forgetting. He can say, truthfully, that he has forgotten something. But if forgetting is the absence of a thing from memory, and he is remembering the forgetting, then the forgetting is present in memory — and if it is present, it is not doing what forgetting does.
He turns it over several ways and none of them works. If what is in memory is an image of forgetting rather than forgetting itself, then the image must have been made from the thing, and the thing would have to have been present to be copied, and we are back where we started.
And in the middle of that, a phrase this series has been waiting for since Part Three.
In Book Four, standing beside the death of his friend, the young Augustine wrote that he had become a great question to himself. Part Three flagged it, said the answer was in Book Ten, and left it there. Here it is: labouring over the problem of forgetting, he writes that he is working hard and working on himself, and that he has become to himself a soil of hard labour and too much sweat.
And a few pages later, arguing with himself about church music, the phrase returns in almost its original words: he has become a question to himself.
That is the payment of the debt, and it is worth being clear about what kind of payment it is. The line from Book Four is not answered in Book Ten. It is confirmed. The young man who did not understand his own grief and the bishop who cannot get to the bottom of his own memory are making the same report, twenty years apart, and the older one has a better vocabulary for it and no more of a solution.
4.5Admitting he is out of his depth
The tour ends without a floor plan, and Augustine says so.
The power he has been describing is his. It is part of his nature. And he does not grasp the whole of what he is. He calls it a wonder that stuns him. Then, rather than solving it, he does something that is either an evasion or the point of the book depending on your view: he says he will go past memory too, looking for what he is looking for.
A position, declared. I think the admission is the most valuable thing in the chapter and it is routinely skipped by both his admirers and his critics. Sixteen hundred years of writing about the interior life has proceeded on the assumption that looking inward yields answers. The man who started that tradition wrote down, in the middle of doing it, that he had looked as hard as he could and could not see the back wall.
Memory holds two kinds of thing: images left by the senses, and things held without images — grammar, argument, mathematics. Nobody has ever seen a valid inference, and it is in there anyway.
Since those did not come in through the senses, Augustine concludes they were already there, and that learning is gathering up what was lying about unsorted. He notes the Latin for “think” is built on the verb for driving things together.
Memory holds emotions you are not having. You can remember a terror in full detail with a steady pulse. He compares memory to a stomach in which food is stored but not tasted, and admits the comparison is ridiculous.
It also holds forgetting, which should be impossible, and he cannot solve it. In the middle of that he writes that he has become to himself a soil of hard labour and too much sweat — the phrase from Book Four, returning.
The debt from Part Three is paid, and not the way it was expected to be. Book Ten does not answer “I have become a great question to myself”. It confirms it, with better equipment and no solution.
5Everybody Wants It and Nobody Can Name It
A puzzle about searching, a claim that every human being wants the same thing, and one of the sharpest observations about people anywhere in ancient literature.
5.1How you look for something
Augustine is looking for God in memory, and he stops to ask what looking for something in memory even involves. The example he uses is a woman in a parable who loses a coin and lights a lamp to hunt for it.
Here is the difficulty. She finds it. How does she know it is the one she lost? Only because she still had a hold on it — an image of it, a recognition — the whole time she was searching. If she had genuinely lost all trace of it she could have picked it up and put it down again without knowing.
Which gives the general rule: you cannot search for something you have no hold on at all, because you would not know when to stop. Every search runs on a partial memory of the thing being searched for.
This is an old problem — Plato put it in a dialogue seven hundred years earlier and answered it with a doctrine of the soul remembering a previous existence. Augustine knows that answer and does not take it. He wants to see how far he can get with ordinary recognition.
5.2The one thing everybody wants
And then he applies it to the biggest case he can find.
Beata vita — the happy life: the standard subject of ancient moral philosophy, and a much heavier term than “happiness” in English. It means a life that has gone well as a whole, not a mood. Every ancient school had a position on it. The Stoics said it was virtue and nothing else. The Epicureans said it was the absence of disturbance. Aristotle said it was activity of the soul in accordance with excellence, with enough good fortune not to be wrecked.
Why it matters: when Augustine says everybody wants the happy life he is not making a bland observation about people liking to be cheerful. He is entering a specific, centuries-old technical debate, and his readers knew the field.
His claim is that if you ask any two people whether they want their life to go well, both say yes at once and without hesitating. There is no hesitation because there is no ambiguity about what is being asked.
But that, on the rule from the last section, is very strange. To want a thing you must have some hold on it. So where did everybody get their hold on this? Nobody has had a completed life to compare against. And people who have never in their lives had a good year still know exactly what the question means.
Augustine turns over the possibilities. Perhaps everybody was happy at some point and remembers it. Perhaps everybody was happy in the first human being, in whom, on his account, we all fell — he raises this and is careful to say the question is difficult and he is not going to settle it here. What he keeps is the smaller and firmer point: whatever the mechanism, the thing is in memory, because otherwise the word would not work.
That the universal want is real, and the universal knowledge is real, and the account of how both got there is missing, is a fair summary of where he leaves it.
5.3Wanting the thing and wanting the road
Then a distinction that does a lot of work and gets very little space.
Everybody, he says, wants the happy life. Not everybody wants what produces it. The universal agreement is on the destination and it collapses the instant anybody names a route.
It is worth pausing on how much that concedes to his opponents. The unanimity he just established turns out to be unanimity about a word. The moment it is filled in — this is what a life that has gone well consists of — the agreement is gone and the ancient schools are at each other’s throats again. Augustine notices this and does not pretend it away, and it is one of the reasons his argument is worth more than the standard version of it.
5.4Joy in the truth
His own answer, when it comes, is four words long. The happy life is joy in the truth.
That is doing two things. It is joy, so it is a state of a person and not an abstract achievement. And it is in the truth, so it is not available by being fooled. On this account a contented life built on something false is not a defective instance of the happy life; it is not an instance of it.
And then, having got there, he makes the observation that has outlived every doctrinal part of the chapter.
People love the truth, he says, when it shows them something. They hate it when it shows them themselves. And because they do not want to be wrong, but do want the thing they are wrong about, they end up loving truth in a strange shape: they love some other thing and want that to be the truth. Then, since they will not give it up, they hate the real thing for exposing it.
He adds the closing turn of the screw. They hate it, and it does not reciprocate: the truth will not let them hide from it, and they cannot stop it being there.
You do not need any of Augustine’s theology to recognise that paragraph. It is the most accurate thing in Book Ten about other people, and it was written by a man who spent nine years in a religion he later decided was false, which is presumably where he got it.
You cannot search for something you have no hold on at all, because you would not recognise it if you found it. Every search runs on a partial memory of what is being looked for.
Apply that to the happy life, which Augustine says every human being wants without hesitating. If the rule holds, everybody must somehow already have it in memory — and he cannot establish how it got there. He raises the possibilities and leaves the question open.
The unanimity is thinner than it looks. Everybody wants the destination; nobody agrees on the road. He concedes this rather than hiding it.
His own answer: the happy life is joy in the truth. Which rules out a contented life built on something false — not as a poor version of the thing, but as not the thing.
And the observation that outlives the argument: people love truth when it lights something up and hate it when it lights them up, so they love some other thing instead and want that to be the truth.
6Late Have I Loved You
The search through memory ends by going past memory, and produces the most quoted paragraph in Latin prose. Underneath the beauty there is a position, and it is the position Chapter Nine goes to war over.
6.1Not in the building
Augustine has been searching memory for what he is looking for, and he goes through it methodically and comes up empty. Not among the images of physical things, because it is not a physical thing. Not among the states of the mind, because those come and go and it does not.
Not, he decides, in the mind itself either, on the grounds that the mind changes and what he is looking for does not.
Then he catches himself and says something that reframes the whole search: why am I looking for a place in there, as though there were places in there?
The palace was a metaphor. He has been walking round it for twenty pages as though its rooms were rooms, and now he stops and says so. That is a good instinct and a rare one — most people who build a working image forget they built it.
His conclusion is that what he was looking for was not in memory before he learned it, and so he did not find it in memory at all. He found it, he says, above himself. And then he adds, with a precision that is easy to skim: nowhere is there a place. We go backwards and forwards, and there is no place.
6.2The passage
What follows is about six lines long in Latin and it is the most anthologised thing Augustine wrote.
Rendered plainly, without trying to reproduce the sound of it: I came to love you late, beauty so old and so new, I came to you late. And there you were, inside, and I was outside, and outside is where I was looking. Ugly, I went crashing about among the beautiful things you made. You were with me and I was not with you. The things that kept me away from you were things that would not have existed at all unless they were in you.
And then the five verbs. You called and shouted and broke my deafness open. You flashed, you blazed, and my blindness went. You gave off a scent, and I breathed in, and now I am short of breath for you. I tasted, and now I am hungry and thirsty. You touched me, and I caught fire for your peace.
Now look at what those five verbs are. Hearing, sight, smell, taste, touch — in that order, all five, one each.
That is not decoration. Chapter Two of this part opened with Augustine ruling out all five senses as an answer to what he loves, and then saying that he does nevertheless love a light, a voice, a fragrance, a food and an embrace of the inner person. Twenty pages later the book comes back and cashes the cheque, in the same order, with the five inner senses now described as things that happened to him rather than things he possesses. The famous passage is the closing bracket of an argument that opened at the start of the book, and most readers meet it as a standalone poem.
6.3Who is doing what
Now read it again for grammar, which is where the doctrine is hiding.
In every one of those five clauses, the active verb belongs to God. He calls. He shouts. He breaks. He flashes. He blazes. He drives away. He breathes out. He touches.
And every verb belonging to Augustine is a reaction. I breathed in. I am short of breath. I am hungry. I caught fire. There is not one thing in the sequence that he initiates. Even the “I came to love you late” that opens it is a confession of having been slow, not a claim to have arrived.
That is the whole of Augustine’s mature doctrine of grace, expressed in the arrangement of verbs, several years before anybody picked a fight with him about it.
Gratia — grace: a gift given without being earned. In Augustine’s mature account it is not only forgiveness after the fact; it is what enables a person to want the good in the first place. Later writers call this prevenient grace, from the Latin for “coming before”: the help that arrives before the wanting, and produces it.
The alternative available at the time, and held by decent people, is that God supplies the law, the example and the forgiveness, and the wanting is the person’s own contribution. On that account grace assists a free movement. On Augustine’s, grace causes it.
Why it matters: the difference between “helped to choose” and “enabled to want” looks like a technicality and is not. It decides what a person can be praised or blamed for, and it is the whole subject of Chapter Nine.
A position, declared. The passage is beautiful, and treating it purely as a lyric — which is how it is usually read aloud — quietly removes the thing it is arguing. I do not think that is a scandal; a paragraph is allowed to outgrow its book. But a series that has spent six parts on Augustine’s arguments should say plainly that this one is an argument, and that if you find the theology of grace uncongenial you have, in these six lines, been agreeing with it.
6.4Whether any of this was new
Which raises the question the last four chapters have been walking towards. Everybody says Augustine invented the inner life. Did he?
Book Ten is routinely called the first piece of writing in which a human being examines their own consciousness. It is a large claim and it is made by serious people, and it is disputed by serious people.
Point to anything earlier that does what Book Ten does. Not a memoir of events — those exist. Not a philosophical treatise on the soul — those exist. Something in which a named individual investigates his own mind as an object he cannot get to the bottom of, in the first person, in the present tense, and reports failure.
The genre after him is enormous — the medieval confessional literature, Petrarch, Montaigne, Rousseau, the modern novel, psychoanalysis — and it traces back through here. Twentieth-century philosophers of mind who have no interest whatever in his religion still quote the memory chapters, which is not a thing that happens to derivative work.
Every component is older. The idea that learning is recollecting something already held is Plato’s. The ascent from bodies through soul to what is above soul is Plotinus, as Part Six set out at length. The daily examination of conscience was a Stoic exercise with a manual. Cicero had written on the happy life, memory and the passions, and Augustine had taught Cicero for a living.
What is new is the combination and the voice, and “new voice” is a literary achievement rather than a philosophical discovery. Calling him the inventor of the inner life flatters him and erases four hundred years of people who supplied the parts.
A third position, and the sharpest. “Interiority” is a modern category. It carries the idea of a private inner theatre that belongs to the person and is authoritative about itself — the self as its own best witness.
That is not what Book Ten has. Augustine’s inner life is not private: somebody else is already in it and knows it better than he does. It is not authoritative: he says outright that there are things about him he cannot see. And it is not a possession; it is a place where he is summoned and cannot answer.
So the question “did he invent the modern self” is asking whether a thing that is not the modern self was the modern self. What he invented is something else, and calling it interiority makes it harder rather than easier to see.
Where things stand: Nobody disputes what is in Book Ten, or that the later tradition of self-examination runs through it. What is disputed is whether “first” and “invented” are the right words, and whether the category being applied fits.
What would settle it: Nothing will settle it, because it is a question about how to describe a text rather than about a fact. The nearest thing to a settlement would be an earlier work doing the same job, and none has turned up.
Why people care so much: Because the claim is load-bearing in a lot of intellectual history. If Augustine is where the inner life begins, then a great deal of modern psychology has a Christian doctrine of the soul at the bottom of it, whether or not anybody involved would put it that way.
A position, declared, and it is the third one. The most useful thing this series can say about Book Ten is that it is not an early draft of modern introspection. It is a description of a self that is opaque to itself and legible to someone else, which is a stranger idea than the one it usually gets translated into, and Chapter Seven is where it gets stated outright.
Augustine searches memory for God, finds nothing, and then stops to notice that he has been treating a metaphor as a place. The palace has no rooms. What he was looking for was not in it. The passage that follows — “late have I loved you” — runs through all five senses in order, deliberately closing an argument that opened in Chapter Two of this part. It is the second half of a bracket, not a standalone poem.
Read it for grammar and the doctrine appears. Every active verb belongs to God; every verb Augustine gives himself is a reaction. That is his whole theory of grace, laid out years before anybody objected to it.
Which is why reading the passage purely as a lyric quietly removes what it is arguing — and why anyone who loves it has, in six lines, agreed with the position Chapter Nine is a war about.
Did he invent the inner life? The most useful answer is no, and the question is the wrong shape: his self is opaque to itself and known better by someone else, which is not what “interiority” means.
7The Man at the Desk
Halfway through Book Ten the tense changes for good. From here to the end Augustine is not describing a man he used to be. He is filing a report on the one writing.
7.1The turn
The hinge is a single question, and it is a bleak one. Is not the whole of a human life on earth a trial, without a break in it anywhere?
Temptatio: usually translated “temptation”, which is too narrow. The Latin means a test, a trial, a putting-to-the-proof. Metal is temptatum when it is assayed.
So the sentence is not “life is full of enticements”. It is closer to “life is one continuous examination, and you do not get to see the paper in advance”.
Why it matters: it sets the register for everything in Chapters Seven and Eight. Augustine is not going to give you a list of things he wants and should not. He is going to describe a man who does not know, from one day to the next, what he will turn out to be.
He puts the reason for the trial plainly. He is a burden to himself, and the burden is that he is not full of what he is looking for.
And then the change of tense. Books One to Nine confessed what he had been. From here on he confesses what he is — on the day of writing, in the state of not knowing how it ends.
7.2What he can and cannot find out about himself
Before he starts the survey, he stops to say what kind of report it can be, and this is the most important paragraph in the second half of the book.
There are things about a person, he says, that only that person’s own spirit knows. Fine — that is the ordinary claim about privileged access, and it is where most accounts stop.
But there is more, he says, that even that spirit does not know. And the person can only know it if it is shown to them from outside.
The instruction to look inside yourself for the truth about yourself is so standard now that it barely registers as a claim. Know yourself. Be honest with yourself. Nobody knows you like you do. The whole apparatus assumes the inside is accessible and that the difficulty is willingness rather than capacity.
The man usually credited with starting that tradition says it does not work. His account is that self-examination has a floor, that the floor is not the bottom, and that what is under it is not reachable from where he is standing.
And notice how much he stakes on it. This is the second half of a book he has just described as a confession of what he is now. He opens it by telling you the instrument is not good enough for the job.
The modern version of this point is not religious and does not need to be. People are routinely and confidently wrong about their own motives; the reasons we give for what we did are frequently constructed after the fact and feel exactly like the real ones. Augustine’s conclusion — that the truth about a person is available, but not to them, and not by looking harder — is one option. It is not the only one. What is not available is the comfortable version in which sincerity is sufficient.
The general form: assuming that an instrument can measure itself.
This is also, incidentally, the answer to anyone who reads Book Ten as an exercise in self-absorption. A self-absorbed book does not open its central section by conceding that the author is not a reliable witness about himself.
7.3The shape he borrows
Having said he cannot see everything, he still has to organise the search. He borrows a frame rather than inventing one, and the frame comes from a single verse in one of the letters of John.
Concupiscentia, and the three-part division: concupiscentia is strong desire — wanting, in the sense of craving. The word is neutral in itself and Augustine almost always uses it for disordered wanting.
The verse he takes his structure from divides everything in the world into three: the craving of the flesh, the craving of the eyes, and the ambition of this life. Augustine reads them as the pleasures of the body, the appetite for experience for its own sake, and the desire to be well thought of.
Why it matters: this survey became the standard scheme of self-examination in the western church for a thousand years, and it is the ancestor of the examination of conscience in later manuals. It is also the reason Chapter Eight has the shape it has, and — as that chapter argues — it may be the reason certain things are missing from it.
Then, before beginning, he sets out what he thinks a person is up against and what would fix it. The image is one he has used before and it is worth having in plain words: a person leaks outward into many things, and what pulls them back into one is self-command. He is scattered and he wants to be gathered.
And it is at exactly this point, at the head of the survey and before a single temptation has been named, that he writes the four words this series has owed you since Part One.
Give what you command, and command what you will.
He then repeats it twice more before the book ends, both times at a moment of maximum difficulty. Chapter Nine is about what happened when somebody read it out loud in Rome. For now it is enough to notice where it sits: not as a conclusion, but as the ticket of admission to the examination that follows. He is saying, before he starts, that he cannot pass this test under his own power and is not going to pretend otherwise.
Halfway through Book Ten the tense changes for good. Augustine stops describing a man he used to be and starts filing a report on the one writing.
The frame is a question: is the whole of human life not one continuous trial? Temptatio means a test, not merely an enticement.
Before starting he concedes the thing that undercuts the whole exercise: there are things about a person that even that person cannot see, and looking harder will not produce them. The instrument cannot measure itself.
He organises the survey on a three-part division from a letter of John — the body, the appetite for experience, and the desire to be well thought of. It became the standard scheme of self-examination in the west for a thousand years.
And at the head of it, before a single temptation is named, sit the four words: give what you command, and command what you will. He says it before he starts, because he is saying he cannot pass the test on his own.
8What He Is Still Tempted By
A bishop in his forties writes down, for publication, what still gets him. The list is unlike anything before it, the third item is the one he is frightened of, and what is missing from it is an argument in itself.
8.1The body, and the one he cannot renounce
He starts with sex, because his readers will expect it, and the passage is not what they expect. He has been celibate for eleven years. Awake, he says, it is quiet. Asleep, it is not.
Images fixed in memory by long habit come back in dreams, he says, and in sleep they carry him not only to pleasure but to consent, and to something very like the act.
Then the question, which is a real one and is still live in philosophy: am I not myself at that moment? The difference between waking and sleeping is enormous, and yet the sleeper is the same man. He does not resolve it. He asks for the help to reach that far in as well.
It is worth stopping to register what has just happened. A serving bishop, in a book he knows will be copied and read by his ecclesiastical enemies, has volunteered a paragraph about his erotic dreams. Nothing obliged him to. Whatever else Book Ten is, that paragraph is not a safe move.
Then food, and the treatment is completely different, for a reason he names.
The distinction he draws here is the most practically useful thing in the chapter and it has nothing to do with religion.
Some appetites can be cut off. He gave up sex, and once given up it is gone; there is no daily negotiation. The decision is made once and enforced by absence.
Eating cannot be cut off. You have to do it, several times a day, for life. So there is no clean line to hold, only a shifting boundary between what the body needs and what the appetite would like, and the boundary has to be found again at every meal by the very faculty that has an interest in getting it wrong.
His own phrasing: he was taught to take food as a medicine. The trouble is that the medicine is pleasant, and while he is moving from need towards enough, pleasure comes along as an escort and he cannot always tell which of them is leading.
Anyone who has quit something outright and also tried to moderate something else knows which of the two is harder, and it is not the one people admire you for.
He adds a detail that shows how careful he is being. It is not the food he is afraid of, he says, but the craving. He is not building a theory in which some things are dirty. He is describing a relation.
Smell gets three lines. He does not think it troubles him — and then immediately adds that he may be wrong about that, which is Chapter Seven’s concession being applied in practice rather than announced.
8.2Music in church, and a man arguing with himself
Hearing gets much more, because of a specific problem he cannot settle: the singing in his own cathedral.
He grants the benefit without hedging. Words that are sung move the mind more than the same words spoken; he has felt it; the congregation is carried by it. Part Six described him weeping at exactly this in Milan, three months before his mother died, and he has not forgotten it.
And he cannot get past the fact that sometimes the tune moves him more than the meaning, and that when that happens he would rather not have heard it.
He goes back and forth. He notes that a bishop in Alexandria had the psalms recited with so little inflection that it was closer to speech than to singing, and that this is one solution. He inclines — his own word, and he explicitly refuses to give a final ruling — towards keeping the singing, on the grounds that people with weaker attachments are lifted by it.
And then, in the middle of it, the phrase from Book Four again. He has become a question to himself.
He has now used that formula twice in one book: once at the limit of what he can understand about memory, and once about whether it is all right to enjoy a hymn. That is not bathos. It is the same discovery arriving at two scales.
Sight gets the last of the bodily section, and he is generous about it. Light he calls the queen of colours, and he does not pretend to be indifferent. He notes that human beings have gone far past necessity in making beautiful things — clothes, shoes, pots, pictures — and that the people who make them get the ability from the source of beauty and mostly do not look up from the work to notice.
8.3The second: wanting to know for the sake of knowing
Curiositas: in Augustine this is a vice, and it is not what the English word “curiosity” now means. It is the appetite for experiencing and finding out, pursued for the taste of the finding out, and travelling — as he puts it — under the respectable name of knowledge.
The test he uses to separate it from ordinary pleasure is exact. People run to look at a mangled body in the road. Nobody enjoys the sight. So whatever is pulling them is not pleasure; it is the wanting to have seen.
Why it matters: this became one of the more damaging things in the western tradition, because a vice called curiosity was later used against inquiry in general. Augustine’s own targets are narrower — spectacle, divination, magic, and the prying into things that will not help anybody — but the word did him no favours.
The examples he gives of his own case are so domestic that they are almost embarrassing to read, and they are the reason the section survives.
He sits at home and watches a lizard catching flies, or a spider wrapping them, and notices that he has been watching. He no longer goes to the games, which he treats as settled — and then admits that if he is riding past a field and a dog puts up a hare, the chase can take his whole mind, and he does not turn his horse aside but his attention has gone.
He says the small ones are the problem. The large temptations he has been given the strength to leave. It is the innumerable little ones, laid across an ordinary day like snares, that he keeps walking into.
Whatever one makes of the theology, that is an accurate description of attention, written sixteen hundred years before anybody built a machine to exploit it.
8.4The third, which is worse
And then the one he is actually afraid of: the wish to be feared and loved by other people for no reason except the pleasure it gives.
Three things make it worse than the other two, and he names all three.
It comes with the job. A bishop has to be respected to function. People must take him seriously or he cannot adjudicate a dispute or hold a congregation together. So the standing cannot simply be renounced the way sex was; it is part of the work, and the pleasure it gives rides in with it.
It attaches to real goods. When somebody praises him for something that genuinely is good, the danger is not that the praise is false. It is that he finds he is pleased about the praise rather than about the good, and the two are almost impossible to separate from the inside.
And it feeds on the resistance to it. He records catching himself being pleased at being the sort of man who is not taken in by praise. There is no floor to that. Every level of detachment from it becomes a new thing to be admired for, including this sentence.
Here is the move Augustine makes that nobody else in the ancient world makes, and it is a piece of methodology rather than theology.
With food, he says, you can find out where you stand. Go without for a few days and observe. The abstinence is a measurement, and it is available on demand, and it costs nothing but hunger.
With the good opinion of other people, there is no equivalent. You cannot arrange to be despised in order to see how you take it. And you certainly cannot do it by behaving badly on purpose to find out — he raises that and rejects it, because you would have ruined the thing you were trying to measure.
So the whole apparatus of self-testing, which works well enough on the appetites, fails completely on the vice he thinks is most dangerous. The only instrument left is the conscience — and Chapter Seven has already told you the conscience cannot see the bottom of itself.
The general point survives every specific: some dispositions can only be observed under conditions you are not allowed to create. Anything you can only test by wrecking your own life is untested, and most people carrying such a disposition have quietly concluded they do not have it.
The general form: mistaking “no test came back positive” for “no test was possible”.
And at the end of the list, underneath all three, he puts a fourth thing that is not really a fourth item: being pleased with himself. Not pleased about anything in particular. Just the standing satisfaction of being who he is, which he thinks is where the other three come from.
8.5What is not on the list
Which brings us to the question that has to be asked about this chapter, and that admirers of the book usually skip.
Book Ten is the first part of the work with a declared human audience, written by a bishop with enemies, and consisting largely of a list of his own faults. Those three facts sit together awkwardly and readers have divided on what to make of it.
Look at the erotic dreams. There is no version of fourth-century church politics in which volunteering that paragraph helps him. He could have written the whole survey without it and nobody would have noticed the gap.
And look at the praise section, which names precisely the accusation a famous churchman is most exposed to — that he enjoys being important — and then says he has no way of finding out whether it is true of him. A performance does not hand the audience the charge sheet and then concede it cannot mount a defence.
Then look at what is not there. Nothing about the woman he lived with for over a decade and sent away, whom Part Four covered and who is never named. Nothing about money, or the property of the church, which he administered. Nothing about the exercise of power — and by the time he wrote this he could make life difficult for people, and did.
Nothing about the conflict with the Donatists, which was consuming African Christianity while he wrote and in which he would later come to approve the use of the state against dissenters.
What is on the list instead: erotic dreams, eating slightly too much, enjoying hymns, watching a lizard. Every item universal, every item safe, every item flattering to a man who wants to be seen examining himself. That is exactly what the best rhetorician of the age would produce if the goal were to look honest.
A third position. The form Augustine is writing in did not exist before he wrote it, and it is a public act of private disclosure. That is a contradiction at the level of genre, not a flaw in his character.
Any confession written for readers is shaped for readers. That does not make it false — a letter of condolence is composed and still meant. Asking whether Book Ten is sincere or performed is asking a question the form cannot answer about itself.
The useful question is narrower: is any particular claim in it likely to be true? On that footing the dreams look true because they cost him, and the omissions look like omissions, and both conclusions can be held at once without a verdict on the man.
Where things stand: The contents of Book Ten are not in dispute, and neither are the omissions. What is disputed is whether the omissions are evidence of management or artefacts of the structure.
What would settle it: The strongest counter-evidence to the performance reading is the structure itself: the survey follows a three-part scheme taken from scripture, and it goes where that scheme goes. A list organised on the senses and the appetites has no slot for what he did to one woman in 385.
Why people care so much: Because Book Ten is the founding document of a whole way of writing about oneself, and every memoir since has had the same problem: the disclosure is real and the selection is still doing work.
A position, declared. I think the third reading is right about the genre and the second is right about the omissions, and that the two are compatible. The scriptural frame explains why the woman is not in the list; it does not explain why a man conducting the most searching self-examination ever attempted chose a frame with no place for her in it. Part Four said the striking thing about that episode is that it is the one place where his self-examination stops. Book Ten is where that observation gets its confirmation, by omission.
On sex, he reports that waking is quiet and sleep is not, and asks whether he is himself in a dream. A serving bishop volunteering that, in a book his enemies would read, is not a safe move. On food, the point that outlives the theology: some appetites can be cut off once, and some must be practised daily, and the second kind is harder because the boundary has to be found again at every meal by the faculty with an interest in getting it wrong.
On church music he argues with himself and refuses to rule, and in the middle of it uses the phrase from Book Four again: he has become a question to himself.
Curiositas — wanting to have found out, for the taste of it. His examples are a lizard catching flies and a dog putting up a hare. He says the small ones are what get him, which is an accurate account of attention written sixteen centuries early.
The third temptation is worst because it comes with the job, attaches to real goods, and feeds on resistance to it — and above all because there is no way to test for it. You can fast. You cannot arrange to be despised.
9Four Words
A prayer four words long, in the middle of a book about memory, read out loud in Rome around the year 405. This series has owed you this chapter since Part One.
9.1What the sentence says
In Latin it is four words. Rendered into English it takes seven: give what you command, and command what you will.
Augustine writes it three times in Book Ten. Once at the head of the examination of himself, as Chapter Seven set out, and twice more later, both times at a point where he has just admitted he cannot manage something.
It looks mild. It is not.
Take the two halves separately.
“Give what you command.” The obvious prayer, when a requirement is beyond you, is to ask for the requirement to be reduced. Make it easier. Augustine asks for the opposite: leave the requirement where it is and supply the ability to meet it. Not lower the bar — give me the legs. “And command what you will.” This is the half people skim, and it is where the force is. Once the ability arrives with the requirement, the content of the requirement stops mattering. He is writing a blank cheque. Ask anything, because whatever you ask, you will also be supplying.
Put the two together and you have a claim about human beings, not just a request. It says: I cannot generate the wanting. If the wanting appears, it will have been given. And therefore nothing I manage is a thing I did on my own.
That is a statement most people would find hard to sign in any area of life — the recovering addict, the person trying to be less cruel, the person trying to work. Which is roughly why it caused what it caused.
And notice where the four words come from. Chapter Six of this part took the “late have I loved you” passage apart and found that every active verb in it belonged to God and every verb Augustine gave himself was a reaction. This sentence is that grammar turned into a request. The lyric and the doctrine are the same thing said twice.
9.2Somebody read it out
What happened next is the best-known consequence of any sentence in the book, and the evidence for it is thinner than its fame suggests.
The story: some years after the Confessions was published, in Rome, a bishop quoted this line in company. A British ascetic named Pelagius was present, could not bear it, contradicted the man who had quoted it with some heat, and the two of them nearly came to blows.
The source is Augustine himself, in a work written around 428 or 429, more than twenty years after the event.
Augustine was not there. He is reporting what he was told, at second hand, about a man he had by then spent sixteen years fighting in print, in a book whose purpose was to defend the position that started the quarrel. He does not name the bishop who quoted the line, and he does not give a date.
What that supports: that Augustine believed, and told others, that his own sentence set the controversy off, and that he thought the four words were the point of disagreement. That much is solid, and it is worth something — he knew what the fight was about better than anybody.
What it cannot support: the scene. A single hostile witness, at second hand, two decades late, is not evidence that a man lost his temper in a room in Rome. The origin story of the largest doctrinal fight in western Christianity rests on an anecdote of exactly the kind this series has declined to take at face value five times already.
What is not in doubt is that the two men disagreed, that the disagreement was about this, and that it did not stay a disagreement.
9.3Why anybody would object
The reaction is easier to understand once you know what Pelagius was doing in Rome.
Pelagius: a lay ascetic from Britain, or possibly Ireland, working in Rome from the 380s. Not a bishop and not a heretic-hunter: a moral teacher with a following among wealthy Roman Christians, urging them to take the religion seriously enough to change how they lived.
His central conviction was that people can do what they are told to do. God, being just, does not issue commands that are impossible to obey; therefore the capacity to obey must already be there; therefore anybody who says they cannot is making an excuse.
He is usually described as the enemy of grace. He was not. He held that grace is real and everywhere — in the creation of a good human nature, in the law, in the teaching and example of Christ, and in forgiveness. What he would not accept was that grace supplies the wanting itself.
Why it matters: he and Augustine agreed that God is just and that the commands are real. They split on what follows from that, and the split goes all the way down.
So a room full of comfortable Roman Christians hears a bishop read out a prayer that says nobody can do anything unless it is handed to them first, and the man who has spent twenty years telling those exact people that they can and should do better hears it as the end of his life’s work.
His objection was not metaphysical. It was that the sentence supplies an excuse to everybody who wants one, and that most people want one.
Pelagius’ objection has a shape that is worth naming, because it is everywhere and it is usually invisible. It runs: if people believed this, they would stop trying; therefore it is false.
That is not an argument about whether the claim is true. It is an argument about what would happen if it were believed. The two come apart completely. A true claim about human capacity would have exactly these consequences whether or not the consequences were bad.
Augustine saw this and made the reverse move, which is not obviously better: he treated the moral usefulness of his own position — that it produces humility — as though it counted in favour of its truth. Both men reached for consequences when the question was about capacity.
The pattern is still standard. Determinism is rejected because it would make punishment unfair. A finding about heredity is rejected because of what people would do with it. A theory of addiction is rejected because it “gives permission”. Each of those may be a good reason to be careful about how something is said. None of them is a reason to think it false.
The general form: answering “is this true?” with “what would happen if people believed it?”
It should be said in Pelagius’ defence that the consequence he feared is not imaginary. Within a century Augustine’s position had produced people who concluded that effort was pointless, and the church had to convene a council to say that it had not meant that. The argument was bad and the worry was sound, which is a combination that turns up more often than it is admitted.
9.4Twenty years
What began as a quarrel about a line in a devotional book became the largest doctrinal conflict of Augustine’s life and it outlived him.
The shape of it, briefly, because the timeline at the back has the dates. Rome fell to the Goths in 410 and Pelagius and his sharper associate Caelestius left Italy for Africa. Caelestius stayed and was condemned at Carthage in 411. Pelagius went on to Palestine, where a synod of eastern bishops examined him in 415 and acquitted him. The African bishops, unimpressed, condemned him at two councils in 416 and wrote to Rome; the pope agreed, died, and his successor reopened the case and initially found for Pelagius.
What settled it was not a council. On 30 April 418 the western emperor issued a decree banishing Pelagius and Caelestius from Rome. The Council of Carthage, with something over two hundred bishops present, condemned the teaching the following day. Rome then fell into line.
That order of events is worth stating flatly rather than in a footnote. The imperial decree came first. The council came second. Whatever one thinks of the theology, the mechanism by which it was settled was political, and the losing side has a permanent and legitimate grievance about it.
The fight then ran for another twelve years against Julian of Eclanum, an Italian bishop deposed for refusing to sign the condemnation, who was a better philosopher than Pelagius and much harder to dismiss. Augustine was still writing against him when he died in 430, with the book unfinished on the desk.
9.5What it did to Augustine
And here is the part that matters most for reading the Confessions, and that gets left out of the standard account.
The Augustine of Book Ten is not the Augustine of 428. Twenty years of argument moved him a long way, and it moved him in one direction.
The Confessions says that the wanting is given. It does not say who it is given to, or on what basis, or what happens to those who do not get it. By the end of his life Augustine was arguing that the number of those who receive it is fixed, that the choice is not made on the basis of anything foreseen in them, and that unbaptised infants are not saved. Those are positions the book you are reading does not contain.
A position, declared, and it is the last one in this part. Reading the Confessions backwards from the anti-Pelagian works is the single most common error made about it, and it is made by admirers and critics alike. The four words in Book Ten are the seed of the later system, and a seed is not the tree. What is in the book is a man saying he cannot produce his own wanting. What came out of the fight is a doctrine about who gets to. The first can be read with sympathy by people who would refuse the second, and a great many have.
Almost everything most readers know about Pelagius comes from the man who beat him. That is a situation this series has flagged before — with Faustus in Part Three and the Manichees in Part Two — and it applies here at full strength.
The genuine surviving Pelagian writings — a commentary on the letters of Paul, and a long letter of moral advice to a young Roman noblewoman — do say what Augustine says they say. Human nature retains the power to choose the good; the capacity is itself the gift; Adam harmed us chiefly by example.
And Augustine engaged with the actual texts at length rather than with a caricature. He quotes his opponents extensively, which is why so much of them survives at all. A man building a straw target does not preserve the real one in his own footnotes.
A synod of his peers examined him in person in 415 and acquitted him. That is not a detail; it means bishops who heard him explain himself did not think he was teaching what the Africans said he was teaching.
The sharpest positions attributed to “Pelagianism” — on infant baptism above all — are Caelestius’, and Caelestius was condemned first and separately. Lumping them produced a doctrine that no single person clearly held.
And his works were suppressed. What survives does so largely because it was copied under other men’s names, his own having become unusable. A defendant whose case is preserved mainly inside the prosecution’s files has not had a fair hearing, whatever the verdict.
A third reading, and the one that explains the temperature. Pelagius was out of the picture by 419. The controversy ran another eleven years, against Julian of Eclanum, and Augustine’s hardest positions were hammered out against him, not against Pelagius.
Julian pressed the obvious objections — that the doctrine makes God the author of a nature it then condemns, that it turns marriage and the body into something shameful, that it cannot be squared with divine justice — and pressed them for a decade.
Reading the whole controversy backwards from 430 therefore distorts what was at stake in 405. The quarrel over four words was about excuses. The war it turned into was about justice.
Where things stand: The dates, the councils and the surviving texts on both sides are not in dispute. What is disputed is whether the condemned position was ever held by the man condemned for it.
What would settle it: More of Pelagius would settle a great deal, and the recovery of his Pauline commentary from manuscripts circulating under other names has already changed the picture once. The possibility that more turns up is not negligible.
Why people care so much: Because the outcome fixed what the western church would say about the human will for a thousand years, and it was fixed with an emperor’s help. Anyone who thinks that does not matter should ask how they would feel about the same procedure producing the other result.
The second and third positions are both largely right, which is an awkward thing to have to say and is nevertheless where the evidence sits. Augustine engaged with the real texts; the condemnation still landed on a composite figure that was sharper than the man; and the twelve years after Pelagius left the stage did most of the damage in both directions.
“Give what you command, and command what you will.” Not a request to lower the bar — a request for the legs, plus a blank cheque on what is demanded next.
It is the same claim the “late have I loved you” passage makes with its grammar: the wanting is given, not generated.
A bishop read it out in Rome around 405 and Pelagius could not bear it. The source for that scene is Augustine, twenty years later, at second hand, in the middle of the fight — solid on what the quarrel was about, weak on the anecdote.
Pelagius’ objection was that the sentence hands everybody an excuse. That is an argument about consequences, not about truth — and Augustine made the mirror-image mistake in reply. The fight ran twenty years. Pelagius was acquitted by a synod in 415 and condemned in 418 — the day after an imperial decree banished him. And the Augustine of Book Ten is not yet the Augustine of the late anti-Pelagian works: the seed is not the tree.
10An Honest List of What We Do Not Know
Every part of this series ends the same way, by separating what is genuinely open from what is settled, and saying why in each case.
10.1Genuinely unknown
These are open questions, not gaps in this document. Each is followed by the reason it is open, because “we do not know” is only useful when it comes with a cause.
Whether the scene in Rome happened as described. Why: one source, written more than twenty years afterwards, by a man who was not present, about an opponent he had been fighting in print for sixteen years, who does not name the other witness or give a date.
When Book Ten was written relative to the rest. Why: no dated letter or external notice pins the sequence down, and internal evidence about composition order can be read both ways, as Chapter One set out.
What Pelagius actually held on the disputed points. Why: most of his writing survives only as quotations inside the work of men trying to have him condemned. What survives whole survives largely because it was copied under other men’s names.
Whether Augustine’s reports about his own inner states are accurate. Why: he tells you himself that they cannot be checked from outside, and separately that he cannot see the bottom of them from inside. Both admissions are in the book, and together they make the question formally unanswerable.
Where the contents of memory that did not come through the senses came from. Why: he rules out the senses, considers and declines Plato’s answer about a previous existence, and never supplies a third account. The gap is his, not this document’s.
Whether everybody really does hold the happy life in memory. Why: the argument requires it, he cannot establish how it got there, and he says so and moves on.
What is missing from the list of temptations, and why. Why: the scriptural frame explains the shape of the survey but not the choice of frame, and no external evidence bears on it. Chapter Eight gives the argument and declines to close it.
10.2Solid
And these are as firm as anything about a fourth-century book gets. Again with reasons, because “solid” is a claim that has to be paid for.
The text itself. Why: the Confessions is among the best-attested Latin works of antiquity. Hundreds of manuscripts survive, the oldest from within about a century and a half of composition, and the text is unusually stable across them. When this part quotes a sentence in paraphrase, there is no serious doubt about what stood there.
That Augustine wrote it around 397 to 400 and regarded it as thirteen books in two parts. Why: his own review of his works, written near the end of his life, states both.
That the four words are in Book Ten, three times, and that Augustine believed they started the controversy. Why: the text for the first, and his own statement for the second.
That Pelagius existed, taught in Rome, was acquitted by a synod in Palestine in 415 and condemned in 418. Why: council records, papal correspondence, an imperial rescript, and dated writings from both sides of the dispute. This is one of the better-documented controversies of the period.
That the imperial decree preceded the council. Why: both documents are dated, and the dates are one day apart in that order.
That Augustine’s positions hardened substantially between 412 and his death. Why: a long sequence of dated works, each responding to a named opponent, in which the change can be followed year by year.
10.3What survives of the losing side
It is worth being specific about the evidence for Pelagius, because the situation is unusual and it is better than the standard account suggests.
Three things survive substantially intact. A commentary on the letters of Paul, which circulated for centuries under other names — including, at one point, Jerome’s — and was identified as his in the twentieth century. A long letter of moral advice written in 413 to a young Roman noblewoman who had decided against marriage. And a handful of shorter letters and fragments.
Everything else comes through his opponents, chiefly Augustine and Jerome, who quote him at length in order to refute him.
What that supports: enough of the real man to say that the position condemned in 418 is recognisably related to what he wrote, and that it is sharper and flatter than what he wrote.
What it cannot support: any confident account of what he would have said under cross-examination on the hardest points. The one time he was cross-examined, in Palestine in 415, he satisfied the bishops examining him — and the record of that hearing comes to us mainly through Augustine, who thought they had been fooled.
This series takes no view on who was right. It notes only that a defendant whose case survives mainly inside the prosecution’s files is in a position that ought to make any reader careful, and that Augustine is the one who preserved the files.
10.4Where the last two parts go
Book Ten ends by looking for a mediator — something that could stand between a mind that cannot reach the bottom of itself and what it is looking for. He rejects the candidates offered by the Platonists and names his own, and the answer is the one Part Four said the Platonist books could not supply. The circle that opened in Book Seven closes here.
And then the work turns and does something nobody expects. Having spent ten books on a man, it spends its last three on the first chapter of the book of Genesis.
That is not a change of subject either, though it looks like one. The question Book Ten could not answer — what is a self, that it can hold the past and not hold itself — turns into the question of what time is, and the question of what time is turns into the question of what it means for anything to have begun.
Book Eleven is where that happens, and it is the most philosophically influential thing Augustine wrote.
Genuinely open: whether the scene in Rome happened; when Book Ten was written; what Pelagius actually held; whether Augustine’s reports about his own mind are accurate — which he says himself cannot be checked either from outside or from inside.
Also open on his own account: where the non-sensory contents of memory came from. He rules out the senses, declines Plato’s answer, and never supplies a third.
Solid: the text, which is among the best-attested in Latin antiquity; the dating and the thirteen-book shape, from his own late review; and the whole documented chronology of the Pelagian controversy, including that the emperor moved one day before the council did.
Of Pelagius there survive a commentary on Paul that circulated under other names and a long letter of advice from 413. Everything else comes through the men who beat him.
Book Ten ends looking for something that could stand between a mind and what it cannot reach. Then the work turns to the first chapter of Genesis — which is not a change of subject, because the question of what a self is has become the question of what time is.
Timeline
Book Ten has no narrative, so this timeline covers two things instead: when the book was written, and what the four words in it set off. The public dates are firm and mostly documentary.
| Date | Event | Where it appears |
|---|---|---|
| 354 | Augustine born at Thagaste. | Book I |
| 387 | Baptised at Milan. Monica dies at Ostia. The life story in the Confessions ends here. | Book IX |
| 391 | Ordained priest at Hippo, against his intentions. | outside |
| c. 395–7 | Made bishop of Hippo. | outside |
| c. 397–400 | The Confessions is written and circulated. Book X skips the eleven years above without mentioning them. | Book X |
| c. 405 | In Rome, a bishop quotes “give what you command” in the presence of Pelagius, who objects violently. Reported by Augustine about twenty-three years later. | outside |
| 410 | Rome is sacked. Pelagius and Caelestius leave Italy for Africa; Pelagius goes on to Palestine. | outside |
| 411 | Caelestius is tried and condemned at Carthage. | outside |
| 412 | Augustine’s first works against the Pelagian position. | outside |
| 413 | Pelagius writes his long letter of advice to Demetrias — one of the few of his works to survive whole. | outside |
| Dec 415 | A synod of eastern bishops at Diospolis in Palestine examines Pelagius in person and acquits him. | outside |
| 416 | Councils at Carthage and Milevis condemn the teaching and write to Rome. | outside |
| 417 | Innocent I confirms the condemnation and dies. His successor reopens the case and initially finds for Pelagius. | outside |
| 30 Apr 418 | An imperial decree banishes Pelagius and Caelestius from Rome. | outside |
| 1 May 418 | The Council of Carthage condemns the teaching in nine canons. Rome falls into line. | outside |
| 418–430 | Julian of Eclanum, deposed for refusing to sign, carries the fight. Augustine writes against him for twelve years. | outside |
| 430 | Augustine dies at Hippo with the last book against Julian unfinished. | outside |
| 529 | The Council of Orange adopts a moderated version of Augustine’s position and rejects the harder consequences drawn from it. | outside |
| — | Part Eight starts here: Book XI — what time is, and why a twentieth-century philosopher took the question seriously. | — |
Glossary
Every hard word used in this part, in plain English. Terms carried over from earlier parts are repeated so this part stands alone.
About the Author
This series is read and explained by Lovepreet Singh, and published as part of The Living Archive at misterlove.in.
The Living Archive takes difficult material and rebuilds it so that anyone can follow it, without dropping the hard parts along the way. Its first series was a fourteen-part history of Punjab running to around nine hundred and fifty pages. This is its second.
I read the book, I did the reading around it, and I wrote down what I found. I am not a specialist in late antiquity and I am not going to pretend to be one. Where a question is genuinely contested among people who are, I have said so and given each side its strongest case rather than choosing for you. Where I could not find out, I have written that I could not find out.
The method used here was worked out on the Punjab series and is written up as a portable specification, so that any part of this work can be checked, continued or rebuilt by somebody else. That is the point of an archive that calls itself living.
[Editorial note, to be replaced before publishing: the method requires a one-line statement of the author’s own position with regard to religious material, placed where readers will see it. This is the only field in the document still unfilled, and it has now been carried forward seven times. It matters more in this part than in any so far, because Chapter Nine is about a doctrine of grace and readers are entitled to know where the person explaining it stands.]
The Living Archive
The Living Archive is a method for writing hard things simply without making them shallow. It was developed by Lovepreet Singh and tested on a fourteen-part history of Punjab. It runs on three rules.
One. Simplification is not compression. The simple version is usually longer than the difficult one. Words are being spent to buy comprehension.
Two. Two dials, turned in opposite directions. Language as simple as it can be made. Content as deep as the subject actually goes. Turning both down produces something readable and useless.
Three. Every side gets its strongest case, and where nothing settles a question, that gets said. A verdict that honestly reports a deadlock is worth more than a tidy answer that is not true.
A word on sources
This document carries no footnotes. It is a synthesis of the published record rather than new primary research, and it is written for readers who want to understand a book, not for specialists who want to check my workings.
What the archive is good at here. The Pelagian controversy is one of the best-documented disputes of late antiquity. Council canons, papal letters, an imperial rescript and dated treatises from both sides all survive, and the chronology in the timeline rests on documents rather than on inference. The composition and transmission of the Confessions is also unusually well served, by Augustine’s own late review of his works and by a biography written by a colleague.
What it is bad at here. Everything interior, which is most of the book. Book Ten is a report on one man’s mind by that man, and both the external check and the internal one are ruled out — the first by the nature of the case and the second by Augustine himself, who says he cannot see the bottom of his own motives. Chapter Ten sets out what that does and does not license a reader to conclude.
On the memory chapters. The secondary literature is very large and comes from two directions that rarely speak to each other: historians of late antiquity, and philosophers of mind who read Book X as a text in their own subject. I have tried to keep both in view and to avoid the two standard errors — treating the memory material as devotional decoration, and treating it as a lost draft of cognitive science.
On the Pelagian material. I have tried to give Pelagius his own position rather than the position he was condemned for, and to be explicit about where the evidence for him comes from. Readers who want to go further should know that the recovery of his commentary on Paul in the twentieth century changed the field, and that the standard accounts written before it are correspondingly harder on him.
On translations. All renderings of Augustine in this document are my own plain-English paraphrases, made deliberately loose so that they explain rather than reproduce. They are not quotations from any published translation and should not be cited as such. This applies with particular force to the “late have I loved you” passage in Chapter Six, which exists in dozens of celebrated English versions, none of which is used here. Where the exact wording matters, go to a proper edition.
Corrections are welcome and will be acted on. Later editions will record what was changed and who caught it.
What Comes Next
Part Eight covers Book Eleven, and it is the last hard climb. The book begins as a commentary on the opening sentence of Genesis and turns, about a third of the way in, into a sustained investigation of what time is — the most influential passage Augustine ever wrote, and the one that gets him read by people with no interest in anything else in the work. These are the questions it takes on.
- What was God doing before making the world — and why Augustine refuses the joke answer he says other people give.
- What it means for something to be made out of nothing, and why that is a stranger claim than it sounds.
- Why the past and the future do not exist, on Augustine’s account, and why we can nevertheless measure them.
- “If nobody asks me, I know what time is. If I have to explain it to somebody who asks, I do not know.” The most quoted sentence in the philosophy of time.
- His answer: that time is a stretching-out of the mind, and that past, present and future exist as memory, attention and expectation — which is why Part Seven had to come first.
- And the debt from Part One: Wittgenstein on Augustine. He opens his most famous book by quoting the Confessions as an example of a misleading picture — and later returns to Augustine on time, and treats the question as a serious one about how we can know something perfectly well and be unable to say it.
That last one has been outstanding since Part One, Chapter Seven, and it is the final promise in the series. After Part Eight there is one part left, covering Books Twelve and Thirteen, and the last Hidden Assumption of the whole series has to turn on the series itself.