The Immutable Past · What Is
What Is Flattery
Flattery is ordinarily a lie told to please, and the cure is taken to be seeing through it. Read against the foundation, it is praise that would have been the same whatever the work was, because another condition governed the mark, and the size of the pleasure it gives is measured against the hearer’s own doubt and cannot report which kind it was.
What the pleasure measures
Your doubt, not the work
Ines is unsure of her taste. A stranger in a shop tells her she has a good eye.
The plain picture says flattery pleases because it is false and the hearer is vain. I ask what the pleasure is measured against. It is measured against what Ines predicted of her own taste, and a sentence lands larger the lower that prediction sat. The size does not report whether the sentence was a reading of her choice.
Contents
- The picture under the word
- Three cases the picture cannot hold
- The neighbours
- What the foundation holds
- A praise is a mark about a mark
- What the hearer feels
- What it leaves behind
- What it costs
- An ordinary Tuesday
Dana works on the operations team of a regional insurer. On Thursday she tells her manager, Tomas, that the new intake form is the clearest thing the team has shipped this year. She believes it, and it is a good form. She also needs his signature on her leave request by Friday, and the request is in the same folder as the form.
In March she told him much the same of the claims dashboard, which had broken twice by April. In June she said it of a slide deck. She has said some version of it about each thing he has shown her since January. Tomas nods, feels the warmth of it, and signs the request on Friday morning. Then, for a second, with the pen still in his hand, he feels something he cannot place, like finding that the coin he has just been given is the one he was given last month.
That small snag is the subject of this piece. The sentence was true. It was warm, and it was said by someone who meant it. By the ordinary picture of flattery it is not flattery at all, and Tomas cannot say what is wrong with it. I think something is, and that it can be found without calling Dana a liar or Tomas a fool.
The picture under the word
The ordinary picture of flattery is clear, and I do not think it is stupid. Flattery is praise that is false, or inflated past what the facts support, offered to please someone and usually to gain something. The courtier does it, the salesman does it, and the rest of us see it coming. It explains the old cases well.
Underneath is a picture of where the fault sits. It sits in the sentence, which is untrue or stretched, and in the speaker, who has a motive. The defence follows from that. The one who is praised should look for insincerity, and the ones who get taken in are the vain, who are glad to hear it, or the careless, who did not look.
That last step is where the picture does its work, and it is the step I want to test. It commits to three things. Flattery is a property of what is said. It needs a plan on the speaker’s side. And it works on the ones who could have known better. Each of those is a claim about ordinary cases, and the three cases below are ordinary.
Three cases the picture cannot hold
Case one · the kind friend
Joss reads each draft his friend Mira sends him and says that it is great. He means it as warmth. He takes nothing from her, has no plan, and is fond of her. She has sent him four drafts. She has not sent him the fifth, and when she is asked why she cannot say.
Case two · the stranger
Ines is unsure of her taste. In a shop a stranger turns her chosen jacket over and says that she has a good eye. Ines carries the sentence home like a warm stone. She is not vain. She is the opposite, and it is because she doubted that the sentence was worth so much to her.
Case three · the same words
On one afternoon two colleagues read Hana’s report. Rosa says, “This is excellent.” Pim says, “This is excellent.” The tone is the same, the pause is the same, and Hana is equally pleased by both. She could not tell you which of them was reading it. Six months later she can, and she learned it from what each of them said about other people’s reports.
The three cases share a seam with Dana’s Thursday. The picture put flattery in the sentence’s truth or in the speaker’s plan. Dana’s sentence is true. Joss has no plan. Rosa’s and Pim’s sentences are the same word for word, so that nothing in the sentence separates them. The thing that differs sits in the relation between each sentence and the work it was said about, and the picture did not look there.
The neighbours
Several earlier pieces stand close, and a reader could think I am repeating one of them. I have read each to its last paragraph. None takes the word as its subject, and each takes a different clause from the one used here.
What Is Perfection is the nearest in subject. There, praise is read as a report on the reader’s resolution: “perfect” says that the miss fell below what this reader could distinguish. That is a claim about how much of a mark registered. I ask a different question of a praise, which is which condition governed the mark, and it can be asked of a reader with a fine resolution as easily as a coarse one. A sharp reader can praise from something other than the work.
What Is a Bluff is the nearest in shape. A bluff is a relation between a mark and a holding settled before it, and the reader meets only the mark. Flattery has some of that: the condition that governs it is not on the surface. The differences are two. A hand is closed and can be reached only by a showing, while the work in the folder is open in the present and the hearer holds it. And a bluff needs a holding that the mark misstates, where Dana holds no hidden belief at all. Its last paragraph also says that a true sentence offered for how it will read stands in the same relation as a false one. I take that sentence at a different point. It is about how a mark will read, and I ask which condition governed the mark when it was made.
What Is a Reputation owns marks held in each reader’s denominator, including what a name costs its owner once her conforming marks stop carrying information. What praise does to the one who gives it, over years, is that piece’s, and I do not repeat it. What Is Unmoved owns the point that a small felt surprise reports a quotient and is silent on change. Here the felt surprise is large, and what it is silent on is something else.
One earlier essay deserves a line. Confidence Cannot Be Delivered Whole (March 2026) draws the same line in practice, praise as evidence language and not flattery. It predates the foundation and I read it as report. It is not contradicted, and the derivation below is what it did not have. No piece is superseded, and the earlier essay stays as it is.
What the foundation holds
Six statements do the work, and I use them as the foundation states them. An artifact is an imperfect historical mark made in relation to an exemplar, and by P10 every condition has an exemplar, a complete and exact form against which any artifact can be measured. By the foundation’s account in Part III an artifact carries a magnitude and a phase, and the phase is which condition it serves. The phase is relational: it is fixed by which idea has the actualizer, and not by the mark or by its maker. Fidelity is the alignment of the artifact’s phase with the idea’s. Corruption is the phase drifting away from it.
By P8 a condition is prior to what it conditions, and ideas have people. So the condition a mark serves is the one that had the speaker at that moment, and the speaker did not choose it. By R11 the evaluative question about any artifact is fidelity and not success. By R8 the human experiences only the quotient of the actual against her expectation, and by P4 that expectation comprises subconscious prediction and ideas. Surprise is the logarithm of the quotient’s magnitude. By R2 records are write-protected: they update expectation forward and are never re-opened, and the record retrains prediction.
I treat a spoken sentence as a mark, since an act that leaves a trace in someone’s record has generated an artifact. One caution carries over from earlier pieces. The foundation gives no expression for comparing two hosts’ quotients and no rate or scale for the update to an expectation. Where the argument needs either I say so and leave it.
A praise is a mark about a mark
Start with the form. Dana’s intake form is an artifact, made in relation to an exemplar of what a clear form is. By R11 the question to ask of it is fidelity: how near did the mark come to the form it was made in relation to? That question has an answer, and a good form has a high one.
Now Dana’s sentence. It is a second artifact, an evaluation of the first. It claims a condition: it presents itself as a reading of the form against its exemplar. But the sentence has its own phase, and by Part III and P8 the phase is fixed by which condition had Dana when she spoke. The leave request was in the folder. If that condition governed the sentence, then its phase sits near “getting the signature” and some distance from “an evaluation of the form”. Its fidelity to the second is the alignment of its phase with that condition, and the alignment is low, even though the content is true.
That is how a true sentence can be flattery. Truth belongs to what a sentence says about its subject. Fidelity belongs to the relation between the sentence and the condition it claims to serve. The two come apart as easily as a true answer that dodges the question. Flattery is the case in which they have come apart in a particular direction: an evaluation in form, governed by another condition in fact.
It does not need a plan. Joss’s sentence is governed by something like “keeping this friendship easy”, and he did not choose it, in the sense P8 gives. The condition had him. That is also why the word cannot be settled by looking into the speaker. From the inside, Dana can feel that she means it, and she does. What she cannot see from the inside is which condition would have changed her sentence if the form had been worse.
How would anyone know? Not from one mark, since one sentence shows only its own content. From many, yes. A mark governed by the work’s fidelity moves when the work moves. A mark governed by another condition follows that condition and stays put. So the question to put to a praise is a question about a different world: what would she have said of a worse form? Dana: the same. Joss: the same. Rosa: the same, as Hana found out. The test needs works that differ, and where the works do not differ it tests nothing.
Marked construction 01 — consequence, not doctrine
Praise is flattery when the condition that governs the mark is not the condition the mark claims to read. It is a fact about the relation between the sentence and the work, and it can be present with a true sentence, a warm speaker and no plan. What shows it, across several marks, is that the praise does not move when the work does.
Grounds. P10, for the work’s exemplar and so for what an evaluation claims to read against. Part III, for the phase of a mark as the condition it serves, for fidelity as alignment, and for corruption as drift. P8, for the governing condition as the one that has the speaker. R11, for fidelity as the evaluative question. Marked as the piece’s own: the identification of flattery with a mismatch between the claimed and the governing condition, and the reading of an unmoving praise as the sign of that mismatch and not its definition. The foundation gives no measure of how far a phase has drifted, and the piece gives none. Would be wrong if people asked to sort a set of praises into flattery and not-flattery sort them by whether the sentence is true, or by whether the speaker gained, and the sorting stays that way once they are shown how each speaker’s praise varied across several works. Or: a speaker whose praise does not move across works of plainly different fidelity, and whose praise a later, independent reading of each work confirms.
What the hearer feels
Go back to Ines in the shop. Before the stranger spoke she held a prediction of her own taste, built from the jackets she had bought and returned. It was low. The sentence “you have a good eye” landed against it. By R8 and P4, what she met was the quotient of that sentence against her prediction, and with the prediction low the quotient was well above one. The surprise was large and positive, and it was hers.
Now put a second shopper beside her, a woman who has bought well for thirty years and holds a high prediction of her taste. The same stranger says the same words in the same voice. Her quotient is near one, and what she feels is a small nod. One sentence, from one speaker, in two sizes. The size was set by the hearer’s own prediction, and the speaker’s sincerity had no part in it.
Here is the step that matters. The locked definition of surprise is the logarithm of the quotient’s magnitude. It has no term for which condition the mark served. The feeling is built from that quotient, and so the feeling has no access to the phase. Hana’s two colleagues produce one quotient. If Rosa’s sentence was a reading of the report and Pim’s was a courtesy, nothing in what Hana felt could have reported it.
Figure 01
Four praises, what governed each, and whether it moved with the work
| Speaker to hearer | What governs the mark | What the speaker gets | What she says of the next work | Reading |
|---|---|---|---|---|
| Dana to Tomas | The form is clear; the leave request is in the folder. | Her leave request is signed. | The same sentence for the next thing he shows her. | Unmoving. Governed by another condition. |
| Joss to Mira | Mira is a friend and the friendship should stay easy. | Four drafts, four sentences. | The same sentence for the fifth. | Unmoving. Governed by another condition. |
| Rosa to Hana | Not visible to Hana on the day. | Hana is pleased. | The same sentence for each report on the team. | Unmoving. Learned from the record. |
| Pim to Hana | Not visible to Hana on the day. | Hana is pleased. | Said of this report and not of two others. | Moves with the work. A reading. |
This is unwelcome in a particular way. People reach for the feeling as evidence: it felt right, it felt true, I could tell. The felt size is a fair measure of one thing, and it is evidence about the distance between a sentence and the hearer’s doubt. It is not evidence about the sentence. A true reading and a flattering one land the same when the same quotient meets the same prediction. And the lower the prediction sat, the larger the landing, which means the hearer with the most doubt is the one least able to tell by feel.
Marked construction 02 — consequence, not doctrine
How large a praise feels is set by the quotient of what was said against the hearer’s own prediction of the work, and the definition of surprise carries no term for the condition the mark served. The felt size therefore cannot report whether the praise was a reading of the work, and it is largest on the hearer who predicted least.
Grounds. R8, for the hearer meeting only the quotient. P4, for the prediction as the hearer’s own and composed of subconscious prediction and ideas. The foundation’s locked definition of surprise, as the logarithm of the quotient’s magnitude. Marked as the piece’s own: the identification of the absence of a phase term in that definition with the feeling’s silence on phase. The foundation does not say that the pleasure is nothing more than the surprise, and the piece rests on the weaker reading, that the quotient is what the hearer meets. Would be wrong if hearers shown the same sentence from a flatterer and from a reader, with no cue about the speaker, reliably feel the flatterer’s as smaller, or larger, at the same prediction. Then the size carries something about the phase, and the construction fails.
What it leaves behind
Mira’s fourth draft is weak. Joss says it is great. By R2 the sentence stays where it was made, and it is added to her record beside the first three. The record retrains prediction, so her expectation of her own drafts has been moved up by four sentences, and the move is forward. She sends the fifth draft to an editor who reads it carefully and says what she read. The editor’s reading is the same reading it would have been. Against the raised expectation it lands as a quotient below one. It reads smaller than it would have read against the prediction she had before Joss.
That is the cost, and its direction is all the foundation gives. A flattery that is taken in does not leave a false belief that can be removed by correcting a fact. It leaves a mark in the record that has already done its work on the expectation, and the record cannot be re-opened. I do not know how far one sentence moves a prediction, and the foundation has no rate for it.
What can be done is the only thing that was ever available, which is to add. The repair of a flattery is a new mark: a reading, by someone, of the work, with something in it that the work could have been wrong about. It sits beside the old one and neither removes it nor takes it back. It is also why Mira’s silence about the fifth draft makes sense. She had met, without a name for it, the fact that Joss’s sentences did not move.
Marked construction 03 — consequence, not doctrine
A flattery taken in is added to the hearer’s record, and the record retrains her expectation of her own work, forward. A later honest reading then lands against a higher expectation and reads smaller than it would have. The repair is a new mark and not an edit of the old.
Grounds. R2, for write-protection and for updating forward. P4, for the expectation as the quotient’s denominator. The foundation’s section on the host’s own act in its own denominator, for the record retraining prediction. Marked as the piece’s own: the step from “the record retrains prediction” to “a praise of one’s work, taken in, is retrained as a higher expectation of that work”. The foundation says the first. The second is an application of it and the piece’s own. No size or rate is claimed. Would be wrong if a hearer whose reading of her own later work, taken over many instances, is unchanged in size after praise she took in, or reads larger. Then the expectation was not retrained by the mark in the way the construction says.
What it costs
The first cost is that a good deal of kind speech turns out to be flattery in this sense. A parent’s praise, a friend’s, a team’s warmth on a Friday: many of those sentences do not move when the work moves, because the condition that has the speaker is love, or loyalty, or keeping the room easy. I do not say they are faults. They are marks, and they are marks of something, and the something is genuine. What they lose is their use as a reading of the work. Taken as a reading of the speaker’s care, they are accurate, and often the most accurate thing on offer.
The second cost is that the test cannot be run on the day. One sentence shows its content and nothing else, and the feeling it produces is silent on the question. A stranger’s praise, however warm, cannot be graded when it arrives. It can be graded later, by what the same speaker said of other work, and that means by waiting and by keeping a record. That is slow and it is the only way I have found.
The third is smaller and falls on the one giving praise. Run the question on your own sentences. What would I have said if it had been worse? Many of mine would have been the same. That is not a verdict on me, because the condition that had me was usually a decent one. It does mean that the sentence I thought was a reading was a courtesy, and that the person who got it was told less than I meant.
Last, I have not said that Tomas was wronged by the signature, or that Dana will answer for anything, or that anyone was owed the true reading. The ledger is not in this piece, and nothing here is offered as comfort.
If the praise you lived on turns out to have been flattery
One belief can come off here, which is that you were foolish to enjoy it. The pleasure was a quotient doing what a quotient does. It was built from your own prediction and a sentence that landed above it, and you did not do anything wrong by having one.
Understanding the mechanism does not turn the glow off, and there is no schedule for how long it takes to read the old sentences differently. If the thought that arrives is that nothing anyone said to you was true, that is not what I have argued. Many of those sentences were true, and many were marks of care. The work is still there, and it can still be read.
This essay is not the right tool for deciding whether someone loved you, or whether you can trust the people who praised you. If this is looping, or if it has taken the ground from under a relationship you depend on, that is worth saying to someone you trust.
An ordinary Tuesday
These are practice and not doctrine. They are revisable on evidence and they are not grounds for anything above.
- Before you praise, name the part. Say what in the work the sentence is a reading of. If you can point to a paragraph, a joint, a number, the sentence is probably about the work. If you can only point to the room or to what you need, the sentence is about that.
- Ask what you would have said if it had been worse. If the answer is the same sentence, say a different one: “I’m glad you showed me.” That makes no claim about the work, and it is honest about what governs it.
- When you are praised, hold the size apart from the verdict. The size is yours. It tells you how far the sentence sat from your doubt. It does not tell you how close the work is to its form.
- Ask which part. A reading of the work can be pointed back at the work. “Which bit?” is a plain question, and it costs the speaker little if the praise was a reading.
- If you took one in, add a mark. Get a reading from someone whose praise has moved in the past, and read the work again yourself. The first sentence stays where it was, and a second one beside it is all the repair there is.
Praise either moves with the work or carries something else the speaker was holding, and the size you feel cannot tell you which. It can only tell you how far the sentence was from your doubt.
What is left open
- Is flattery a mismatch between the condition a mark claims and the condition that governs it?
- Construction 01 says so. Falsifier: people sort praises by truth or by gain, and the sorting survives once they see how each speaker’s praise varied across works. The piece’s own, marked.
- Is the felt size of a praise silent on the condition the mark served?
- Construction 02 says so. Falsifier: hearers with no cue about the speaker reliably feel a flatterer’s sentence as a different size from a reader’s at the same prediction. The piece’s own, marked.
- Does a flattery taken in raise the hearer’s expectation of her own work?
- Construction 03 says so. Falsifier: a hearer whose later readings of her own work do not shrink after praise she took in, over many instances. The piece’s own, marked.
- How far does one sentence move a prediction?
- The foundation gives no rate or scale for the update to an expectation. The piece states the direction and stops there. Stated, not crossed.
- How do two hosts’ quotients compare?
- The foundation gives no expression for it. The piece does not need the two to be compared, and says so where it comes close, in the two shoppers. Stated, not crossed.
- Does a machine have surprise?
- Listed as open. This piece speaks of hosts who praise and are praised. A system that produces praise is left where the foundation leaves it. Not touched.