What explains the strengths of beliefs and ideologies over critical thinking? How does someone come to believe something that nobody ever argued for? I have been circling this question on this site for some time: in the piece on Althusser and the 2024 US election, where the two camps did not so much disagree about policies, but rather lived in different realities; in the essay on evil, where ordinary people turned out to be the indispensable personnel of extraordinary harm; and in the rereading of the Gorgias, where sophistry showed itself as speech between two, with the third term quietly removed. What I want to add here is small and linguistic: a mechanism that I think runs underneath all of these. My claim is that the beliefs which sit deepest were never propositions in the first place. They were metaphors, and at some point, usually nobody can say when, they lost the word “like.”
Take a biblical example: “All flesh is grass,” says Isaiah; “as for man, his days are as grass,” says the Psalmist. Written as a syllogism: men die, grass dies, therefore men are grass. We see at once that this is invalid, but why? It is worth reflecting on the error, because it lays bare the fault in the way humans use language in order to constitute identity.
Both premises say that something belongs to the class of things that die. That class is the middle term, and in both premises it is the predicate of a universal affirmative, which means it is not distributed: we learn that men lie somewhere inside it and that grass lies somewhere inside it, and nothing more. Two subsets of the same set need not overlap at all. This is the undistributed middle, the syllogistic form of affirming the consequent. But the more interesting fault sits in the copula. “Is” has at least three functions: “Socrates is mortal” predicates; “Socrates is the teacher of Plato” identifies; “man is an animal” places one class inside another. Metaphor exploits the ambiguity. It borrows the “is” of identity to make a predication, that men share with grass the property of being short-lived, and it works because the listener silently supplies the qualification: in this respect, not in others. Classical rhetoric had a name for the respect, the tertium comparationis, the third of the comparison. The literal reader is the one who drops the third. She hears identity.
And identity, once granted, lets everything through. Leibniz’s law says that if a and b are identical, whatever is true of a is true of b. This is valid. The person who reads it literally runs the law backwards, from one shared property to identity, and then, having got identity, runs it forward, correctly. The harm is done not by the fallacy but by the valid step after it. If men are grass, everything true of grass is true of men: it withers, it is trampled, it can be mown. A figure built to carry one feature now carries all of them, and nobody had to say which. One only has to know what grass is.
For the next step, it is useful to consult Lacan, because he reflects on what language does to us before we begin to think. In “The Instance of the Letter in the Unconscious” (1957) he takes from Roman Jakobson the observation that a sentence is built along two axes: selection, which replaces a term by a similar one, and combination, which joins a term to its neighbors. Lacan identifies these two basic linguistic operations with metaphor and metonymy.
In Peter throws a ball, the object slot can hold a ball, a stone, or a glance: the place stays, the term is selected from a set of alternatives. That is selection. The sentence also has to be assembled, subject to verb, verb to object, each piece joined to the next. That is combination. Metaphor is selection pushed one step further: a term takes the slot because it resembles what belongs there, as when Homer writes “the lion leapt” and means Achilles. Metonymy is combination pushed one step further: a term stands in for its neighbor, something it is joined to in the sentence or in the world, as when we say: Peter drank the whole bottle and we understand the wine. The lion was chosen for likeness; the term bottle was reached by proximity. For Lacan, these two operations are what Freud had called condensation and displacement. The unconscious has no mechanisms of its own; it works with the machinery of language.
Why does Lacan insist on keeping the two apart? Because they behave differently at the boundary between signifier and signified, the place where meaning is constituted. Metaphor crosses the boundary: one signifier replaces another, the replaced one sinks beneath the bar but stays latent, and from that tension a new signification appears, a fixed point. Metonymy never crosses; it slides from signifier to signifier, and the meaning is always at the next term, never here. Hence his two formulas, which sound like slogans and are diagnoses: the symptom is a metaphor, a meaning lodged where a repressed signifier was; desire is a metonymy, a movement that no object stops.
Lacan uses linguistics for a clinical reason, but there is a structural one as well. To hold a metaphor as a metaphor, to keep the replaced term latent instead of abolishing it, is for Lacan the very form of a subject who has a place in language. Where that fails, as he thought it fails in psychosis, the signifier is no longer heard as standing for something. It is heard as the thing.
So the absorption of an ideology has two steps. First, a metaphor with the bar removed: the nation is a body, no “like,” no third. This is the only step that asks for consent, and it asks quietly, since it looks like a manner of speaking. The second step is metonymy, and it asks for nothing. Once the identity holds, contiguity does the rest, a body has organs, a body has one will, a body can be infected, and it does this without inference, only through the neighborhood of words. And because desire is metonymic, the person supplies the motion himself: we want what the strong have, the strong are such-and-such, we are such-and-such. Nobody imposes the extension. Each performs it and takes it for his own thought.
The Roman historian Livy has the classic case. In 494 the plebeians left Rome and camped on the Sacred Mount, and Menenius Agrippa went out to them, not with an argument but with the fable Shakespeare later gives him in Coriolanus: the limbs rebelled against the belly, which only seemed to receive and do nothing; they starved it, and starved themselves. The people are the limbs, the Senate is the belly, the strike is suicide. Livy reports that they were reconciled. A political fight about grain had become an anatomical fact. Nobody has to argue that limbs cannot secede; limbs are not made for that. And the vocabulary kept walking after Menenius stopped. A body with one will makes a dissenting member a sick one, and the twentieth century found in the same picture the word “parasite,” with the hygiene that follows. Nobody argued that step either. It was already in the picture.
We can condense this imagery even further. The fundamental misrecognition is expressed in the image of a man who points at the moon for his dog and finds the dog staring at his finger. The Zen masters used the image for the sutras themselves: the teaching is a finger, do not mistake it for the moon. Perhaps the problem is not literalness at all but mistaking the sign for what it was meant to illuminate. “Grass” points at brevity; the dog looks at the grass. It is Lacan’s formula for psychosis seen from the other side.
But consider what happens when we point at ourselves. Here the image breaks down in an instructive way, because there is no moon. For Lacan the “I” is not an object that words point to; it is what appears between the spoken words, and the ego is an image the child once took for itself in the mirror. When someone says “I am a failure,” the finger and the moon are the same finger. There is nothing behind the word to look at except the word, and so the dog’s mistake is almost unavoidable. This is why a self-conception is so much harder to loosen than an opinion about the economy. In the Elegie after the Genjō Kōan I wrote that from one mistake, the boat moves and we say the shore moves, a self is built. This is the same mistake, seen from the side of grammar.
Adorno called this identity thinking: the concept swallows what it names and leaves no remainder. The literalized metaphor is its everyday form, and it answers the question I began with. Nobody has to argue for such beliefs because they are not beliefs; they are names that have become bodies. We take on ideologies very easily. And the traffic runs the other way as well. The concept does not only swallow the flesh, it writes on it: the tattoo, the uniform, the number on the forearm, the badge of the party, the haircut – signs of belonging stamped into skin, the word made flesh in the most literal sense there is. Kafka understood this when he had the sentence inscribed on the condemned man’s body (“In the Penal Colony“, 1914). Freedom from such beliefs cannot consist in finding the right name, the true body, the correct identity; that only exchanges one inscription for another. It consists in a changed relation to the name. Lacan’s bar and Adorno’s non-identical say the same thing from two sides: no signifier covers the subject it represents, and no concept exhausts the thing it tries to grasp. To hear the “is” as a “like” again is to occupy that gap, and it has to be occupied each time the word wants to become flesh.