Why Gadamer insisted that Truth and Method offered no method

Without prejudices there would be nothing to correct#

Vorurteil does in German what its Latin ancestor praejudicium did: it names a judgement formed before the evidence is in, with no implication that the judgement is wrong. Gadamer restored the neutral sense, then pushed further. "The fundamental prejudice of the Enlightenment," he writes, "is the prejudice against prejudice itself, which denies tradition its power" (second revised edition, p. 273). The sentence was designed to irritate, and it worked.

The claim is routinely reported as something sillier than it is. Gadamer is not saying prejudices are benign, or that tradition is a reliable guide, or that the Enlightenment was wrong to distrust inherited authority. He is saying the demand for an understanding purged of all pre-judgement is incoherent, because the pre-judgements make the encounter possible. You cannot read a page of Thucydides without expectations about what a historian does, what a city is, what counts as a reason. Those expectations are not obstacles between you and the text; they are the only things that let it say anything at all, including the things that contradict them.

Discrimination, then, rather than purification. Some prejudices enable and some distort, and the way to find out which is which is to be brought up short: the text resists, the anticipation of meaning turns out not to fit. That failure is the only instrument Gadamer offers, and it cannot be systematised, since one cannot set out in advance to be surprised. The obvious rejoinder — that distorting prejudices are exactly the ones never contradicted, because they arrange the world so as to confirm themselves — is the one Habermas made, and it is good.

History has already worked on us before we start work on it#

Wirkungsgeschichte is usually rendered "effective history", and the consciousness answering to it becomes "historically effected consciousness" in the Weinsheimer and Marshall translation used throughout here. A work's reception is not varnish over the original that scholarship can strip away; it is part of what the work now is, and it has formed the reader before the book is opened.

Gadamer's best illustration is legal. A judge confronted with a statute does not first establish what it means and then, separately, apply it. Understanding and application are one act; the case makes the meaning determinate, and a statute a century of courts have applied is not available in some pristine pre-applied condition. Theology works the same way: the sermon is no appendix to the exegesis. He treats this as the paradigm rather than a peculiarity of law and preaching, which is what separates him from Dilthey.

Hence the fusion of horizons, Horizontverschmelzung, the most quoted and least understood term in the book. A horizon is the range visible from a standpoint, and Gadamer insists that it is mobile: it travels with us, opens and closes, and is never a wall. There are not two sealed horizons, the text's and ours, waiting to be welded; there is one horizon in motion, and understanding is the moment at which the past stops being an object over there and becomes a claim on us here. His test case is the classical. A Greek tragedy is not classical because a period of taste decided so, but because it goes on addressing people with no share in the culture that made it — a fact about effects rather than an aesthetic compliment.

The chapter on play is the argument, not the ornament#

Readers in a hurry skip Part One, which is a mistake: the ontology of the work of art is where Gadamer builds the structure everything else rests on. His model is Spiel, play — a word covering games, theatrical performance and the play of light on water — and what interests him is the grammar. We say the game was a good one, that the play went on for hours; the idiom is exact. "The actual subject of play is obviously not the subjectivity of an individual who, among other activities, also plays but is instead the play itself" (p. 104). Skilled players are absorbed. They are taken up into something with its own rhythm that neither controls and both constitute. The spectator is in the same position: being outside oneself, Gadamer writes, is "the positive possibility of being wholly with something else", a self-forgetfulness in which watching consists (p. 126).

The festival extends the point across time. It has no existence apart from its celebrations, and yet it is not a different festival each year; repetition is its mode of being rather than a threat to its identity. "The festival is a commonality," he writes in The Relevance of the Beautiful — the 1977 essay translated by Nicholas Walker and edited by Robert Bernasconi for Cambridge in 1986 — "and is the representation of commonality itself in its consummated form."

Transpose that structure to reading and the rest of the book follows. Understanding is an event with the form of play: the interpreter participates rather than operates, the text is a partner rather than an object, and what emerges belongs to neither alone. Which is why there can be no method. A method is what a subject applies to an object, and Gadamer has just denied the relation has that shape.

Habermas was right that tradition can lie; Gadamer that critique has an address#

The quarrel began in print in February 1967, when Jürgen Habermas published Zur Logik der Sozialwissenschaften as a 195-page supplement to Philosophische Rundschau: an extended and respectful demolition of Truth and Method. Gadamer answered that year with "Rhetorik, Hermeneutik und Ideologiekritik". Habermas returned in 1970 with "Der Universalitätsanspruch der Hermeneutik", in the Festschrift Hermeneutik und Dialektik got up for Gadamer's seventieth birthday by Rüdiger Bubner, Konrad Cramer and Reiner Wiehl; Gadamer's "Replik" followed in the Suhrkamp collection Hermeneutik und Ideologiekritik in 1971.

A hermeneutics of tradition, on Habermas's charge, has no leverage on what he called, in a 1970 paper in Inquiry, systematically distorted communication. Ideology is no misunderstanding that better listening would clear up. It is a pattern of understanding that works precisely by seeming transparent to those inside it, and a discipline taking the transmission of meaning as its horizon will reproduce the distortion while congratulating itself on its openness. Labour and domination, Habermas argued, operate behind the back of language.

Gadamer's reply concedes more than his reputation suggests. Reflection on a prejudice, he wrote, "brings before me something that otherwise happens behind my back. Something — but not everything." The question he pressed back was where the critic stands. A theory competent to unmask distorted communication is itself a body of concepts with a history, taught in institutions, read by people formed in a tradition of emancipatory thought running from Kant through Marx. It has, so to speak, an address. That does not refute it; it does make critique one more move inside the hermeneutic situation rather than a vantage point outside, and Habermas never fully answered the point.

The exchange did not end in bitterness. When Gadamer received Stuttgart's Hegel Prize on 13 June 1979, Habermas gave the address and called his achievement the urbanisation of the Heideggerian province. It is the best sentence ever written about Gadamer, and it comes from his sharpest critic.

Paris, April 1981: the conversation that did not take place#

On 25, 26 and 27 April 1981 the Goethe Institute in Paris hosted an encounter arranged by Philippe Forget, with Manfred Frank, Jean Greisch and François Laruelle also taking part. Gadamer opened with a lecture, "Text and Interpretation", carrying an unusually warm invitation to French thought. Jacques Derrida did not reply to it. He offered three short questions instead, published as "Three Questions to Hans-Georg Gadamer": whether the good will to understand that Gadamer presupposes belongs to the metaphysics of the will Heidegger had taught them both to distrust; whether psychoanalysis does not break the continuity of context hermeneutics assumes; and whether the relation to the other is secured by understanding at all, rather than by interruption. The next day he read a paper on Nietzsche and Heidegger, "Interpreting Signatures", which did not mention Gadamer.

Gadamer's answer was short and slightly nettled: good will is no Kantian axiom but the minimal condition of speaking to anyone, and even immoral people, he observed, try to understand each other. The exchange and the commentaries were collected by Diane Michelfelder and Richard Palmer as Dialogue and Deconstruction: The Gadamer-Derrida Encounter (SUNY Press, 1989), a book largely devoted to explaining why its central event failed to occur. Gadamer returned to the non-meeting for a decade afterwards, in "Destruktion and Deconstruction", in a letter to Fred Dallmayr, and as late as 1994 in "Hermeneutik auf der Spur".

The coda is better than the encounter. In 2023 Antoine Pageau-St-Hilaire published in Continental Philosophy Review a correspondence between the two men running from March 1977 to July 2000, together with Derrida's tribute after Gadamer's death in 2002, "How Right He Was! Gadamer, My Cicerone". The dialogue of the deaf had been going on in private for twenty-three years.

The conservatism charge deserves a fuller hearing than it usually gets#

Three objections are worth separating. The first is political, and it has a biography attached. Gadamer remained in Germany throughout the Nazi period, moved to a chair at Leipzig at the end of the 1930s and was made rector there in 1946 under Soviet occupation, and his post-war account of his own conduct was, at best, economical. Teresa Orozco's Platonische Gewalt: Gadamers politische Hermeneutik der NS-Zeit (Argument, Berlin, 1995) read his lectures of the 1930s and 1940s as political interventions of a recognisable kind; Richard Wolin pressed the case for a general readership in "Untruth and Method: Nazism and the Complicities of Hans-Georg Gadamer", in The New Republic of 15 May 2000. A philosopher's conduct does not settle his philosophy. A philosophy assigning positive epistemic weight to authority and tradition should still expect the question.

The second is conceptual. If understanding runs through tradition, and tradition names what the transmitters chose to keep, then "tradition" does the work "authority" used to do with the coercion left out of the description. Gadamer's answer — that authority rests on recognition of superior insight rather than obedience — describes the good case and passes over the ordinary one.

The third is sharpest. Oliver Ramsbotham, writing in the Journal of Dialogue Studies in 2016, argued that in intense and asymmetric conflicts the parties do not treat one another's positions as horizons at all. They contest the very distinctions — between fact and value, between reality and opinion — within which a fusion could be described, and reaching for fusion early mistakes an unresolved conflict for a misunderstanding. The objection generalises. Gadamer's model conversation is two people of good faith attending to a subject matter; he says remarkably little about the text composed to mislead, or the message whose author has no interest in the subject matter whatever. Advertising, propaganda and fraud are not marginal cases of communication, and a hermeneutics that makes openness the right stance towards any text is disarmed before them.

What a language model has, and what it has not#

Take the machine's case seriously, because it is stronger than hermeneuts usually admit. A language model stands in a tradition of texts, in a fairly literal sense: it is a compression of them. It has something answering to a horizon, a settled sense of how words have been used, shaping what it can say before any exchange begins. The exchange is open-ended; neither party knows where it will finish; contradicting it with a good reason changes the reply. The question-and-answer structure Gadamer took from Plato and made the engine of interpretation looks, on the face of it, satisfied.

Consciousness is the least useful place to locate the difference, since nobody can adjudicate it and the argument about talking to language models has been stuck there for years. Gadamer offers something more discriminating: understanding alters the one who understands. "Being experienced does not consist in the fact that someone already knows everything," he writes; "the experienced person proves to be, on the contrary, someone who is radically undogmatic" (p. 350). Experience in his sense is the experience of negation, of finding that things are not as one took them to be, and it accumulates into a way of standing towards the future. The interpreter who ends a conversation holding exactly the anticipations she began with has not understood. She has processed.

That is the condition a model fails, structurally rather than contingently. The weights are the same afterwards as before; the horizon cannot move, being fixed at training time, and whatever happens within the exchange is discarded when the context window closes. Something that cannot be changed by what it meets cannot be at stake in meeting it, and being at stake is not, for Gadamer, an optional intensity of understanding. It is what understanding is.

He never wrote that sentence, and the extension is mine: the case did not exist for him to rule on. It invites a reply, too. The human at the other end is changed, so perhaps the event occurs and only one participant undergoes it. Gadamer would find that an odd thing to call a conversation — though a book undergoes nothing either, and he still called reading a dialogue. The honest position is that the machine sits between the book and the interlocutor, and his vocabulary was not built to locate it.

The literature is arriving in quantity. Communicating with AI: Philosophical Perspectives, edited by Herman Cappelen and Rachel Sterken, went online at Oxford on 31 August 2026: sixteen chapters in four parts, open access, with Cappelen and Josh Dever on hyper-externalism, Anandi Hattiangadi and Anders Schoubye arguing that the outputs of large language models are meaningless, Diane Proudfoot on Wittgenstein's lion, Regina Rini on autonomy traps. Oxford Academic lists the print edition for 29 September; the press's catalogue gives 29 December, 448 pages. This site reviews it separately. Hermeneutics barely appears in the contents. The debate is being conducted almost wholly in the vocabulary of analytic philosophy of language, a defensible choice and a narrowing one, and precisely the narrowing Gadamer spent a career describing.

He would not have been surprised. The modern reflex, that securing a procedure is what secures a truth, misdescribes what happens when one mind meets something it did not make. Whether he was right is a live question, and a better one than whether the machine is conscious.

Confucius is a name most readers know and few can place. This profile introduces the philosopher behind it: his ideas of ren, ritual, the exemplary person and the rectification of names; his claim that good government begins in character; and the compiled, argued-over text we call the Analects.

Confucius