Self, Culture, and Language: a reading of Merleau-Ponty

“Our contact with ourselves always takes place through a culture, or at least through a language that we have received from the outside and which guides us in our knowledge of ourselves. So that, finally, the pure self, the mind, without instruments and without history, if it is indeed like a critical instance that we oppose to the pure and simple intrusion of ideas suggested to us by the environment, is only accomplished in effective freedom through the instrument of language and by participating in the life of the world.”

— Maurice Merleau-Ponty, Talks (Causeries), 1948 ^[1]

At the meeting point of philosophical reflection and introspective inquiry, Maurice Merleau-Ponty’s Causeries — published in 1948 and translated into English as Talks — condenses, in a single sentence, a demanding claim about the relationship between self-awareness, language, and culture. The sentence is built on two moves that are easy to run together but worth separating from the outset. The first is descriptive: contact with oneself “always” passes through culture, “at least” through language — a graded claim, since “at least” concedes that language may be the minimal, not the only, mediation at work. The second is normative, even hopeful: the self is not thereby diminished but “accomplished in effective freedom” through that same instrument. The difficulty, and the real subject of this essay, is to understand how a self so thoroughly mediated by something external to it can still be called free — and what becomes, on this view, of the “pure self” Merleau-Ponty names only to qualify out of existence.

Born in 1908 in Rochefort-sur-Mer, Merleau-Ponty became the leading French exponent of an existentialist phenomenology centred on embodiment and perception, developed above all in his 1945 Phenomenology of Perception and revisited in his late, unfinished ontology of The Visible and the Invisible. Talks, a set of radio interviews given nearly a decade after that first major work, restates its technical arguments in an accessible register, without renouncing their precision — which is what makes the sentence quoted above worth reading closely rather than paraphrasing.

It is worth recalling that 1948 falls in the immediate aftermath of the Second World War, at a moment when French intellectual life was reassessing, often under the shadow of Hegel and Marx, what remained of the autonomous, self-transparent subject inherited from Descartes. Read against that background, the quotation is not simply a psychological observation about how people come to know themselves; it is a deliberate move within a specific philosophical dispute about whether a subject can ever ground itself, or whether it is constituted only in and through what lies outside it — culture, others, language, history.

This essay follows that dispute through four steps that mirror the sentence’s own structure:

• how cultural norms mediate self-perception (I); • how language, more precisely, performs that mediation (II); • what becomes of the idea of a “pure self” once this mediation is taken seriously (III); • and in what sense such a mediated self can nonetheless be called free (IV).

I. Cultural norms as the first mediation — and where the argument for it needs correcting

Merleau-Ponty’s claim that self-contact “always” passes through culture invites a first hypothesis: that identity is not generated by the individual in isolation but assembled out of norms received from the surrounding society — unwritten codes of conduct, expectation, and value that function less like explicit rules than like an atmosphere, present everywhere and rarely noticed as such precisely because it conditions perception before it becomes an object of perception.

This is not, for Merleau-Ponty, a purely private matter. In Sens et non-sens (Sense and Non-Sense), the companion volume to Talks published the same year, he writes that religion belongs to culture “not as dogma, not even as belief, but as a cry” — a formula written in the specific context of a France taking stock of the war years, where inherited frameworks of meaning had to be re-earned rather than simply inherited ^[4]. The remark matters here because it shows cultural mediation operating even where a person believes they are speaking most intimately and spontaneously — a cry, not a proposition.

The most influential account of how other people shape self-perception in Merleau-Ponty’s own generation is Sartre’s theory of “the look”: to be seen by another is to be turned, at least momentarily, into an object for a consciousness not one’s own, and intersubjectivity is therefore built, for Sartre, on an original structure of conflict. It is tempting to borrow this concept wholesale to explain how cultural expectation reorganises self-perception from the outside. But this is precisely where Merleau-Ponty parts ways with Sartre, and the difference is not a minor nuance: where Sartre’s ontology retains a Cartesian split between a subject who sees and an object who is seen, Merleau-Ponty argues that my body is from the start “primordially interconnected” with the bodies of others, so that intersubjectivity is not primarily a contest between two isolated consciousnesses but a shared, pre-reflective situation — what commentators call intercorporeality ^[3]. On Merleau-Ponty’s own terms, then, cultural norms do not reach me by way of a hostile external gaze that fixes me as an object; they reach me because I am, from the outset, bodily woven into a common world with others, before any explicit confrontation of gazes takes place. The correction matters for the essay’s thesis: it makes cultural mediation something closer to atmosphere and habit than to domination.

Gender offers a sharper test of the same idea, and one far better anchored in the philosophical moment of Merleau-Ponty’s own text than a general appeal to “feminist movements.” In 1949, one year after Talks, Simone de Beauvoir — herself a phenomenologist working in the same Parisian circle as Merleau-Ponty and Sartre — published The Second Sex, arguing that “one is not born, but rather becomes, a woman”: femininity, on her account, is not a biological given but a role assembled, day after day, out of cultural expectation, in a movement structurally identical to the one Merleau-Ponty describes for the self in general ^[5]. Read alongside Beauvoir, Merleau-Ponty’s sentence stops being an abstract claim about “identity” in general and becomes a specific, testable one: that even categories which feel most given — including one’s sex — are, in the strict sense, received and rehearsed rather than simply possessed.

What this section corrects, then, is a certain looseness in speaking of “cultural norms” as an external force acting on an otherwise separate self. On Merleau-Ponty’s own terms, the norms are not added on top of a self that could, in principle, do without them; they are part of the fabric out of which that self is woven in the first place. This is the first sense in which contact with oneself “always” passes through culture — not as an external imposition, but as the very medium of self-formation.

II. Language as the more precise mediation, and the limits of that mediation

Merleau-Ponty’s second, more precise clause — “at least through a language” — narrows the claim considerably. Culture is diffuse; language is a specific, structured system, and the question of how far it shapes thought has its own long and contested history, one the essay’s earlier draft treated too casually by simply listing striking examples. It is worth stating the actual state of that debate before drawing on it.

Linguistics distinguishes sharply between linguistic determinism — the strong claim that language fixes what its speakers can think — and linguistic relativity, the weaker and considerably better supported claim that the language one speaks habitually influences, without determining, patterns of attention and judgement ^[6]. Lera Boroditsky’s studies of temporal metaphor in Mandarin and English speakers, for instance, found that habitual talk of time as vertical rather than horizontal correlates with faster judgements along that axis, without preventing English speakers from grasping vertical time when primed to do so ^[7]. Similarly, because Russian obligatorily distinguishes sinij (dark blue) from goluboj (light blue) where English uses only “blue,” Russian speakers discriminate those colours measurably faster — a genuine but modest effect, not a difference in what can be perceived at all ^[8]. The Australian language Guugu Yimithirr, which favours cardinal directions over relative terms such as “left” and “right,” correlates with a demonstrably more precise habitual spatial memory in its speakers, according to Stephen Levinson’s fieldwork ^[9].

The two most quoted illustrations of this idea, however, turn out to be considerably weaker than their popular versions suggest, and a rigorous essay should say so rather than repeat them uncritically. Benjamin Lee Whorf’s claim that Hopi has no grammatical means of referring to time was subjected to a six-hundred-page rebuttal by Ekkehart Malotki in 1983, who documented extensive Hopi tense marking and time vocabulary; later relativist linguists have in turn contested some of Malotki’s own analysis, so the case remains genuinely disputed rather than settled in either direction ^[10]. And the folk claim that Inuit languages contain dozens or hundreds of words for snow was traced by the linguist Geoffrey Pullum to a chain of exaggeration with almost no basis in the actual, quite modest, root inventory of Inuit and Yupik dialects ^[11]. Both cases are worth keeping in the essay, but as cautionary examples of how easily a genuine, modest thesis — language mildly inflects habitual cognition — gets inflated into a much stronger and less defensible one.

This distinction matters directly for reading Merleau-Ponty’s sentence. If language merely determined thought outright, the “effective freedom” he goes on to claim for the self would be strictly impossible — one cannot be free within a system that leaves no play. The weaker, empirically supported thesis of linguistic relativity is, by contrast, entirely compatible with his claim: language sets a habitual grain to thought without fixing it absolutely, which is exactly the kind of mediation that leaves room for the “instrument,” rather than the prison, of language that the final clause of the quotation describes.

Local and minority languages illustrate the same point at the level of collective rather than individual identity. Meänkieli, the Tornedalian Finnish spoken along the Sweden–Finland border, was for most of the twentieth century treated as a deficient dialect and actively discouraged in schools, before its recognition as an official Swedish minority language in 2000; its history shows a linguistic community reasserting an identity that assimilationist policy had tried to dissolve, and doing so specifically through custody of the language rather than through any other cultural marker ^[12]. The case confirms, from the outside, what Merleau-Ponty argues from within: that language is not a neutral tool added to an already-formed identity but part of what constitutes the identity of a community in the first place.

As Merleau-Ponty writes in his posthumously published Le Visible et l’invisible (The Visible and the Invisible, 1964), “by breaking silence, language achieves what silence wanted and did not obtain” ^[13] — a formula that anticipates the book’s central concept of the chiasm: self and world, speaker and language, are not two separate terms joined afterwards by communication, but two sides of a single, reversible fabric, each formed through its contact with the other. This concept, developed through his readings of embodiment and touch — to touch is already, in some sense, to be touched — gives the clearest available account of why language can mediate the self without simply overriding it: mediation, on this model, is not a wall between two separate things but the very tissue connecting them ^[13].

III. The “pure self” reconsidered: not a Cartesian ideal, but a tacit cogito

Merleau-Ponty’s phrase “the pure self, the mind, without instruments and without history” names, on a first reading, an entity untouched by culture or language — and this sits awkwardly beside his claim that self-contact “always” takes place through culture, or “at least” through language. If self-perception is invariably mediated, how can there be room, even hypothetically, for a self stripped of every mediation? The earlier version of this essay left the tension unresolved, gesturing vaguely at a “non-pure self.” But Merleau-Ponty had already worked out a precise answer to this exact question three years earlier, in Phenomenology of Perception, and it is worth using it rather than improvising around it.

There, Merleau-Ponty subjects the Cartesian cogito — the “I think” taken as the one thing a subject can know with total, instrument-free certainty — to sustained criticism. Against Descartes, he argues that we have no more direct or certain access to our own inner states than to the things we perceive: willing, feeling, and thinking are available to us only in a “situated and ambiguous way,” never as pure, self-transparent data. In place of the Cartesian cogito, he proposes what he calls the tacit cogito — a cogito “in action”: we do not first achieve certainty of “I think” and then deduce “I am” from it; rather, the certainty of thinking already rests on a more basic, pre-reflective “I am” of existential engagement with the world, prior to and presupposed by explicit self-consciousness ^[2].

This reframes the apparent contradiction in the quotation rather than dissolving it by fiat. The “pure self” Merleau-Ponty names is not a lived possibility that culture and language then corrupt; it is the Cartesian fiction of a self fully present to itself, without instruments, which his whole philosophy was built to dismantle. What survives the critique is not that fiction but the tacit cogito: an ambiguous, embodied, pre-reflective self-presence that is real, but that never amounts to the total transparency the word “pure” implies. Read this way, Merleau-Ponty’s sentence is not asserting a contradiction between a pure self and its cultural mediation; it is using the idea of a pure self as a foil, precisely in order to show that no such thing survives serious phenomenological analysis — and that what remains in its place, the tacit cogito, is exactly the kind of self that can be “accomplished” rather than simply given.

Merleau-Ponty gestures toward the productive use of this non-transparency in Signes (Signs, 1960): “the ‘holy’ man is not one who has eliminated contradictions from himself: he is one who uses them and draws them into his work” ^[14]. Selfhood, on this view, is not compromised by its own opacity to itself; it is partly constituted by what it does with that opacity.

Antoine François Prévost’s Manon Lescaut (1731) offers a literary case study of exactly this dynamic. Set during the Regency, as Enlightenment currents began to unsettle a rigid class and moral order, the novel follows the Chevalier des Grieux’s ruinous love for Manon, a young woman of modest birth who embraces libertinage — a deliberate refusal of prevailing religious and moral norms. Recent scholarship on the novel reads this passionate love less as a simple moral failing than as a test case for Enlightenment questions about individual identity and autonomy: des Grieux does not step outside cultural determination altogether — he remains legible only within the very social and moral categories he defies — but he does something with those categories other than passively occupy them ^[15]. This is closer to what Merleau-Ponty’s tacit cogito allows than to any fantasy of a self escaping culture altogether: not purity, but a margin of reworking within an inheritance one never fully authored.

IV. Effective freedom: using the structures of language rather than escaping them

The final clause of the epigraph states that the self is “accomplished in effective freedom” through the “instrument” of language and by “participating in the life of the world.” Two words matter especially here: instrument, which implies something used rather than merely endured, and effective, which rules out a merely notional or inward freedom disconnected from actual engagement with others.

The rhetoric of political emancipation gives a concrete case. Martin Luther King Jr.’s 1963 “I Have a Dream” speech is often treated, loosely, as proof that words alone can “reshape collective identity.” A closer discourse-analytic reading complicates that claim usefully: King does not step outside the existing vocabulary of American civic identity — the Constitution, the Emancipation Proclamation, the promissory language of the Declaration of Independence — to invent a wholly new one; he builds new relationships and a new political identity out of that inherited vocabulary, redirecting terms the culture already possessed toward ends it had not yet honoured ^[16]. This is a more precise, and more defensible, model of “effective freedom” than the image of language as something one simply breaks free of: freedom here consists in what is done with an inherited instrument, not in discarding the instrument.

The same structure clarifies the apparent tension between freedom and linguistic constraint raised in the previous section. Every word carries connotations and associations an individual did not choose, and in that sense language always limits what can be said and, to some extent, thought. But this same structure is what makes communication — hence any shared, effective action in “the life of the world” — possible at all. Freedom, on Merleau-Ponty’s account, is not the absence of this structure but a certain skill within it: the capacity to bend inherited terms, as King does, or as poetry does through metaphor and wordplay, toward meanings the structure did not originally contain.

This is also why the concept of the chiasm introduced in Section III does real work here rather than remaining a decorative flourish. If self and language were simply two separate things, freedom would indeed require escaping one to preserve the other. But if they are, as Merleau-Ponty argues, two sides of a single reversible fabric, then acting skilfully within language is not a compromise on freedom — it is what freedom, for an embodied, historically situated self, actually consists in.

Conclusion

Returning to the sentence that opened this essay, its two moves can now be read together rather than as an unresolved tension. The descriptive claim — that contact with oneself always passes through culture, at least through language — is best understood not as domination by an external force but, following Merleau-Ponty’s own critique of Sartre and of Descartes, as the constitutive medium of a self that was never separable from its cultural and linguistic situation in the first place. The normative claim — that this same self achieves “effective freedom” through language — follows once the “pure self” is correctly identified as a Cartesian fiction rather than a lost possibility: what remains after that fiction is dismantled is not a diminished self but a tacit, embodied one, capable of reworking the very inheritance that formed it, as Beauvoir’s account of gender, Prévost’s des Grieux, and King’s civic rhetoric each show in their own register.

What this reading corrects, finally, is the temptation — present in less careful treatments of Merleau-Ponty, and in earlier drafts of this very essay — to treat culture and language as a cage from which an authentic self might one day escape. On Merleau-Ponty’s own terms, that image gets the geometry wrong. Self and world, on the model of the chiasm, are not separated by a wall that freedom might someday breach; they are woven together, reversibly, such that acting freely within language is not a lesser substitute for some purer freedom beyond it, but the only kind of freedom an embodied, historical self was ever going to have.