You are here: Home / Concepts / Metaphor/Metonymy

This project is funded by an Arts and Humanities Research Council (AHRC) research grant and is supported by the Centre for Research in Modern European Philosophy (CRMEP) and Kingston University's Faculty of Arts and Social Sciences.

AHRC logo

CRMEP logo

La métaphore/la métonymie

The related functions of metaphor and metonymy were central to Lacan’s rethinking of psychoanalysis in terms of structural linguistics. These concepts were a crucial resource for the Cahiers pour l’Analyse, especially in their investigations into structural causality and the concept of ideology.

Though metaphor has been a longstanding trope of philosophical thought dating back to Aristotle’s Poetics, the specific distinction between metaphor and metonymy put to use in the Lacanian enterprise was developed by the linguist Roman Jakobson in a 1956 article titled ‘Two Aspects of Language and Two Types of Aphasic Disturbances’. There Jakobson argued that language is comprised of two axes, a metaphorical axis and a metonymic axis. The metaphoric axis was the site of substitution, the domain wherein linguistic terms may be substituted for one another in the production of meaning. By contrast, the metonymic axis was the site of sequential ordering, that is, the domain in which signifiers concatenate to form syntactically ordered sentences or expressions. Thus, Jakobson’s distinction was consistent with Ferdinand de Saussure’s rethinking of language as split between paradigmatic relations, which, qua sites of linguistic meaning, hold in absentia, and syntagmatic relations, which eo ipso hold only in presentia.

The metaphor/metonymy distinction correlates to many other conceptual binaries that were important for Lacan and for the more specific project of the Cahiers pour l’Analyse, which was equally inspired by the tradition of French epistemology. As the site of meaning, metaphor concerned the relation of signifier to signified, that is, the process whereby the signifier comes to produce meaning in a given context. (Here some of the original, etymological sense of the Latin metaphora or Greek metaphorá is retained, as respectively the ‘carrying over’ or ‘transference’ of meaning, with the latter term having an especially pointed resonance in the psychoanalytic context). Likewise, metonymy, as the domain of sequential ordering, concerns solely the chain of signification, that is, the sequence of signifiers irrespective of the metaphorical function that generates meaning. Lacan’s appropriation of Jakobson’s distinction, combined with his engagement with Saussurean linguistics, was one of the major innovations of his thought. But it was also by and large consistent with a distinction developed in Jean Cavaillès’s essay Sur la logique et la théorie de la science (1946), a key text for the Cahiers. Through a critical analysis of logical positivism, Cavaillès developed a conceptual distinction between two logical sequences, ‘thematisation’ and ‘paradigmatisation’, wherein the former has a reflexive structure that generates meaning (like metaphor) and the latter describes the ‘actualising’ or ‘longitudinal’ process that serves as thematisation’s base but remains abstract without the reflexive movement of the ‘thematic’.1

While Cavaillès was a crucial resource for the engagement with logic in the Cahiers, it remains the case that Lacan was the primary influence. The metaphor/metonymy distinction receives its most sustained elaboration in one of Lacan’s most famous écrits, ‘The Instance of the Letter in the Unconscious, or Reason Since Freud’ (1957). Here Lacan makes use of the more generic sense of metonymy as meaning ‘the part taken for the whole’ to describe the signifying function of language in toto. The signifying chain that is language itself is essentially metonymic in that it is comprised of a sequence of letters or signifiers that ‘stands in’ for its putative reference. Lacan writes: ‘I shall designate as metonymy the first aspect of the actual field the signifier constitutes, so that meaning may assume a place there’ (E, 506). The ‘assumption’ of meaning is named metaphor:

Metaphor’s creative spark does not spring forth from the juxtaposition of two images, that is, of two equally actualized signifiers. It flashes between two signifiers, one of which has replaced the other by taking the other’s place in the signifying chain, the occulted signifier remaining present by virtue of its (metonymic) relation to the rest of the chain.

One word for another: this is the formula for metaphor... (E, 507).

What is more, ‘we see that metaphor is situated at the precise point at which meaning is produced in nonmeaning’ (E, 508). For Lacan, then, metaphor is essentially a process of condensation, the production of meaning in a discrete instance, whereas metonymy is essentially one of displacement, the process whereby meaning is always deferred or displaced within a signifying chain. In this regard, ‘the symptom is a metaphor’, as a locus of condensed meaning, and ‘desire is a metonymy’ (E, 528) as the procedural operation that displaces or defers symptoms and their ‘meaning’. Lacan ties his new frame to the Freudian edifice most pointedly in his rethinking of the Oedipus complex as the site wherein the ‘paternal metaphor’ - the phallus, functioning as the ‘Name-of-the-Father’ and inducting the child into the symbolic order - converts the originary desire for the mother into the metonymic chain of desire in language, itself a displacement without end constitutive of the subject of the unconscious.

In the Cahiers pour l’Analyse

The Lacanian conception of metaphor and metonymy is integral to Jacques-Alain Miller’s reading of Frege in ‘La Suture: Éléments d’une logique du signifiant’ (CpA 1.3). There Miller mobilizes the sense of metaphor as a ‘vertical’ condensation that establishes the displacement of the signifying chain in his assessment of the genesis of the whole number line out of the zero in Frege’s Grundlagen der Arithmetik. Miller argues that the ‘verticality’ of the movement from zero, by which ‘the 0 lack comes to be represented as 1 […], indicates a crossing, a transgression’; the successor operation installs a ‘horizontal’ sequence of numbers on the basis of this primary ‘verticality’ (CpA 46-47, trans. 31). Whereas ‘logical representation’ tends to collapse this construction, the Lacanian concepts of metaphor and metonymy are capable of articulating this construction within a logic of the signifier. The primary ‘metaphor’ of the substitution of 1 for 0 is the motor for ‘the metonymic chain of successional progression’. Miller contends that that we thus arrive at ‘the structure of repetition, as the process of the differentiation of the identical’ (46/32). He concludes:

If the series of numbers, metonymy of the zero, begins with its metaphor, if the 0 member of the series as number is only the standing-in-place suturing the absence (of the absolute zero) which moves beneath the chain according to the alternation of a representation and an exclusion - then what is there to stop us from seeing in the restored relation of the zero to the series of numbers the most elementary articulation of the subject’s relation to the signifying chain? (47/32)

The suture that simultaneously establishes and annuls the subject is legible in Frege’s discourse only through the conceptual frame of the metaphor/metonymy relation developed by Lacan.

In her development of the logic of specularisation in Lacan’s and Miller’s arguments in Volume 3, Luce Irigaray points to the predominant role of metaphor over metonymy in psychosis, as well as their inverse relation in neurosis. The psychotic ‘comes face to face with the metaphoric layering of life and death rather than living their metonymic succession, which alone is bearable’ (CpA 3.3:50, trans. 19). Hence the anxiety of the psychotic occurs at a different level to the anxiety of the neurotic, who is unable to ‘metaphorise’ since he is bound to a signifying chain in which he feels constitutively inadequate, carried forth in an endless metonymic sequence. ‘Riveted to what he has been’ the obsessional neurotic is unable to become.

In his reading of Balzac’s Sarrasine in Volume 7, Jean Reboul emphasises the metonymic nature of desire in Sarrasine’s regard for Zimbanella and its ultimate condensation in a metaphoric bond occurring in the imaginary. ‘Out of a partial and metonymic desire of the object, he bonds with the specular image of a structured being, and projected into this other little imaginary, he constitutes himself at the same instant that the other finally appears to him as constituted’ (CpA 7.5:94).

Thomas Herbert’s ‘Remarques pour une théorie générale des idéologies’ (CpA 9.5) in Volume 9 is the most ambitious attempt to take the metaphor/metonymy relation out of the strictly psychoanalytic framework into a more general theory of ideology. Herbert (the pseudonym for Michel Pêcheux) relies heavily on the theories of Althusser and Nicos Poulantzas to distinguish between empiricist ideologies that seek to correlate signifiers to ‘actual’ objects and the more tenacious ‘speculative’ ideologies that determine a field of meaning as such regardless of correlation to a putatively external reality. Taking inspiration from the relationship between (semantic) metaphor and (syntactical) metonymy in Lacanian thought, Herbert shows how metonymic relations in one domain, e.g., the economic, become metaphorically displaced into, and as a consequence establish relations with, other domains, such as the political or the ideological (CpA 9.5:85-87). For example, in capitalism, economic relations are effectively metonymical: its constitutive ‘terms’ - salary, worker, contract, boss, etc. - only make sense in their differential relationship to one another. Through the very organization of the economic field of production, however, these metonymic sequences become condensed into ‘semantèmes’, units of meaning; each term is effectively shorthand for the whole sequence. This very compression metaphorically displaces these meanings into the adjacent field of the political, wherein they constitute a ‘politico-juridical axiomatic’ whose own internal coherence blinds it to its origin.

Jacques-Alain Miller’s ‘Action de la structure’ (CpA 9.6), also in Volume 9, is one of the programmatic texts of the Cahiers enterprise. Drafted in 1964, this text sketches the lineaments of a theory of structural causality conceived as metonymic causality. Putting forth a position more fully developed in Herbert’s preceding article, Althusser himself cites Miller’s formulation in his own contribution to Reading Capital as follows: ‘The absence of the cause in the structure’s “metonymic causality” on its effects is not the fault of the exteriority of the structure with respect to its economic phenomena; on the contrary, it is the very form of the interiority of the structure, as a structure, in its effects’.2 For Miller, again metonymy is a sequence of displacement that is inaugurated through the metaphorisation of the cause qua lack that determines a given discourse or sequence. He writes:

We will therefore need to explore the space of the displacement of the determination. At once univocal, repressed and interior, withdrawn and declared, only metonymic causality might qualify it. The cause is metaphorised in a discourse, and in general in any structure for the necessary condition of the functioning of structural causality is that the subject takes the effect for the cause. Fundamental law of the action of the structure (CpA 9.6:102).

In the preamble that introduces the ‘Chemistry Dossier’ that concludes this volume of the Cahiers, the editors suggest the limitations of a metaphorical approach to science (CpA 9.11:168). This metaphorical or analogical approach results whenever one specific science serves as the basis for a general discourse on science. What is advocated here, by contrast, is a grasp [reprise] of science ‘as a whole’ which betokens a focus solely on relations, and hence, by implication, the metonymic sequences of science’s conceptual development over the metaphoric substitution at its source.

Select bibliography

  • Althusser, Louis, et. al. Lire le Capital. Paris: Maspero, 1968. Reading Capital, trans. Ben Brewster. London: New Left Books, 1970.
  • Aristotle. Poetics, trans. I. Bywater, in The Complete Works of Aristotle, ed. Jonathan Barnes. Princeton: Princeton University Press, 1984.
  • Jakobson, Roman. ‘Two aspects of language and two types of aphasic disturbances’ (1956). In Selected Writings, vol. II, Word and Language. The Hague: Mouton, 1971.
  • Lacan, Jacques. Écrits. Paris: Seuil, 1966. Écrits, trans. Bruce Fink, in collaboration with Héloïse Fink and Russell Grigg. New York: W.W. Norton, 2006.


1. Cavaillès’s use of paradigm here is different from Saussure’s, in that Cavaillès uses the concept to describe the actuality of syntax, whereas for Saussure it is the site of the established meaning. Cavaillès’s deliberately more historical approach is grounded in the semantics of Alfred Tarski.

2. Reading Capital, 188.