The distributed invention of enunciation theory
Abstract
In the second half of the 20th century a linguistic approach emerged that aimed to complement the analysis of
language structure: énonciation, centred on speakers and the act of speaking. Émile Benveniste has had his role
raised to author of the theory, despite its developing simultaneously in work by Roman Jakobson and Jacques Lacan, and later
Tzvetan Todorov, with each of whom he had professional and personal ties. Others who figure in its formulation are J. L. Austin,
Charles Bally, Leonard Bloomfield, Jacques Damourette and Édouard Pichon, Bronisɬaw Malinowski, Hendrik Pos and, to some extent,
Karl Bühler. Of particular significance is work published in 1969Pos, Hendrik
J. 1939 “Phénoménologie et
linguistique”. Revue internationale de
philosophie 1/2.354–365.
by Jean Dubois and
Michel Foucault, both of whom give enunciation a clearer and fuller treatment than is found in the 1970 paper by Benveniste
regarded as the locus classicus. The present article argues for, not sidelining Benveniste, but approaching the
invention of enunciation as dialogic – a case of distributed cognition – instead of treating it on the lines of the ‘Great Man
theory of history’ which this journal’s founder, E. F. K. Koerner, worked hard to oppose.
Publication history
1.Benveniste
When the Oxford English Dictionary defines linguistics as “The scientific study of language and its structure”, it echoes what most linguists claim. Some would prefer ‘languages and their structures’; some would be more concerned about ensuring that ‘language’ be conceived broadly enough to include the people for whom, in whom, through whom it exists, and that any structural analysis take account of them and their actions in speaking. In the second half of the 20th century, énonciation emerged as such an approach.
Conceiving of language in terms of énonciation – notoriously difficult to translate for reasons that will
be discussed below, but ‘enunciation’ will have to do – is an innovation routinely attributed to Émile Benveniste (1902–1976) (see
e.g. Lie 2020Lie, Sulgi 2020 Towards
a Political Aesthetics of Cinema: The outside of film. Trans. by Daniel Fairfax. Amsterdam: Amsterdam
University Press. (Orig. publ. as Die Aussenseite des Films: Zur
politischen Filmästhetic, Zürich &
Berlin: Diaphanes 2012.): 26; Smith 2021Smith, Russell 2021 “Enunciation”. Oxford
Research Encyclopedias: Literature ed. by John Frow. Oxford: Oxford
University Press. 
), with Benveniste (1970a) 1970a “L’appareil
formel de
l’énonciation”. Langages 17.12–18 (repr.
in Benveniste 1974: 79–88). cited as the locus classicus. In presenting enunciation
theory Angermuller (2014Angermuller, Johannes 2014 Poststructuralist
Discourse Analysis: Subjectivity in enunciative pragmatics. Houndmills, Basingstoke & New
York: Palgrave Macmillan. French
version, Analyse du discours poststructuraliste: Les voix du sujet dans le langage chez Lacan,
Althusser, Foucault, Derrida, Sollers, trans. by Rosine Inspektor & Johannes Angermuller. Limoges: Lambert-Lucas 2013 : 25) says that “Its pioneer is Émile Benveniste”, and in the
book’s French translation Benveniste is le fondateur ‘the founder’ of the theory (Angermuller 2014 [2013]: 29).11.“À l’origine, la théorie de l’énonciation est inspirée par le structuralisme. Le fondateur en est Émile Benveniste
[…]”. Most unusually, the French version of Angemuller’s book, labelled as a translation, appeared before the English original. In
the eyes of some, like Henri Meschonnic (1932–2009) and Chloé Laplantine, Benveniste was not a structuralist, principally because
of his work on enunciation. For my part I have characterised him as a ‘resistant’ structuralist (Joseph 2019 2019 “The Resistant Embrace of
Formalism in the Work of Émile Benveniste and Aurélien Sauvageot”. Form and Formalism in
Linguistics ed. by James McElvenny, 141–174. Berlin: Language
Science Press.). The present article is however not the first to point to its use by some of his contemporaries
and predecessors.
My suspicion is that in some respects this is like the conquest of Mont Blanc by Horace-Bénédict de Saussure, Ferdinand’s great-grandfather. Schoolchildren have been taught this for 200 years. It is depicted on a Swiss 20-franc note, and in Chamonix there is a statue of H.-B. de Saussure with his trusty local guide Balmat. Yet it is well documented that Michel Paccard made it to the top a year earlier, accompanied by Balmat’s brother. What mattered is that none of the others had a de in his name. It was a team effort, inasmuch as Paccard was inspired by H.-B. de Saussure’s public offer of a reward to anyone who could identify a path to the top. And H.-B. de Saussure was one of the first up. But his status made his ascent the first.
Likewise, the claim that énonciation originated with Benveniste requires qualification. Amongst those who took it up, he was the one with the chair of comparative grammar in the Collège de France – de France. Only with the aristocratic index de in the name of the man or his institution, it seems, has the event officially transpired. The preposition must reach the summit first.
Another factor is that Benveniste (1970a) 1970a “L’appareil
formel de
l’énonciation”. Langages 17.12–18 (repr.
in Benveniste 1974: 79–88). was published after its author suffered
the stroke in December 1969 which left him paralysed and unable to speak for the remaining seven years of his life. Benveniste never
made any claim to being the sole originator of enunciation, and as a rigorous scholar he might well have tempered the claims to that
effect made by others from the 1970s onward, had he been aware of them and able to address them. But he was not; and we are hampered
in our efforts to assess his role by the loss of his personal papers during the years in which he was being moved from one hospital or
care home to another. The aim of the present article is not to sideline Benveniste, but to situate him appropriately within the
dialogic process through which enuncation theory came about. I am terming it a ‘distributed invention’, by analogy with ‘distributed
cognition’, a well-established concept across the humanities and social sciences which has made some inroads into linguistics (see
Joseph 2018 2018 Language, Mind and Body: A conceptual
history. Cambridge: Cambridge University
Press.), though much remains to be done.
Normand (1986)Normand, Claudine 1986 “Les
termes de l’énonciation de Benveniste”. Histoire Épistémologie
Langage 8:2.191–206. traces the development of énonciation in
Benveniste’s work back to papers he published in 1946Benveniste, Émile 1946 “Structures
des relations de personne dans le verbe”. Bulletin de la Société de Linguistique de
Paris 43/1, fasc.
1, no 126, 1–12. (Repr.
in Benveniste 1966: 225–236.) and 1949 1949 “Le
système sublogique des prépositions en latin”. Travaux de Cercle Linguistique de Copenhague 5:
Recherches structurales, 177–184. (Repr.
in Benveniste 1966: 132–139.) , and notes in particular how his 1954 article on current trends in linguistics is oriented toward “the
problem of signification”. This is a key word in Benveniste’s evolving theory; the published English translation
below renders it as ‘meaning’, which is essentially correct, but for a nuance which I shall explain.
[L]e langage a pour fonction de “dire quelque chose”. Qu’est exactement ce “quelque chose” en vue de quoi le langage est articulé, et comment le délimiter par rapport au langage lui-même? Le problème de la signification est posé.(Benveniste 1966 [1954]: 7)
[T]he function of language (langage) is “to say something”. What exactly is this “something” in regard to which language is articulated, and how is it defined with respect to language itself? The problem of meaning is raised.(Benveniste 1971 [1954] 1971 Problems in General Linguistics, trans. by Mary Elizabeth Meek, Coral Gables. Florida: University of Miami Press.
What is said here is situated within, not langue, the language system, but langage, the
overall phenomenon, including langue, parole and the faculté de langage, the
language faculty. Signification is implicitly conceived here as lying outside the language system
(langue), whilst being the overall function, the raison d’être, of language. Signification and
enunciation occupy a ‘semantic’ realm, distinct from the ‘semiotic’ one of the language (see the editors’ introduction to Benveniste 2012 2012 Dernières
leçons: Collège de France, 1968 et 1969. Ed. by Jean-Claude Coquet & Irène Fenoglio. Paris: École
des Hautes Études en Sciences Sociales, Gallimard, Seuil. English version,
see Benveniste 2019
[2012].: 49–51). This is intricate and potentially confusing, given that the word
signification begins with sign(e), and yet is being dissociated from the
semiotic. The trouble with translating it as ‘meaning’ is that this English word is used for both the semantic and semiotic senses,
whereas Benveniste made signification into a ‘term of art’ by restricting it to le sémantique (a
noun which is itself a Benvenistean term of art).
Benveniste (ibid.) identifies the problem of signification as one of three fundamental questions faced by linguists, the others being, first, what is it that will be described under the term langue, and secondly, how will it be described? Both of these fall within what he calls the semiotic, whilst the problem of signification is semantic. The challenge is to identify and delimit meaning with relation to language, which is all the more difficult because language is itself articulated with this function in view.
The presentation of enunciation in Benveniste (1970a) 1970a “L’appareil
formel de
l’énonciation”. Langages 17.12–18 (repr.
in Benveniste 1974: 79–88). addresses a question which
it understandably provoked in the minds of certain other linguists of the time, as to whether enunciation, as use, was not what
Saussure meant by parole, speech. Without directly answering the question, Benveniste indicates how his focus is a
different one.
L’énonciation est cette mise en fonction de la langue par un acte individuel d’utilisation.
Le discours, dira-t-on, qui est produit chaque fois qu’on parle, cette manifestation de l’énonciation, n’est-ce pas simplement la “parole”? — Il faut prendre garde à la condition spécifique de l’énonciation: c’est l’acte même de produire un énoncé et non le texte de l’énoncé qui est notre objet. Cet acte est le fait du locuteur qui mobilise la langue pour son compte. La relation du locuteur à la langue détermine les caractères linguistiques de l’énonciation.(Benveniste 1974 [1970a] 1974 Problèmes de linguistique générale 2. Paris: Gallimard.
Enunciation is this putting the language to work through an individual act of use.
But it will be asked, isn’t this manifestation of enunciation simply parole, the discourse which is produced each time one speaks? – We must take care to focus on the specific condition of enunciation: it is the act itself of producing an utterance, and not the text of the utterance, that is our object. This act is the fact of the speaker who mobilises the language on his or her own behalf. The relationship of the speaker to the language determines the linguistic features of the enunciation.(My translation, as are those that follow except where a published translation is cited)
The speaker is not ‘speaker’ before the act of enunciation. With enunciation, speaker becomes both speaker and subject; the enunciation positions him or her vis-à-vis the language, and at the same time their relationship shapes the enunciation.
In presenting enunciation not as an alternative to structuralist analysis, but as a parallel track, Benveniste can be said
to fulfil a wish expressed by the Neogrammarians Karl Brugmann (1849–1919) and Hermann Osthoff (1847–1909) when they remarked that, in
the past, “Languages were indeed investigated most eagerly, but people speaking, much too little” (Osthoff & Brugmann 1878Osthoff, Hermann and Brugman(n), Karl 1878 Morphologische
Untersuchungen auf dem Gebiete der indogermanischen
Sprachen, Part 1. Leipzig: Verlag
von S. Hirzel.: iii).22.Osthoff & Brugmann (1878Osthoff, Hermann and Brugman(n), Karl 1878 Morphologische
Untersuchungen auf dem Gebiete der indogermanischen
Sprachen, Part 1. Leipzig: Verlag
von S. Hirzel.: iii): “Man erforschte zwar eifrigst die
sprachen, aber viel zu wenig den sprechenden menschen”. Still, they had in mind people not
as individual wilful subjects but as the bearers of physical speech apparatus that could account for the regularity of language
change, and mental apparatus which, through a quasi-mechanical faculty of analogy, could explain exceptions to that
regularity. The approach is however not backward-looking; on
the contrary, it prefigures future ideas on stance, voice, identity and indexicality, in addition to the direct continuations of
enunciation theory in the work of Antoine Culioli (1924–2018), Oswald Ducrot (1930–2024) and others in France, well surveyed by Angermuller (2014)Angermuller, Johannes 2014 Poststructuralist
Discourse Analysis: Subjectivity in enunciative pragmatics. Houndmills, Basingstoke & New
York: Palgrave Macmillan. French
version, Analyse du discours poststructuraliste: Les voix du sujet dans le langage chez Lacan,
Althusser, Foucault, Derrida, Sollers, trans. by Rosine Inspektor & Johannes Angermuller. Limoges: Lambert-Lucas 2013 .33.For a survey of more recent developments within enunciation theory, see Angermuller
(2023) 2023 “The
Analysis of Discursive Subjects”. Handbook of Political Discourse ed.
by Piotr Cap, 180–203. Cheltenham,
U.K.: Edward Elgar. .
Benveniste’s “question of signification” mattered enough for him to try to organise an international conference on
semantics, funded by the Rockefeller Foundation, with preparatory meetings in New York and Nice in 1950 and 1951 (transcripts of the
Nice meeting were published as Benveniste ed. 1951 ed. 1951 Actes
de la Conférence européenne de sémantique (Nice, 26–31 mars 1951), organisée par M. E. Benveniste (Paris), avec la
participation de MM. C. E. Bazell (Istanbul), G. Devoto (Florence), J. R. Firth (Londres), H. Frei (Genève), L. Hjelmslev
(Charlottenlund), J. Lotz (Columbia U. New-York), A. Sommerfelt (Oslo), S. Ullmann
(Glasgow). Paris: Société de Linguistique de
Paris.). It proved to be unworkable. The
European linguists were too far apart in their approaches for real dialogue amongst them to take place – they talked at each other,
rather than with each other – and as one of those attending wrote to Benveniste, “the anti-semantic heritage of the ‘Americans’ makes
it difficult for them to deal with semantic problems, as does their dearth of information on current and past semantic work” (Lotz 1951Lotz, John 1951 Letter
to Émile Benveniste, 9 Jan. 1951. Papiers
d’Orientalistes 29 (2), f. 22. Bibliothèque
Nationale, Paris.). The heritage was that of Leonard Bloomfield’s (1887–1949) behaviourism, which
even Bloomfield had practised less than he preached: however much he and his followers might have wanted to isolate language for
analysis, at every turn recourse to meaning was needed, for instance to determine which sounds constituted a phoneme or separate
phonemes, or where word or morpheme boundaries lay.
The insight particular to Benveniste is that the language system and the speaking person occupy different conceptual
spheres that nevertheless intersect with one another. He explores this initially, and in greatest detail, in his papers on person and
deixis.44.These include, following on from the 1946 and 1949 papers cited above, Benveniste
(1956b) 1956b “La
nature des pronoms”. For Roman
Jakobson, 34–37. The
Hague: Mouton. (Repr.
in Benveniste 1966: 251–257.) and (1958 1958 “De
la subjectivité dans la langue”. Journal de
Psychologie, juillet-sept., 257–265. (Repr.
in Benveniste 1966: 258–266.)). It is surprising that, in his review of Benveniste
(1966), Winfred P. Lehmann (1916–2007) categorised these papers as “psycholinguistics” (Lehmann 1968Lehmann, Winfred
P. 1968 Rev. of Benveniste
(1966b). Language 44:1.94–96.). Equally surprising is Lehmann’s view that “If in any of his essays Benveniste discusses linguistic
theory as such, it is in the first three, which treat the development of linguistics”. In other words, for Lehmann, what
Benveniste is doing is more than mere linguistic theory, which was a compliment from the pen of a non-Chomskyan
like Lehmann in 1968Lehmann, Winfred
P. 1968 Rev. of Benveniste
(1966b). Language 44:1.94–96.. For Benveniste the key innovative thinkers of two orders of language and
signification are Ferdinand de Saussure (1857–1913) and Charles Sanders Peirce (1839–1914) (see Benveniste 1969, 2012 2012 Dernières
leçons: Collège de France, 1968 et 1969. Ed. by Jean-Claude Coquet & Irène Fenoglio. Paris: École
des Hautes Études en Sciences Sociales, Gallimard, Seuil. English version,
see Benveniste 2019
[2012].). With Peirce, Benveniste folds in the later phenomenology of Edmund Husserl (1859–1938) and of the
Husserlian linguist Hendrik Pos (1898–1955). Saussure stands at the head of the tradition of modern linguistics in which Benveniste
himself was trained. Benveniste aligns the traditions as in Table 1.
| Peirce and phenomenology represent the order | Saussure and structural linguistics represent the order |
|---|---|
| semantic | semiotic |
| intention/intended | signifier/signified |
| énonciation ‘enunciation’ | langue ‘language system’ |
| énoncé ‘utterance’ | parole ‘speech’ |
| words and things in the world | signs and social structure |
Structural linguistics is based on the Saussurean semiotic order, which excludes consideration of writing. The new linguistics of enunciation envisioned by Benveniste would integrate the semantic with the semiotic, and one of the main aims of his last lectures is to understand how the two are bridged by writing.
Énonciation is translated as ‘enunciation’ despite the potential misunderstanding this creates with the
meaning ‘careful articulation’. In my translator’s introduction to Benveniste (2019 [2012]) 2019
[2012] Last Lectures: Collège de France, 1968 and 1969 trans.
by John
E. Joseph. Edinburgh: Edinburgh
University Press.
I followed the trail laid out by Claudine Normand (1934–2011) and Aya Ono, suggesting that énonciation began as
Benveniste’s equivalent for the English term ‘utterance’. Whilst it is tempting to translate énonciation as
‘utterance’, that English word needs to be reserved for énoncé. Énonciation denotes the
process of uttering, including the context, broadly construed so as to include speaker, hearers and referents in
the world, as well as the text uttered; énoncé is the product of uttering, the utterance as text.
The potential for confusion here is obvious: both French words can mean ‘utterance’ in its usual English sense. But
énonciation is the novel conception, a fact which justifies distancing it from normal usage by translating it as
‘enunciation’ in an abnormal sense, discomfiting as it may feel.
Ono (2007Ono, Aya 2007 La
notion d’énonciation chez Émile
Benveniste. Limoges: Lambert-Lucas.: 27–57) has shown how in Benveniste’s writings from 1945 until the
definitive formulation in Benveniste (1970a) 1970a “L’appareil
formel de
l’énonciation”. Langages 17.12–18 (repr.
in Benveniste 1974: 79–88). , the meaning of énonciation
is often ambiguous, or even indicates quite clearly what he will eventually refer to as énoncé (see also Coquet 1987Coquet, Jean-Claude 1987 “Linguistique
et sémiologie”. Actes
sémiotiques-Documents
IX
, 88.5–20.). As indicated above, Benveniste is clear that énonciation does
not align with Saussure’s parole, the text envisioned as product of the socially-shared system of
langue. We can partly equate parole with énoncé, with the nuance that
parole envisions the text semiotically, and énoncé semantically, in Benveniste’s sense of the
semantic, which incorporates reference, whereas the semiotic does not.
Indeed we find no indication in the Cours de linguistique générale that Saussure, in the linguistics of
parole which he planned to develop before his rather early death, intended for it to include the big picture of
why a speaker utters an utterance, for what pragmatic purpose, and how therefore the utterance relates to meaning beyond language –
the sorts of questions that drove Peirce, whom Benveniste places in direct contrast to Saussure. Still, Benveniste might have noted
that we do not find anything in the Cours which excludes such a possibility for the envisioned
linguistique de la parole; and that we do find a proto-pragmatics in some of Saussure’s linguistic work,
particularly his doctoral thesis on the Sanskrit genitive absolute (Saussure 1881Saussure, Ferdinand
de 1881 De l’emploi du génitif absolu en
sanscrit, thèse pour le doctorat présentée à la Faculté de Philosophie de l’Université de
Leipzig. Genève: Imprimerie Jules-Guillaume
Fick.).
The clearest statement of what Benveniste means by enunciation is the following (repeated here from a longer citation above):
Il faut prendre garde à la condition spécifique de l’énonciation: c’est l’acte même de produire un énoncé et non le texte de l’énoncé qui est notre objet. Cet acte est le fait du locuteur qui mobilise la langue pour son compte. La relation du locuteur à la langue détermine les caractères linguistiques de l’énonciation.(Benveniste 1974 [1970a] 1974 Problèmes de linguistique générale 2. Paris: Gallimard.
We need to pay attention to the specific condition of enunciation: it is the act itself of producing an utterance, and not the text of the utterance, that is our object. This act is the deed of the speaker who mobilises the language on his or her own behalf. The speaker’s relationship to the language determines the linguistic features of the enunciation.
Enunciation is the act, the doing, not the text: that much is clear. Less clear is how the act has “linguistic features” that can be distinguished from those of the text. And are those features really determined by the speaker’s relationship to the language, ‘determines’ being the strongest possible claim, as opposed to, say, ‘affects’, or ‘helps to shape’?
2.Dubois
The article in which the preceding quote appears was written by Benveniste for a special issue of the journal Langages with “L’énonciation” as its title. The preceding year, Jean Dubois (1920–2015) had published an article in the same journal entitled “Énonciation et énoncé”, laying out essentially the same distinction as in the preceding quote:
L’opposition entre l’énoncé, le texte réalisé et l’énonciation, acte de production du texte, apparaît avec les analyses de la linguistique européenne, à la convergence des études des formalistes sur les structures narratives et de celles de la stylistique pragoise et genevoise. […] L’énonciation est présentée soit comme le surgissement du sujet dans l’énoncé, soit comme la relation que le locuteur entretient par le texte avec l’interlocuteur, ou comme l’attitude du sujet parlant à l’égard de son énoncé.(Dubois 1969Dubois, Jean 1969 “Énoncé et énonciation”. Langages 13.100–110.
The opposition between the utterance, the realised text, and the enunciation, act of producing the text, appears with the analyses of European linguistics, when the studies of the formalists on narrative structures converge with those of Prague and Geneva stylistics. […] The enunciation is presented either as the emergence of the subject in the utterance, or as the relationship which the speaker entertains with the interlocutor through the text, or as the attitude of the speaker with regard to the utterance.
Dubois here is both clearer and more nuanced than Benveniste. He does not cite Benveniste, nor does Benveniste cite Dubois,
so from the published record we cannot tell whether one was relying on the other, or both drawing on other sources. Indeed the special
issue of Langages of 1970 on “L’énonciation” contains no reference to Dubois’ “Énonciation et énoncé”, just a mention
en passant, in the editor’s introduction (Todorov 1970a: 8), of the special issue on discourse analysis in which
Dubois (1969)Dubois, Jean 1969 “Énoncé
et
énonciation”. Langages 13.100–110. appeared. This despite Todorov’s discussion (on which see § 3 below) including Mikhail Bakhtin (1895–1975) and V. N. Voloshinov (1895–1936), prominent
amongst those “formalists on narrative structures” pointed to by Dubois; and despite the opening section of Todorov (1970a) having the
title “Énonciation et énoncé”, exactly that of Dubois (1969)Dubois, Jean 1969 “Énoncé
et
énonciation”. Langages 13.100–110. .
Neither Dubois’ “European linguistic […] studies of the formalists on narrative structures”, which presumably includes the Russian formalists, nor Prague stylistics get cited in later historical treatments of enunciation. These tend to centre on the term rather than what it designates, and Dubois does not say what Russian or Czech terms he has in mind. Geneva stylistics is cited by others, as we shall see. Dubois says concerning enunciation that
Dans cette conception, la notion jakobsonienne d’embrayeurs (ou de shifters) a joué, on le sait, un grand rôle: ce sont les points perceptibles de la présence du sujet parlant dans le texte; mais ces éléments appartiennent aussi à la structure linguistique, même s’ils en sont des moyens privilégiés. Le dégagement de la classe des embrayeurs (je, ici, maintenant) est une première élaboration de l’énonciation, identifiée avec le sujet qui énonce.(Dubois 1969Dubois, Jean 1969 “Énoncé et énonciation”. Langages 13.100–110.
In this conception, the Jakobsonian notion of shifters has played, as we know, a great role: these are the perceptible points of the speaker’s presence in the text; but these elements also belong to the linguistic structure, even if they are privileged means of it. Releasing the class of shifters (I, here, now) is a first elaboration of enunciation, identified with the subject who utters.
The term shifter, introduced by Otto Jespersen (1860–1943) in Jespersen
(1922)Jespersen, Otto 1922 Language:
Its nature, development and origin. London: George Allen
& Unwin; New York: Henry
Holt., will be discussed further in § 7. Dubois’s mention of the “linguistic
structure” is not neutral. He positions himself not as a structuraliste but a transformationniste,
aligining himself with Chomsky and Zellig Harris (1909–1994), who figures prominently in the introduction to this special issue (Sumpf & Dubois 1969Sumpf, Joseph & Jean Dubois 1969 “Problèmes
de l’analyse du
discours”. Langages 13.3–7. ).55.Harris originated the term ‘discourse analysis’, and his work in this area has always been much appreciated in France,
starting with Harris (1952)Harris, Zellig
S. 1952 “Discourse
Analysis”. Language 28:1.1–30. , a French translation of which is included in this special
issue. Angermuller (2014Angermuller, Johannes 2014 Poststructuralist
Discourse Analysis: Subjectivity in enunciative pragmatics. Houndmills, Basingstoke & New
York: Palgrave Macmillan. French
version, Analyse du discours poststructuraliste: Les voix du sujet dans le langage chez Lacan,
Althusser, Foucault, Derrida, Sollers, trans. by Rosine Inspektor & Johannes Angermuller. Limoges: Lambert-Lucas 2013 : 7) locates “The origins of discourse analysis in France” in
“distributionalism, inspired by the American linguist Zellig Harris, and the structuralism of Ferdinand de Saussure”. Dubois points out the limits
of the structuralist approach he associates with Jakobson:
On se heurte donc à des obstacles sur les deux voies où l’on a défini l’énonciation en linguistique structurale. Identifiée avec le procès, elle ne peut rendre compte du non-achèvement des textes, de la production continue, de l’infini des énoncés. Identifiée avec le sujet parlant, elle implique une double structure de même nature sans laquelle toute théorie du reflet est impossible; mais en ce cas on réduit le sujet d’énonciation à n’être qu’un texte que l’on peut lire entre les lignes de l’énoncé manifesté; et les communications intersubjectives deviennent des intertextualités.(Dubois 1969Dubois, Jean 1969 “Énoncé et énonciation”. Langages 13.100–110.
One thus bumps up against obstacles on the two paths by which enunciation has been defined in structural linguistics. Identified with the process, it cannot take account of the incompleteness of texts, of the continuous production, the infinity of utterances. Identified with the speaker, it implies a double structure of the same nature without which any theory of reflect is impossible; but in this case one reduces the subject of enunciation to being only a text that can be read between the lines of the manifested utterances; and intersubjective communications become intertextualities.
Jakobson’s role will be taken up in § 8 below; others whom Dubois (1969)Dubois, Jean 1969 “Énoncé
et
énonciation”. Langages 13.100–110. prominently cites are Roland Barthes (1915–1980), Luce Irigaray and Todorov, names which do not
figure in linguists’ accounts of enunciation, but then this special issue of Langages is devoted to discourse
analysis. It includes an article by Irigaray (1969)Irigaray, Luce 1969 “L’énoncé
en
analyse”. Langages 13.111–122. , as will Todorov’s special issue of the
following year (Irigaray 1970 1970 “Le
sexe fait ‘comme’
signe”. Langages 17.42–55. ).
Dubois gives considerable attention to Uriel Weinreich (1926–1967) on modalisation (Weinreich 1966Weinreich, Uriel 1966 “Explorations
in Semantic Theory”. Current Trends in Linguistics III ed.
by Thomas
A. Sebeok, 395–477. The
Hague: Mouton.), essentially what later linguists and discourse analysts will call ‘stance’.66.Dubois does not in fact specify more than “U. Weinreich”, but Weinreich
(1966)Weinreich, Uriel 1966 “Explorations
in Semantic Theory”. Current Trends in Linguistics III ed.
by Thomas
A. Sebeok, 395–477. The
Hague: Mouton. is presumbably the work referred to. Recall that the first quotation included in this section alluded to “the analyses of European linguistics”: by this
Dubois means structuralism, in contrast to the transformationism of American linguistics. Although Weinreich is mainly remembered for
his work on bilingualism and historical linguistics, he tried, like his student William Labov, to align himself with Chomsky in the
1960s, much as we see Dubois doing. It was a widespread notion in France that ‘static’ structuralism had reached a dead end, and
‘dynamic’ transformation was the way forward.
The particular transformationalist path carved out by Dubois (1969)Dubois, Jean 1969 “Énoncé
et
énonciation”. Langages 13.100–110. with regard
to enunciation merits more attention than the scope of the present paper allows; suffice it to say that it must have proved
unattractive, indeed off-putting, to others pursuing enunciation theory, or the article would not have fallen into such obscurity. The
one full-length treatment of the concept of enunciation in Benveniste’s work, Ono (2007)Ono, Aya 2007 La
notion d’énonciation chez Émile
Benveniste. Limoges: Lambert-Lucas.,
does not mention Dubois (1969)Dubois, Jean 1969 “Énoncé
et
énonciation”. Langages 13.100–110. . This is curious, since Michel Arrivé (1936–2017), who
supervised Ono’s thesis and wrote the introduction to her book, talks about Dubois’s article in Arrivé (1986)Arrivé, Michel 1986 Linguistique
et psychanalyse: Freud, Saussure, Hjelmslev, Lacan et les
autres. Paris: Méridiens-Klincksieck. English
version, see Arrivé 1992.. He says it appeared “at the same time” as Benveniste (1970a) 1970a “L’appareil
formel de
l’énonciation”. Langages 17.12–18 (repr.
in Benveniste 1974: 79–88).
(Arrivé 1992 [1986] 1992 Linguistics
and Psychoanalysis: Freud, Saussure, Hjelmslev, Lacan and others. Amsterdam &
Philadelphia: John Benjamins. (Trans.
by James Leader of Arrivé 1986.) : 95), which had been given as a conference paper in 1968, as his
presidential address to the first meeting of the International Semiotics Association in Warsaw. The word énonciation
pops up there a couple of times, in passing, without Benveniste commenting on it or saying what he means by it. Dubois (1969)Dubois, Jean 1969 “Énoncé
et
énonciation”. Langages 13.100–110. spells out its meaning clearly, in the sense subsequently attributed to Benveniste as his
creation.
3.Todorov
A major figure in literary theory, Todorov edited the special issue of Langages in which Benveniste (1970a) 1970a “L’appareil
formel de
l’énonciation”. Langages 17.12–18 (repr.
in Benveniste 1974: 79–88). appeared. Although Todorov’s introduction to the volume names a wide array
of figures whose contributions to the concept of enunciation predate Benveniste’s, he says that
[D]ans ce domaine, Benveniste a joué (et continue de jouer) un rôle de véritable défricheur, en proposant à la fois une théorie générale des signes indiciels (en particulier dans son étude sur La nature des pronoms) et les premières descriptions rigoureuses des formes françaises de plusieurs d’entre eux : le temps, la personne, les verbes de parole.(Todorov 1970Todorov, Tzvetan 1970 “Problèmes de l’énonciation”. Langages 17.3–11.
In this domain, Benveniste has played (and continues to play) a role of veritable pioneer, proposing both a general theory of indexical signs (particularly in his study on “The Nature of Pronouns” [Benveniste 1956b 1956b “La nature des pronoms”. For Roman Jakobson, 34–37. The Hague: Mouton. (Repr. in Benveniste 1966: 251–257.)
After completing a degree in philology at the University of Sofia in 1963, Todorov emigrated to Paris, where he enrolled for a doctoral degree in 1966. Like many others at the time, he was less inspired by his university studies than by Benveniste’s lectures in the Collège de France, which were open to the public, not part of any degree programme, and subject to the one restriction that lectures could not be repeated from year to year. Hence it was good fortune for him to hear Benveniste lecture on enunciation sometime between 1966 and 1968, since it was a one-time opportunity.
Not long after, Todorov was amongst the first to rediscover the Russian linguistic and literary theorists who had fallen
into obscurity in the 1930s, when the stadialism of Nikolaj Marr (1865–1934) became the quasi-official doctrine in the USSR, including
Voloshinov and Bakhtin, mentioned in § 2 above.77.
Todorov (1981) 1981 Mikhaïl
Bakhtine: Le principe dialogique, suivi de: Écrits du cercle de
Bakhtin. Paris: Seuil. introduced them to a broad French reading public, who were
intrigued by his suggestion that Bakhtin had actually written the works attributed to Voloshinov and other members of their
circle, a suggestion which has since been roundly demythified. One of the anonymous reviewers of the present paper has pointed out
Voloshinov’s and Bakhtin’s use of the term высказывание [vyskazyvaniye], corresponding to ‘utterance’, and suggests that its
echoes may have resonated for Todorov and his fellow Bulgarian émigrée Julia Kristeva when they encountered
énonciation in Benveniste’s lectures. Neither mentions this in their contributions to the published version
of Benveniste’s last lectures (Kristeva 2012Kristeva, Julia 2012 “Préface:
Émile Benveniste, un linguiste qui ne dit ni ne cache, mais
signifie”. In Benveniste (2012: 13–40)., Todorov 2012 2012 “Postface:
Émile Benveniste, le destin d’un savant”. In Benveniste (2012: 183–197).), but it is a logical suggestion which merits further research. At this
time when structuralism was at its apex, approaches to language which challenged it – or challenged the reductivist version of it that
was being so much bandied about – seemed logical candidates for aligning with one another.
4.Malinowski
Ono traces Benveniste’s use of enunciation principally to ‘utterance’ in work by Bloomfield, the Anglo-Polish anthropologist
Bronisɬaw Malinowski (1884–1942) and the Oxford philosopher J. L. Austin (1910–1960). These three do indeed figure prominently in
Benveniste’s work, with Benveniste’s own translation of a long passage from Malinowski
(1923)Malinowski, Bronisɬaw 1923 “The
Problem of Meaning in Primitive Languages”. Supplement to C. K. Ogden & I. A. Richards, The
Meaning of Meaning: A study of the influence of language upon thought and of the science of
symbolism, 451–510, London: Kegan
Paul, Trench, Trubner & Co. taking up nearly one-fifth of his most important paper on enunciation (Benveniste 1970a 1970a “L’appareil
formel de
l’énonciation”. Langages 17.12–18 (repr.
in Benveniste 1974: 79–88). ). For Bloomfield and his successors, utterance serves as a basic unit of language analysis; for Austin,
it is bound up with his theory of performativity. In my translator’s introduction to the Last Lectures, I examine how
the English word ‘utterance’ underwent a semantic shift across the 20th century. In 1900 it was most often an action noun, denoting
the act of speaking; by 2000, it almost always designated the product of this action. Malinowski, Bloomfield and Austin all use
utterance in both senses.
The long extract from Malinowski which Benveniste translates has the word utterance three times (the second time in the plural), and Benveniste renders them with three different words.
The case of language used in free, aimless, social intercourse requires special consideration. […] Language here is not dependent upon what happens at that moment, it seems to be even deprived of any context of situation. The meaning of any utterance [Benveniste: énoncé] cannot be connected with the speaker’s or hearer’s behaviour, with the purpose of what they are doing.(Malinowski 1923Malinowski, Bronisɬaw 1923 “The Problem of Meaning in Primitive Languages”. Supplement to C. K. Ogden & I. A. Richards, The Meaning of Meaning: A study of the influence of language upon thought and of the science of symbolism, 451–510, London: Kegan Paul, Trench, Trubner & Co.
[T]he situation in all such cases is created by the exchange of words, by the specific feelings which form convivial gregariousness, by the give and take of utterances [Benveniste: propos] which make up ordinary gossip. The whole situation consists in what happens linguistically. Each utterance [Benveniste: énonciation] is an act serving the direct aim of binding hearer to speaker by a tie of some social sentiment or other. Once more language appears to us in this function not as an instrument of reflection but as a mode of action.(ibid., p. 315)99. Benveniste (1974 [1970a] 1974 Problèmes de linguistique générale 2. Paris: Gallimard.
The full paper by Malinowski repeats the word utterance 35 times. Its meaning shifts tacitly from what is said, a text, to the act of speaking, and this already in the first two examples (which Benveniste does not cite):
Analysis of a savage utterance, showing the complex problems of Meaning which lead from mere linguistics into the study of culture and social psychology.(ibid., p. 296)
Difference in the linguistic perspectives which open up before the Philologist who studies dead, inscribed languages, and before the Ethnographer who has to deal with a primitive living tongue, existing only in actual utterance.(ibid.)
Then the third occurrence returns to the meaning ‘what is said’:
Let us assume further that there is an ideal interpreter at hand, who, as far as possible, can convey the meaning of each utterance, word for word, so that the listener is in possession of all the linguistic data available.(ibid., p. 300)
But the fourth, which follows directly, is ambiguous between the two meanings:
Would that make you understand the conversation or even a single utterance? Certainly not.(ibid.)
No one seems to have commented on these shifts. Benveniste and others who have translated Malinowski let them go unremarked. If the majority of examples of utterance in Malinowski suggest the translation énoncé, the overall thrust of his paper is to draw attention to énonciation. This is affirmed overtly:
What I have tried to make clear by analysis of a primitive linguistic text is that language is essentially rooted in the reality of the culture, the tribal life and customs of a people, and that it cannot be explained without constant reference to these broader contexts of verbal utterance.(ibid., p. 305)
Here the sense is of the action of producing discourse. It is this action sense that is found in dictionaries of the time – and still today, it is the first meaning not indicated as archaic in the Oxford English Dictionary. In fact, though, this meaning is archaic in current English, where utterance nearly always means what is said, the énoncé. Under utterance, the meaning “that which is uttered or expressed in words” is the last one given in the OED, followed in the 2nd edition of 1989 by the comment: “freq. from c. 1865, esp. with an and pl.”, in other words as a count noun rather than mass noun.
The date explicitly indexes the relative recentness of the semantic shift. Still, the fact that in 1923 Malinowski, in a paper with the goal of establishing the central importance of the utterance in its older sense of the act of uttering, himself uses the word more often in its newer sense of the text produced by the act of uttering, is evidence of how far the shift had advanced. It is true that English was not his mother tongue, but no traces of Polish linguistic influence are to be found in his writings.
It would hardly be surprising if a scholar as meticulous as Benveniste looked up utterance in the OED when translating the extract from Malinowski. If he did, he found a justification for his different translations of the word in his article, as well as an authoritative source for his attempt to restrict énonciation to the historical meaning of utterance as the action, rather than the text produced. The OED, which hierarchises definitions historically, gives quite early in the list of definitions of utterance “The action of uttering with the voice; vocal expression of something; speaking, speech”, with examples from the 15th to the mid 19th century, but nothing after. Later on, it gives the definition which corresponds to énoncé, what one says when speaking, and characterises it as a special use in linguistics.
5.Bloomfield
It could be that academic usage was in the forefront of this semantic evolution of utterance from the action of speaking to the text produced; certainly utterance in its more modern sense of what is said is dominant in Bloomfield, and in 1951 his student Zellig Harris (1909–1992) defines utterance as “any stretch of talk, by one person, before and after which there is silence on the part of the person” (Harris 1951: 14).
Bloomfield figures prominently in the very last of Benveniste’s Last Lectures, from 1 December 1969, a few
days before the massive stroke which left him paralysed and unable to speak for the remaining seven years of his life. Benveniste (1954) 1954 “Tendances
récentes en linguistique générale”. Journal de Psychologie Normale et
Pathologique 47e–51e années:1/2.130–145. (Repr.
in Benveniste 1966: 3–17.), on “Recent trends in general linguistics”, presents Bloomfield’s Language (1933) 1933 Language. New
York: Henry Holt &
Co.; London: George Allen &
Unwin. as the principal representative of American linguistics,
alongside Saussure (1916) 1916 Cours de linguistique
générale. Ed. by Charles Bally & Albert Sechehaye, with the collaboration of Albert Riedlinger. Paris
and Lausanne: Payot. (2nd
ed. 1922; subsequent eds essentially
unchanged). English version, Course in General
Linguistics, trans. by Wade Baskin, New
York: Philosophical
Library 1959. for European linguistics; and he brings up Bloomfield’s knowledge
of Saussure in a 1968 interview (Benveniste 1974 [1968] 1968 “Structuralisme
et linguistique” (interview with Pierre Daix). Les
lettres
françaises, no 1242 (24–30 juillet), 10–13. (Repr.
in Benveniste 1974: 11–28.): 15). It is clear from Benveniste (1954) 1954 “Tendances
récentes en linguistique générale”. Journal de Psychologie Normale et
Pathologique 47e–51e années:1/2.130–145. (Repr.
in Benveniste 1966: 3–17.) how deeply he himself knew Bloomfield’s magnum opus of
1933. In that book, and again in Bloomfield (1935) 1935 “Linguistic
Aspects of Science”. Philosophy of
Science 2.499–517. , utterance is
Bloomfield’s usual term to designate what a linguist studies. It occurs dozens of times, mostly, as with Malinowski, in the textual
sense:
In a speech-community some utterances are alike or partly alike in sound and meaning.(Bloomfield 1933 1933 Language. New York: Henry Holt & Co.; London: George Allen & Unwin.
Any utterance can be fully described in terms of lexical and grammatical forms; we must remember only that the meanings cannot be defined in terms of our science.(ibid., p. 167)
When someone says ‘I’m thirsty,’ we all say ‘He is thirsty,’ and proceed to treat him as one in need of drink. The linguist alone, when acting in his professional capacity, responds to the structure (phonetic, grammatical, lexical) of the utterance.(Bloomfield 1935 1935 “Linguistic Aspects of Science”. Philosophy of Science 2.499–517.
In any one utterance, a sentence is a word or phrase which, within the arrangement habits (grammar) of the language, is not a part, in this utterance, of any larger phrase.(ibid., p. 515)
It is true that each of these occurrences could also be interpreted as the action of uttering, but the context suggests instead the product of this action, the uttered text. In other cases, it is clearer that we are dealing with the action, especially when, as in all except the first instance below, Bloomfield speaks of utterance of a sentence, form, phrase or phoneme:
Utterances which in ordinary life we describe as consisting of ‘the same’ speech forms – say, successive utterances of the sentence I’m hungry – evidently contain some constant features of sound-wave, common to all utterances of this ‘same’ speech form.(Bloomfield 1933 1933 Language. New York: Henry Holt & Co.; London: George Allen & Unwin.
Hence a language can be replaced by a few dozen unit signals (say, visual marks), each of which replaces the utterance of a phoneme.(Bloomfield 1935 1935 “Linguistic Aspects of Science”. Philosophy of Science 2.499–517.
The utterance, in a phrase, of the word not produces a phrase such that simultaneous parallel response to both this phrase and the parallel phrase without not cannot be made.(ibid., p. 506)
Bloomfield (1933) 1933 Language. New
York: Henry Holt &
Co.; London: George Allen &
Unwin. sometimes says speech-utterance, especially
toward the start of the book, where he also says speech-event without distinguishing clearly between the two. A
speech-event would seem to be an énonciation in Benveniste’s sense, and for Bloomfield the
speech-event is precisely what language specialists are interested in. He says that, although “worthless in
itself”, it is “a means to great ends” in the form of “real or practical events, stimuli and
reactions; and it is these that can be called its ‘meaning’” (Bloomfield 1933 1933 Language. New
York: Henry Holt &
Co.; London: George Allen &
Unwin.: 26–27). He
goes on:
Accordingly, we say that speech-utterance, trivial and unimportant in itself, is important because it has a meaning; the meaning consists of the important things with which the speech-utterance (B) is connected, namely the practical events (A and C).(ibid., p. 27)
– A and C being the events which precede and follow the speech act (ibid., p. 23). Here again, interpreting (speech)-utterance involves a certain ambiguity amongst énonciation, énoncé, and a vaguer conception which does not distinguish between the two.
There is also an instance of speech-utterance in Bloomfield’s
1924Bloomfield, Leonard 1924 Rev.
of Saussure (1922 [1916]). Modern Language
Journal 8:5.317–319. review of Saussure (1922 [1916]), where he equates it with Saussure’s parole:
Actual speech-utterance, la parole, varies not only as to matters not fixed by the system (e. g., the exact phonetic character of each sound), but also as to the system itself: different speakers at times will violate almost any feature of the system.(Bloomfield 1924Bloomfield, Leonard 1924 Rev. of Saussure (1922 [1916]). Modern Language Journal 8:5.317–319.
When Benveniste translates this passage for inclusion in a 1964 1964 “Documents
pour l’histoire de quelques notions saussuriennes”. Cahiers Ferdinand de
Saussure 21.131–135 article, he turns
“Actual speech-utterance, la parole” into “L’énonciation réelle, la parole”, rendering speech-utterance as
énonciation despite Bloomfield’s having already translated it as parole, and despite
Benveniste’s own later assertions of the distinction between énonciation and parole as discussed in
§ 1 above:
L’énonciation réelle, la parole, varie non seulement en ce qui n’est pas réglé par le système (par exemple, le caractère phonétique exact de chaque son), mais également en fonction du système lui-même: des locuteurs différents enfreignent parfois presque toutes les caractéristiques du système.(Benveniste’s translation, in Benveniste 1964 1964 “Documents pour l’histoire de quelques notions saussuriennes”. Cahiers Ferdinand de Saussure 21.131–135
The absence of an article before speech-utterance does indeed suggest the mass noun, hence the action meaning.
6.Austin
The other English-language scholar who plays an important role in the history of énonciation is the Oxford
‘ordinary language’ philosopher J. L. Austin, with his conception of performative and constative utterances, which Benveniste (1963) 1963 “La
philosophie analytique et le langage”. Les études
philosophiques 18/1.3–11. (Repr.
in Benveniste 1966: 267–276.) treats at length. Unlike his discussions of Malinowski and Bloomfield,
Benveniste’s treatment of Austin is based on a paper published originally in French (Austin
1962a 1962a “Performatif-constatif”. La
philosophie analytique, Cahiers de
Royaumont, 271–304. Paris: Minuit. English
version, “Performative-Constative”, trans. by G. J. Warnock, in Philosophy
and Ordinary Language ed. by Charles
E. Caton, 22–54. Urbana: University
of Illinois Press 1963.), having been given by Austin in a colloquium at the Château du Royaumont in 1958. All of Benveniste’s citations from
it include only the word énoncé. Benveniste is more critical of Austin than might be expected, given their common aim
of uncovering the actions performed by speakers in language. When Benveniste remarks that the concept of the performative was already
essentially laid out in his own work, one detects a feeling of injustice.
[N]ous y portons d’autant plus d’intérêt que nous avions nous-même d’une manière indépendante signalé la situation linguistique particulière de ce type d’énoncé. En décrivant, il y a quelques années, les formes subjectives de l’énonciation linguistique, nous indiquions sommairement la différence entre je jure, qui est un acte, et il jure, qui n’est qu’une information.(Benveniste 1966 [1963]: 270)
[W]e are all the more interested in it because we have ourselves independently pointed out the special linguistic situation of this type of utterance. Several years ago, while describing the subjective forms of the linguistic utterance, we gave a brief indication of the difference between I swear, which is an action, and he swears which is nothing but a description of fact.(Benveniste 1971 [1966] 1966 Problèmes de linguistique générale 1. Paris: Gallimard. English version, see Benveniste 1971.
A footnote identifies the work referred to as Benveniste (1958) 1958 “De
la subjectivité dans la langue”. Journal de
Psychologie, juillet-sept., 257–265. (Repr.
in Benveniste 1966: 258–266.), which appeared
ten weeks after the colloquium in which Austin gave his paper. Priority seems to have mattered to Benveniste, who comments after his
citation of Austin (1962a) 1962a “Performatif-constatif”. La
philosophie analytique, Cahiers de
Royaumont, 271–304. Paris: Minuit. English
version, “Performative-Constative”, trans. by G. J. Warnock, in Philosophy
and Ordinary Language ed. by Charles
E. Caton, 22–54. Urbana: University
of Illinois Press 1963. that “It is regrettable that the date at which this conference
took place does not appear anywhere in the publication” (Benveniste 1971 [1966] 1966 Problèmes
de linguistique générale
1. Paris: Gallimard. English
version, see Benveniste 1971.: 309n.)
Benveniste goes on to say that Austin is wrong when he claims that “To say ‘Shut the door’, plainly enough, is every bit as
performative […] as to say ‘I order you to shut the door’” (Austin 1963 [1962a]: 25, cited by Benveniste 1966 [1963]: 269, 274). In
Benveniste’s view, “this is an illusion which is almost enough to create a very serious misunderstanding concerning the very nature of
the performative utterance” (Benveniste 1971 [1966] 1966 Problèmes
de linguistique générale
1. Paris: Gallimard. English
version, see Benveniste 1971.: 237 – the last word is
énoncé in the original). To begin with, the speech act that is accomplished in saying Shut the
door or I order you to shut the door is not the closing of the door, but the uttering of something. The
difference between them is that I order you to shut the door names the act of ordering – and that is the definition
of a performative, according to Austin himself at the start of his paper. A performative takes the form of an utterance which
accomplishes an act in naming it, which Shut the door does not do, because it does not name the act of ordering. It
is difficult to see how Austin could have rebutted Benveniste’s astute critique – perhaps he would have accepted it – but he had been
dead three years when Benveniste’s article appeared.
Benveniste uses énonciation several times in his paper on Austin, but without making a clear distinction between it and énoncé. In this discussion of felicity conditions, I give the French original in brackets:
A performative utterance [énoncé] that is not an act does not exist. It has existence only as an act of authority. Now, acts of authority are first and always utterances [énonciations] proferred by those to whom the right to utter [énoncer] them belongs. This condition of validity, related to the person making the utterance [énonçante] and to the circumstances of the utterance [énonciation], must always be considered met when one deals with the performative.(Benveniste 1971 [1966] 1966 Problèmes de linguistique générale 1. Paris: Gallimard. English version, see Benveniste 1971.
In the first and third sentences, énoncé and énonciation appear to be synonymous. In the fourth sentence, “the circumstances of the énonciation” suggests the meaning of an action which Benveniste will assign to énonciation a few years later, but here it remains ambiguous.
It is certainly possible that Benveniste, whilst not mentioning it, had read Austin in English as well. Austin’s
Philosophical Papers appeared in 1961; his best known book, How To Do Things With Words,
containing lectures given in 1955, came out posthumously (Austin 1962b 1962b How
To Do Things With Words: The William James Lectures delivered at Harvard University in
1955. Oxford: Clarendon
Press. French version, Quand dire c’est
faire trans. by Gilles Lane. Paris: Le
Seuil 1970.) in the same year as
his paper in French. Benveniste would have found uses of utterance in these English texts which are as ambiguous as
those we have seen in Malinowski and Bloomfield. Sometimes utterance clearly means the text pronounced, and not the
act of pronouncing it; this is particularly the case with its first occurrences in How To Do Things With Words:
What are we to call a sentence or an utterance of this type?* I propose to call it a performative sentence or a performative utterance, or for short, ‘a performative’.(Austin 1962a 1962a “Performatif-constatif”. La philosophie analytique, Cahiers de Royaumont, 271–304. Paris: Minuit. English version, “Performative-Constative”, trans. by G. J. Warnock, in Philosophy and Ordinary Language ed. by Charles E. Caton, 22–54. Urbana: University of Illinois Press 1963.
A footnote to the first sentence reads:
*‘Sentences’ form a class of ‘utterances’, which class is to be defined, so far as I am concerned, grammatically, though I doubt if the definition has yet been given satisfactorily. … To issue a performative utterance is, for example, to make a bet.
Other occurrences however correspond unambiguously to the action of speaking:
And we began by distinguishing a whole group of senses of ‘doing something’ which are all included together when we say, what is obvious, that to say something is in the full normal sense to do something – which includes the utterance of certain noises, the utterance of certain words in a certain construction, and the utterance of them with a certain ‘meaning’ in the favourite philosophical sense of that word, i.e. with a certain sense and a certain reference.
Similarly with Austin (1961a)Austin, J. L. 1961a Philosophical
Papers. Ed. by J. O. Urmson & G. J. Warnock. Oxford: Clarendon
Press. French version, Écrits
philosophiques, trans. by Lou Aubert & Anne-Lise Hacker, Paris: Seuil 1994., which includes the transcription of the 1956 BBC
radio broadcast in which he introduced the notion of performatives (Austin 1961b 1961b “Performative
Utterances”. Unscripted broadcast, BBC Third Channel, 24 Aug.
1956, corrected transcript first publ.
in Austin (1961a: 220–239).), and
where the first instances of utterance correspond to the énoncé rather than
énonciation:1111.The French translation of Austin (1962b) 1962b How
To Do Things With Words: The William James Lectures delivered at Harvard University in
1955. Oxford: Clarendon
Press. French version, Quand dire c’est
faire trans. by Gilles Lane. Paris: Le
Seuil 1970. is not sensitive to these
linguistic distinctions. The translation of Austin (1961a)Austin, J. L. 1961a Philosophical
Papers. Ed. by J. O. Urmson & G. J. Warnock. Oxford: Clarendon
Press. French version, Écrits
philosophiques, trans. by Lou Aubert & Anne-Lise Hacker, Paris: Seuil 1994. omits Austin (1961b) 1961b “Performative
Utterances”. Unscripted broadcast, BBC Third Channel, 24 Aug.
1956, corrected transcript first publ.
in Austin (1961a: 220–239)..
I want to discuss a kind of utterance which looks like a statement and grammatically, I suppose, would be classed as a statement, which is not nonsensical, and yet is not true or false. These are not going to be utterances which contain curious verbs like ‘could’ or ‘might’, or curious words like ‘good’, which many philosophers regard nowadays simply as danger signals. They will be perfectly straightforward utterances, with ordinary verbs in the first person singular present indicative active, and yet we shall see at once that they couldn’t possibly be true or false.(Austin 1961aAustin, J. L. 1961a Philosophical Papers. Ed. by J. O. Urmson & G. J. Warnock. Oxford: Clarendon Press. French version, Écrits philosophiques, trans. by Lou Aubert & Anne-Lise Hacker, Paris: Seuil 1994.
But the next sentence has an utterance that corresponds to énonciation rather than énoncé:
Furthermore, if a person makes an utterance of this sort we should say that he is doing something rather than merely saying something.(ibid.)
At first glance, the utterance here looks ambiguous, but what the sentence is telling us is precisely that to make a performative utterance is “doing something rather than merely saying something”, and these two ‘somethings’ are themselves different semantically: the something you say is a text, an énoncé; but the something you do is not really some ‘thing’. The something you do is to nominalise an action – hence not énoncé but énonciation.
This critique of Austin is where Benveniste first speaks in extenso about énoncés and énonciations. As Ono points out, the notion of énonciation was still in the process of becoming a theoretical problem in Benveniste’s thinking. It is noteworthy that, despite this paper being included in Benveniste (1966), the book’s index contains neither énonciation nor énoncé.
7.Lacan
Amongst French writers we find uses of énonciation going back decades. Jacques Lacan (1901–1981) invited
Benveniste to contribute the opening paper to the first issue of his journal La psychanalyse (Benveniste 1956a 1956a “Remarques
sur la fonction du langage dans la découverte freudienne”. La
psychanalyse 1.3–16. (Repr.
in Benveniste 1966: 75–87.)). The paper is on Freud, and cites Lacan’s “brilliant treatise on the function and field of
the individual act of speaking [parole] and language [langage] in psychoanalysis” (Benveniste 1971 [1966] 1966 Problèmes
de linguistique générale
1. Paris: Gallimard. English
version, see Benveniste 1971.: 67, bracketed additions by Meek in her translation).1212.Benveniste (1966 [1956a): 77): “Dans son brillant mémoire sur la fonction et le champ de la parole et du langage en
psychanalyse, le docteur Lacan dit […]”. In translating parole as “the individual act of speaking”, Meek turns it
into what Benveniste would not have called parole, but énonciation. She was
perhaps following Baskin’s practice in his translation of Saussure (1916) 1916 Cours de linguistique
générale. Ed. by Charles Bally & Albert Sechehaye, with the collaboration of Albert Riedlinger. Paris
and Lausanne: Payot. (2nd
ed. 1922; subsequent eds essentially
unchanged). English version, Course in General
Linguistics, trans. by Wade Baskin, New
York: Philosophical
Library 1959. of
translating parole as ‘speaking’. It does not refer to énonciation, though Lacan’s long paper in the same issue
does, in the neutral sense of ‘uttering’ (Lacan 1956 1956 “Fonction
et champ de la parole et du langage dans la psychanalyse: Rapport du Congrès de Rome tenu à l’Istituto di Psicologia della
Università di Roma les 26 et 27 septembre 1953”. La
psychanalyse 1.81–166. (Repr.
in Lacan 1966b: 237–322). English
version, “The Function and Field of Speech and Language in
Psychoanalysis”, in Lacan (2006
[1966b]: 197–268).). In a 1958 seminar, discussed below,
Lacan uses énonciation and énoncé in precisely Benveniste’s late sense, and then in his 1960 paper
“Subversion of the Subject”, the “subject of enunciation” occurs repeatedly (Lacan 1966a
[1960] 1966a
[1960] “Subversion du sujet et dialectique du désir dans l’inconscient
freudien”. Paper read (with the title “La dialectique”) at the Congrès réuni à Royaumont
par les soins des Colloques philosophiques internationaux, 19–23 Sep. 1960, and first published (in rewritten form) in Lacan
(1966b: 793–827). English version, “The Subversion of the Subject and the Dialectic of Desire in the Freudian
Unconscious”, in Lacan 2006
[1966b]: 671–702).).1313.Fink translates le sujet de l’énonciation as “the enunciating subject”, with a note that it “could
also be rendered as ‘subject of (the) enunciation’” (Lacan 2006 [1966b] 2006 Écrits:
The first complete edition in English trans. by Bruce Fink, in collaboration with Hélöise Fink & Russell Grigg. New
York & London: W. W. Norton & Co.: 835n.). In the
quotes which follow, Fink sometimes translates énoncé as ‘statement’, as Foucault’s translators do (see § 10), other times as the ‘enunciated’.
Arrivé (2007 2007 Preface
to Ono 2007, 9–13. : 10n.) admitted to finding this “very disturbing”, but concluded
that the term has “an entirely different sense” in Lacan and Benveniste. As is often the case when a scholar writes “entirely
different”, it sounds defensively hyperbolic. At the other extreme we find statements to the effect that “Lacan borrows the notion of
enunciation from Benveniste” (Mitelman 2015Mitelman, Myriam 2015 “Secrets
de l’énonciation”. Ironik! Le bulletin UFORCA pour l’Université Populaire
Jacques-Lacan no. 9 (sep 2015) https://www.lacan-universite.fr/wp-content/uploads/2015/09/MITELMAN.pdf: 3). Fenoglio (2017Fenoglio, Irène 2017 “Sur
la notion de ‘sujet’ chez
Benveniste”. Linx 74.175–188. http://journals.openedition.org/linx/1767. : 8) notes correctly that no study has yet been done of what Lacan may have taken from Benveniste,
comparable even to Milner’s (2002Milner, Jean-Claude 2002 Le
périple structural: Figures et paradigmes. Paris: Le
Seuil.: 76) brief examination of indications that Benveniste
read Lacan.
The subject of Lacan’s 1958 1958 “Le
désir et son interpretation”. Seminar of 12 Nov. 1958, https://ecole-lacanienne.net/wp-content/uploads/2016/04/1958.11.12.pdf (publ. version
in Lacan 2013). Seminar is desire and interpretation, and he
indicates early on that the year will be devoted to how “desire’s position is profoundly marked by, moored to, and riveted to a
certain linguistic function – that is, to a certain relationship between the subject and the signifier” (Lacan 2019 [2013] 2019 Desire
and Its Interpretation: The seminar of Jacques Lacan, Book VI, trans. by Bruce Fink, Cambridge
& Medford, Mass.: Polity.: 5).1414.“tout mon développement cette année vous le montrera, que la situation du désir est profondément marquée, arrimée,
rivée à une certaine fonction du langage, à un certain rapport du sujet au signifiant”. I have cited the published 2013 version
only where there is no difference from the original typescript (Lacan 1958 1958 “Le
désir et son interpretation”. Seminar of 12 Nov. 1958, https://ecole-lacanienne.net/wp-content/uploads/2016/04/1958.11.12.pdf (publ. version
in Lacan 2013).). “[I]n order to speak”, the subject “has to
enter into language and into its pre-existing discourse” (Lacan 1958 1958 “Le
désir et son interpretation”. Seminar of 12 Nov. 1958, https://ecole-lacanienne.net/wp-content/uploads/2016/04/1958.11.12.pdf (publ. version
in Lacan 2013).: 16).1515.“pour parler il a à entrer dans le langage et dans son discours
pré-existant”. Concerning discourse in Lacan and Benveniste, see Miller (2022)Miller, Alexander 2022 “Formation
and Development of the Concept of Discourse in Lacan and Benveniste”. Psychoanalysis and
History 24:2.151–179. .
À partir du moment où la structure de la chaîne signifiante a réalisé l’appel de l’Autre, c’est-à-dire où le procès de l’énonciation se distingue de la formule de l’énoncé et s’y superpose, la prise du sujet dans l’articulation de la parole, prise qui était d’abord innocente, devient inconsciente.(Lacan 2013 [1958] 2013 Le séminaire, livre VI: Le désir et son interprétation (1958–1959). Ed. by Jacques-Alain Miller. Paris: La Martinière. English version, see Lacan 2019.
From the moment at which the structure of the signifying chain makes us call on the Other [a réalisé l’appel de l’Autre] – in other words, at which the process of enunciation is distinguished from the formulation of the enunciated and is superimposed on it – the taking up of the subject in the articulation of speech, a taking up that was initially innocent, becomes unconsious. (2019 [2013]: 16, bracketed material in the published translation)
This follows Lacan’s introduction of la barre, his strong interpretation of the line separating signifier
from signified in the figures of the linguistic sign in Saussure (1916) 1916 Cours de linguistique
générale. Ed. by Charles Bally & Albert Sechehaye, with the collaboration of Albert Riedlinger. Paris
and Lausanne: Payot. (2nd
ed. 1922; subsequent eds essentially
unchanged). English version, Course in General
Linguistics, trans. by Wade Baskin, New
York: Philosophical
Library 1959. (see Joseph 2017Joseph, John
E. 2017 “The Arbre-Tree
Sign: Pictures and words in counterpoint in the Cours de linguistique
générale
”. Semiotica 217:1.147–171. : 158–160).
In Lacan’s fourth and fifth lectures, the enunciation/utterance distinction takes centre stage, in a way it will never do for Benveniste, even in his 1970 paper which is taken as the locus classicus. In the following passage he is talking about the future perfect:
Un repérage du temps concerne l’acte dont il va s’agir – par exemple, dans l’énoncé À telle époque, je serai devenu son mari, il s’agit du repérage de ce Je qui va se transfomer par mariage. Cependant, puisque vous l’exprimez dans le terme du futur antérieur, il y a d’autre part le point actuel d’où vous parlez, qui vous repère comme le Je de l’acte d’énonciation. Il y a donc deux sujets, deux Je. […] Nous voilà donc arrivés à l’appréhension de ces deux lignes comme représentant, l’une, ce qui se rapporte au procès de l’énonciation, l’autre, au procès de l’énoncé.(Lacan 2013 2013 Le séminaire, livre VI: Le désir et son interprétation (1958–1959). Ed. by Jacques-Alain Miller. Paris: La Martinière. English version, see Lacan 2019.
[T]here are two temporal landmarks. One of them concerns the act that will be involved – for example, in the statement “At such and such a time, I will have become her husband”; this is the landmark of the I that will be transformed by marriage. Nevertheless, since you express it in the future perfect tense, there is a second temporal landmark: the current point from which you are speaking, which situates [repère] you as the I in the act of enunciation. There are thus two subjects – that is, two Is. […] We have thus begun to grasp these two [horizontal] lines as representing what is related to the enunciation process, on the one hand, and what is related to the process of the statement, on the other. (Lacan 2019 [2013] 2019 Desire and Its Interpretation: The seminar of Jacques Lacan, Book VI, trans. by Bruce Fink, Cambridge & Medford, Mass.: Polity.
Benveniste did not attend Lacan’s seminars, which were not published at the time, but he did have access to Lacan’s 1960
paper, which has a section concerning enunciation, at least from its appearance in Écrits (Lacan 1966b 1966b Écrits. Paris: Seuil. English
version, see Lacan 2006.). After introducing the concept of signifier, which he credits to the Stoics, and in modern times
to Saussure and Jakobson, Lacan asks,
La structure du langage une fois reconnue dans l’inconscient, quelle sorte de sujet pouvons-nous lui concevoir?
On peut ici tenter, dans un souci de méthode, de partir de la définition strictement linguistique du ‘Je’ comme signifiant : où il n’est rien que le shifter ou indicatif qui dans le sujet de l’énoncé désigne le sujet en tant qu’il parle actuellement.
C’est dire qu’il désigne le sujet de l’énonciation, mais qu’il ne le signifie pas.(Lacan 1966b [1960]: 800)
Once the structure of language is recognized in the unconsious, what sort of subject can we conceive of for it?
In a concern for method, we can try to begin here with the strictly linguistic definition of I as signifier, where it is nothing but the shifter or indicative that, qua grammatical subject of the statement, designates the subject insofar as he is currently speaking.
That is to say, it designates the enunciating subject, but does not signify him.(Lacan 2006 [1966b] 1966b Écrits. Paris: Seuil. English version, see Lacan 2006.
The paper does not mention Benveniste, nor does this passage from the article deriving from Lacan’s opening seminar of 1965, published in the first issue of the new journal he launched in January 1966:
Le cas de la linguistique est plus subtil, puisqu’elle doit intégrer la différence de l’énoncé à l’énonciation, ce qui est bien l’incidence cette fois du sujet qui parle, en tant que tel, (et non pas du sujet de la science).(Lacan 1966b [1966a]: 860)
The case of linguistics is subtler as it must take into account the difference between the enunciated and enunciation, that is, the impact of the subject who speaks as such (and not of the subject of science).(2006: 730–731)
It is also worth noting that in revising Lacan (1945)Lacan, Jacques 1945 “Le
temps logique et l’assertion de certitude anticipée: Un nouveau sophisme”. Cahiers
d’art 1940–1944, 32–42. (Repr.
in Lacan 1966b: 197–213). English
version, “Logical Time and the Assertion of Anticipated
Certainty”, in Lacan (2006
[1966b]: 161–175.) for republication in
Écrits, Lacan changed “le ‘je’, tierce forme du sujet de la connaissance dans la logique” (“the
‘I’, third form of the subject of knowledge in logic”) to “le ‘je’, tierce forme du sujet de l’énonciation dans
la logique” (“the ‘I’, third form of the subject of enunciation in logic”; Lacan 1966b [1945]: 208n.; 2006 [1966b] 2006 Écrits:
The first complete edition in English trans. by Bruce Fink, in collaboration with Hélöise Fink & Russell Grigg. New
York & London: W. W. Norton & Co.: 175n.).
Lacan cites the originators of terms when he sees them as someone’s creation, but he treats enunciation and
utterance as accepted terms and concepts – just as will, eleven years later, Dubois and Foucault (see § 10 below), and Benveniste himself. Only after Benveniste has fallen into silence, after his
devastating stroke in December 1969Dubois, Jean 1969 “Énoncé
et
énonciation”. Langages 13.100–110. , do others start to attribute these concepts to him,
and I believe that, had he been aware and able, this meticulous philologist would have corrected them. It is our duty to him to do so
now – and to make clear the true importance of his contribution. As the senior linguist in France, his role in
challenging the established structural linguistics by contributing to the analysis of ‘shifters’, embrayeurs in
French (although Lacan uses the English term),1616.
Embrayeur was introduced as a translation of shifter by Nicolas Ruwet (1932–2001) in
his translation of Jakobson (1957) for a 1963 collection of Jakobson’s papers – with a significant shift, as it happens, in
connotation. In a Translator’s Note (p. 178n.) Ruwet writes: “Nous avons choisi ce terme pour traduire l’anglais
Shifter, emprunté par Jakobson à O. Jespersen, Language, pp. 123–124. […] Le mot
‘embrayeur’, qui est utilisé dans le langage technique pour traduire certains des sens de shift, shifter, nous a
paru propre à désigner ces unités du code qui ‘embrayent’ le message sur la situation” (“We have chosen this term to translate
English Shifter, borrowed by Jakobson from O. Jespersen [1922Jespersen, Otto 1922 Language:
Its nature, development and origin. London: George Allen
& Unwin; New York: Henry
Holt.: 123–124]. […] The word embrayeur, which is used in technical language to translate certain of the
meanings of shift, shifter, has seemed to us appropriate to designate these units of the code
which ‘embrayent’ (engage) the message onto the situation” – as a gear shift engages the motor of an automobile.
The OED includes gear-shift as one of the senses of shifter in North American English; but when
Jespersen (1922Jespersen, Otto 1922 Language:
Its nature, development and origin. London: George Allen
& Unwin; New York: Henry
Holt.: 123) uses shifters to designate “a class of words […]
whose meaning differs according to the situation”, this implies the OED’s more general meaning of
shifter as “A person who or thing which changes or moves something; a person who or thing which moves from
one position to another”. where the speaker-interlocutor relationship
cannot be detached from the language system, was crucial. And at the end, belatedly, he alone set out the position that linguistics
needed to pass beyond the ‘semiotic’ fixation of structuralism, not by replacing it with the ‘semantic’ outlook, but by pursuing the
much more difficult aim of doing both.
For Lacan, the subject is split at more levels than this one, in a way Benveniste does not take up. In Lacan’s Freudian schema speaker intentionality begins from the ça, the Id. Although he famously maintained that the unconscious is structured like a language, the language is something distinct. It comes to the speaker already structured via the ‘big D’ – diachrony. The big S, the signifier, lies at the end of this language trajectory, after two encounters with subject intentionality, at two different states, one emanating directly from the Id, the other mitigated through message and code, with further mitigations entering the schema as Lacan develops it.
Le Je dont il s’agit dans le cogito n’est pas simplement le Je articulé dans le discours, le Je en tant qu’il se prononce dans le discours, et que les linguistes appellent, au moins depuis quelque temps, un shifter. Le Je du cogito est un sémantème, qui n’a pas d’emploi articulable qu’en fonction du code, je veux dire en fonction, purement et simplement, du code articulable lexicalement. En revanche, le Je-shifter, comme l’expérience la plus simple le montre, ne se rapporte à rien qui puisse être défini en fonction d’autres éléments du code, donc à un sémantème, mais il est simplement défini en fonction de l’acte du message, il désigne celui qui est le support du message, c’est-à-dire quelqu’un qui varie à chaque instant. […] le Je-shifter est dès lors essentiellement distinct de ce qu’on peut appeler le sujet véritable de l’acte de parler en tant que tel […]. C’est même ce qui donne toujours au plus simple discours en Je une présomption de discours indirect. Je veux dire que ce Je pourrait très facilement être suivi dans le discours même d’une parenthèse – Je (qui parle), ou Je (dis que)’(Lacan 2013 2013 Le séminaire, livre VI: Le désir et son interprétation (1958–1959). Ed. by Jacques-Alain Miller. Paris: La Martinière. English version, see Lacan 2019.
The I involved in the cogito is not simply the I that is articulated in discourse, the I insofar as it is pronounced in discourse, and that linguists have, at least for some time now, been calling a “shifter”. The I of the cogito is a semanteme whose only articulable use depends on the code, I mean that it purely and simply depends on the lexically articulable code. On the other hand, the shifter I, as the most basic experience shows, is related to nothing that can be defined as a function of other elements of the code, thus to a semanteme, but is simply defined as a function of the messaging act. It designates he who underpins the message – namely, someone who varies from one moment to the next. […T]he shifter I is essentially distinct from what one might call the true subject of the act of speaking as such […]. It is even what gives even the simplest I discourse the appearance of indirect discourse. I mean that this I could rather easily be followed in discourse itself with a parenthesis: “I (who am speaking)” or “I (am saying that…)’(Lacan 2019 [2013] 2019 Desire and Its Interpretation: The seminar of Jacques Lacan, Book VI, trans. by Bruce Fink, Cambridge & Medford, Mass.: Polity.
Despite his and Lacan’s overlapping interests in enunciation, and his paper on the role of language in Freud, Benveniste evidently did not consider the Freudian dimension to be part of the linguistic analysis of enunciation, or at least not essential to it. In that sense, Arrivé is right about enunciation having a “different sense” in Lacan and Benveniste, and yes, “entirely different” when it comes to their overall approaches; but we should not let this mask the significant overlap in their shared concern with the role of the subject in the speech act. Benveniste’s articles on pronouns feed into Lacan’s expansive elaboration, but a further gap in our knowledge is how these articles relate to Jakobson’s work on the same topic. The three knew each other and read each other, and if they did not cite each other consistently it is perhaps a sign that they saw themselves as engaged in a shared discourse, of which none of them was singly the originator.
8.Jakobson; Bühler
In discussions of enunciation, Roman Jakobson (1896–1982) figures surprisingly seldom, given his prominence in 20th-century
linguistics. The reason is apparent enough: he did not use the term more than occasionally, though his work contributed significantly
to the concept, particularly through his role in drawing attention to the importance of shifters, as we have seen Dubois allude to in
§ 2 and Lacan in § 7. Moreover, Jakobson was a linch-pin
for structuralists scattered geographically, as is evident from his correspondence with Claude Lévi-Strauss (1908–2009) (Jakobson & Lévi-Strauss 2018Jakobson, Roman & Claude Lévi-Strauss 2018 Correspondance
1942–1982. Ed. by Emmanuelle Loyer & Patrice Maniglier. Paris: Le
Seuil.), where we see that Lacan was also part of their circle, as
was Benveniste, despite his not being an extravert like the other three (see Joseph, Laplantine
& Pinault 2020Joseph, John
E., Chloé Laplantine & George-Jean Pinault 2020 “Lettres
d’Émile Benveniste à Claude Lévi-Strauss (1948–1967)”. Histoire Épistémologie
Langage 42:1.155–181. ). Lévi-Strauss and Lacan were also personally close with Merleau-Ponty (on whom see § 9 below):
Louis Marin (1931–92) used to comment that when he was a young man in the early 1950s, he and his wife Françoise were invited to the apartment of M. and Mme Maurice Merleau-Ponty for what was then described as a ‘dîner intime’. When he and his wife arrived, he discovered that it was indeed a small dinner party: M. and Mme Merleau-Ponty, M. and Mme Lévi-Strauss, and M. and Mme Lacan. That these three were all friends indicates a certain collaboration and dialogue that was highly charged in the early period in which structuralism was gaining hold.(Silverman 1994Silverman, Hugh J. 1994 “French Structuralism and After: De Saussure, Lévi-Strauss, Barthes, Lacan, Foucault”. Twentieth-Century Continental Philosophy ed. by Richard Kearney, 390–408. London & New York: Routledge.
Jakobson’s best-known paper on the topic dates from the second half of the 1950s (Jakobson 1957Jakobson, Roman 1957 “Shifters,
Verbal Categories, and the Russian Verb”. Cambridge,
Mass.: Harvard University, Dept. of Slavic Language &
Literatures. (Repr. in Jakobson, Selected
Writings, vol. 2: Word and
language, 386–392. The
Hague: Mouton 1971; partial repr.
as “Shifters and Verbal Categories” in Jakobson 1990: 386–392.)), but already at the start of that decade he was working out its implications and bringing them to
international attention, for instance in a talk he gave to the Société Genevoise de Linguistique in 1950, as reported in the
Cahiers Ferdinand de Saussure (vol. 9, p. 6):
65me seance (29 juin 1950)
Communication de M. Roman Jakobson (de Harvard University), Les catégories verbales.
Résumé: L’analyse des différences entre un terme ordinaire comme ‘chien’ et certains termes comme ‘moi’ – que les enfants apprennent très tard à employer et dont les aphasiques perdent très tôt l’usage – révèle dans ces derniers mots un trait particulier, un renvoi à l’énoncé dont il font partie. Appliquée aux 3 éléments constitutifs d’une forme verbale: les participants au procès, le procès, les rapports entre participants et procès, cette distinction permet d’établir une division en 6 classes, chacune utilsant des moyens formels différents, qui englobe toute la conjugaison. Illustration de la théorie par des exemples tirés de la conjugaison russe.
65th meeting (29 June 1950)
Talk by Mr Roman Jakobson (Harvard University), Verbal Categories
Résumé: Analysis of the differences between an ordinary term like ‘dog’ and certain terms like ‘me’ – which children learn to use very late and aphasics lose the use of very early – reveal in these latter terms a particular feature, a turning back to the utterance of which they are a part. Applied to the 3 constitutive elements of a verbal form: the participants in the process, the process, the relations between participants and process, this distinction allows us to establish a division into 6 classes, each using different formal means, which englobe the entire conjugation. Illustration of the theory with examples drawn from Russian conjugation.
In Jakobson (1990 1990 On
Language. Ed. by Linda Waugh & Monique Monville-Burston. Cambridge,
Mass.: Harvard University Press.: 545) this is described as “an early draft of Jakobson [1957]Jakobson, Roman 1957 “Shifters,
Verbal Categories, and the Russian Verb”. Cambridge,
Mass.: Harvard University, Dept. of Slavic Language &
Literatures. (Repr. in Jakobson, Selected
Writings, vol. 2: Word and
language, 386–392. The
Hague: Mouton 1971; partial repr.
as “Shifters and Verbal Categories” in Jakobson 1990: 386–392.)”. Note that énoncé occurs in this summary, even if
énonciation does not. It does however appear in the 1957 paper, when he draws a distinction between “the narrated
event (procès de l’énoncé)” and “the speech event (procès de l’énonciation)” (Jakobson 1990 [1957]:
390). He here also references the paper which Benveniste contributed the year before to a volume marking Jakobson’s 60th birthday
(Benveniste 1956b 1956b “La
nature des pronoms”. For Roman
Jakobson, 34–37. The
Hague: Mouton. (Repr.
in Benveniste 1966: 251–257.)).
Šťastná (n. d.)Šťastná, Eva n. d. “Words Referring to Their Own Utterance: Jakobson and Benveniste on shifters”. https://www.academia.edu/5984378/_Words_referring_to_their_own_utterance_Jakobson_and_Benveniste_on_shifters_ investigates the origins of work on shifters in Benveniste and Jakobson, and in particular the possibility that Jakobson was the originator. She concludes that he was not, despite the occasions when their academic and personal paths crossed during the 1940s and 1950s, and that Benveniste should be credited with the theory. What I am suggesting is that actually neither of them alone, nor anyone else alone, is the originator – that enunciation arose dialogically, amongst a similarly-spirited group of roughly the same generation, who read each other’s work and, in addition to their occasional meetings, were part of the same network of scholars who were also personal friends.
Possibly Jakobson’s biggest conceptual contribution to the approach would come independently of the terminology. Jakobson (1960) 1960 “Linguistics
and Poetics”. Style in Language ed. by Thomas
A. Sebeok, 350–377. Cambridge,
Mass.: MIT Press., a paper he had given in 1958 at a conference on style at Indiana University,
is framed around literary works and avoids linguistic jargon for the sake of the expected audience. Read and cited across a wide range
of academic fields in addition to linguistics, it does not include ‘enunciation’ or related terms, and amongst those discussed in the
present article it cites only Lévi-Strauss, Malinowski, Peirce, Propp and Jakobson himself. But its schema of the functions of
language is grounded in how the ‘Addressor’ and the ‘Addressee’ are implicated (Jakobson
1960 1960 “Linguistics
and Poetics”. Style in Language ed. by Thomas
A. Sebeok, 350–377. Cambridge,
Mass.: MIT Press.: 66), in a similar spirit to his work on shifters and to what others were casting in terms of enunciation. Not the
least consideration is that Jakobson was addressing an Anglophone academic audience which was simultanously burgeoning in size and
weakening in its ability to read other languages, including French, the mother tongue of enunciation theory.
Lock (1992Lock, Charles 1992 “Roman
Jakobson and the Future of Linguistics” (rev. art. on Jakobson 1990) Canadian
Slavonic Papers/Revue canadienne des
Slavistes 34:3.311–318. : 312) was neither the first nor the last to charge Jakobson (1960) 1960 “Linguistics
and Poetics”. Style in Language ed. by Thomas
A. Sebeok, 350–377. Cambridge,
Mass.: MIT Press. with failing to give due credit to Karl Bühler’s (1879–1963)
Organon model (Bühler 1934Bühler, Karl 1934 Sprachtheorie:
Die Darstellungsfunktion der Sprache. Jena: Gustav
Fischer.) as a source – therein lies a potential
topic for another enquiry into distributed invention – but Bühler has also been juxtaposed with Benveniste, specifically because of
their shared focus on shifters. As Lösener (2010)Lösener, Hans 2010 “Die
Origo der Subjektivität: ich, jetzt, hier bei Bühler und
Benveniste”. Grammatik — Praxis — Geschichte: Festschrift für Wilfried
Kürschner ed. by Abraham P.
ten Cate, Reinhard Rapp, Jürg Strässler, Maurice Vliegen & Heinrich Weber, 155–165. Tübingen: Narr. has explained in the most probing of
these studies (which also include Angermuller 2013: 30; Alegría Polo 2014Alegría
Polo, Marcos
Alberto 2014 Sentido y evento: El problema de
la significación en Bühler y Benveniste. Tesis de maestría, Universidad de Buenos
Aires. http://repositorio.filo.uba.ar:8080/handle/filodigital/4241; Falco 2022Falco, Mariacristina 2022 “Saussure,
Bühler, Benveniste: Sign and enunciation”. With Saussure, Beyond Saussure: Between linguistics
and philosophy of language ed. by Marina
De Palo & Stefano Gensini, 99–111. Münster: Nodus.), Bühler proposes a ‘two-field’ theory, with Zeigwörter ‘pointing
words’ such as I, now, here, there, under, above, left, right falling into the
Zeigfeld ‘index field’, in contradistinction to Nennwörter ‘naming words’ such as tree,
hope, drink, see, green, which constitute the Symbolfeld ‘symbol field’. For Bühler, the
Zeigwörter are ‘signals’ rather than symbols, and as such form a link between human language and animal behaviour
(see Lösener 2010Lösener, Hans 2010 “Die
Origo der Subjektivität: ich, jetzt, hier bei Bühler und
Benveniste”. Grammatik — Praxis — Geschichte: Festschrift für Wilfried
Kürschner ed. by Abraham P.
ten Cate, Reinhard Rapp, Jürg Strässler, Maurice Vliegen & Heinrich Weber, 155–165. Tübingen: Narr.: 156). They are the basis of a particular dimension of Bühler’s theory
which he terms Origo, concerning the relation of speech and deixis. As Lösener sees it,
An dieser Stelle wird der Unterschied zwischen Bühlers und Benvenistes Ansatz deutlich: Sowohl für Bühler wie für Benveniste benötigt die Deixis eine Origo, aber Benveniste findet diese Origo in der Rede selbst, während Bühler sie außerhalb des Äußerungsaktes und unabhängig von ihm, als psychologisch-situative Disposition (als Ich-Jetzt-Hier-System) voraussetzt, auf das die Indikatoren lediglich verweisen. Als Signal ist das Zeigwort zwar auf die Situation, nicht aber auf die Rede bezogen, in der es steht. Die Ebene der Rede stellt bei Bühler keine eigene semantische Dimension dar, die sich weder auf die Situation noch auf das einzelne Zeichen zurückführen lässt. Sie spielt in Bühlers Erklärung der Indikatoren keine Rolle, während die Einbeziehung der Rede, als eigene aus dem Zeichenrepertoire ableitbare Realität, bei Benveniste erst den Schlüssel zum Verständnis der Indikatoren liefert.(Lösener 2010Lösener, Hans 2010 “Die Origo der Subjektivität: ich, jetzt, hier bei Bühler und Benveniste”. Grammatik — Praxis — Geschichte: Festschrift für Wilfried Kürschner ed. by Abraham P. ten Cate, Reinhard Rapp, Jürg Strässler, Maurice Vliegen & Heinrich Weber, 155–165. Tübingen: Narr.
At this point, the difference between Bühler’s and Benveniste’s approach becomes clear: for both Bühler and Benveniste, deixis requires an origo, but Benveniste finds this origo in the speaking itself, whereas Bühler assumes it outside the act of utterance and independently of it, as a psychological-situational disposition (as an I-Now-Here system) to which the indicators merely refer. As a signal, the indicator word is related to the situation, but not to the speech in which it is found. For Bühler, the level of speech does not represent a semantic dimension of its own that cannot be traced back to the situation or to the individual sign. It plays no role in Bühler’s explanation of the indicators, whereas for Benveniste, the inclusion of speaking, as a separate reality derivable from the repertoire of signs, provides the key to understanding the indicators.(My transl.)
Neither Benveniste nor any of the others discussed here make reference to Bühler’s work on the subject, but he merits mention as a significant figure who was thinking through some of the same key problems, though arriving at a solution altogether different from that of enunciation theory.
9.Pos, Damourette & Pichon, Bally, Sauvageot
Going further back, in an article on “Phenomenology and Linguistics” which appeared at the start of 1939, Pos twice uses the verb énoncer in a way consonant with Benveniste’s, and anticipates what Benveniste will say about enunciation, the subject and things in the world:
[L]e sujet linguistique, tout en se rendant compte de sa fonction, ne s’est pas scindé. Il énonce sa réalité vécue, sans l’observer en spectateur. Aussi, rien ne se perd dans ce qui est énoncé de cette réalité. D’autre part, rien n’y rentre qui dépasse le cercle fermé qui protège la conscience originaire.(Pos 1939Pos, Hendrik J. 1939 “Phénoménologie et linguistique”. Revue internationale de philosophie 1/2.354–365.
The linguistic subject, although aware of its function, is not split. It enunciates its lived reality, without observing it as a spectator. Thus, nothing is lost in what is enunciated about this reality. On the other hand, nothing comes from beyond the closed circle which protects the originary consciousness.1717.Pos does not write as clearly as might be hoped: grammatically, the object pronoun in “observing it” (l’observer) refers to the “lived reality”, but the context makes it seem as though what goes unobserved is the act of enunciating.
The article appears in a special issue which opens with a paper by Husserl, whose views Pos seems to think he is
following. It is unclear whether he actually is, though, and what Pos says about the subject enunciating its lived reality without
observing it as a spectator is just what Maurice Merleau-Ponty (1908–1961) will dismantle in Phenomenology of
Perception (Merleau-Ponty 1945Merleau-Ponty, Maurice 1945 Phénoménologie
de la
perception. Paris: Gallimard.). For Merleau-Ponty the subject is indeed
spectator to his own subjective acts, hence is ‘object’ at the same time as subject, as for instance when I touch my own hand. This
conception of subjecthood is more consonant with Benveniste’s characterisations of enunciation than is Pos’s.
Still, Benveniste’s familiarity with Pos’s work is evident from the fact that he invited Pos to give the opening paper at
the Conference on Semantics which he organised at Nice in 1951, although Pos did not finally turn up (Benveniste ed. 1951 ed. 1951 Actes
de la Conférence européenne de sémantique (Nice, 26–31 mars 1951), organisée par M. E. Benveniste (Paris), avec la
participation de MM. C. E. Bazell (Istanbul), G. Devoto (Florence), J. R. Firth (Londres), H. Frei (Genève), L. Hjelmslev
(Charlottenlund), J. Lotz (Columbia U. New-York), A. Sommerfelt (Oslo), S. Ullmann
(Glasgow). Paris: Société de Linguistique de
Paris.). The simple fact that Pos was talking about subjects enunciating lived experience would
have been significant for Benveniste, even if their views did not align precisely.
Arrivé (2007 2007 Preface
to Ono 2007, 9–13. : 10) suggests Damourette &
Pichon (1911–30)Damourette, Jacques & Édouard Pichon 1911–1930 Des
mots à la pensée: Essai de grammaire de la langue
française. 7 vols. Paris: d’Artrey. as a source, but whilst they created many neologisms, including locuteur and
allocutaire, which Lacan acknowledges (e.g. Lacan 1966b [1966a]: 809, 816), énonciation was not
one of their terms, and Arrivé himself admits that Benveniste hardly ever mentions Damourette & Pichon except to criticise them
(see Arrivé 1996 1996 “Ce
que Lacan retient de Damourette et Pichon: l’exemple de la
négation”. Langages 124.113–124. , condensed from part of Arrivé
1994 1994 Langage
et psychanalyse, linguistique et inconscient: Freud, Saussure, Pichon,
Lacan. Paris: Presses Universitaires de
France.). Pichon in turn was critical of Bally (1932)Bally, Charles 1932 Linguistique
générale et linguistique française. Paris: Ernest
Leroux., but not specifically of his
“Théorie générale de l’énonciation” (“General Theory of Enunciation”), the title of the book’s first section.1818.A very minor point: Chiss (1986Chiss, Jean-Louis 1986 “Charles
Bally: Qu’est-ce qu’une ‘théorie de l’énonciation’?”. Histoire Épistémologie
Langage 8:2.165–176. : 169) writes that in the first edition of
Bally (1932)Bally, Charles 1932 Linguistique
générale et linguistique française. Paris: Ernest
Leroux. the section is entitled “Théorie de l’énonciation”, and that this was
changed to “Théorie générale de l’énonciation” in the second edition of 1944. In fact the first edition has “Théorie générale de
l’énonciation” both on the section title page and in the Table of Contents, but the running head has been reduced to “Théorie de
l’énonciation”, creating the confusion. Bally was the first to use enunciation as a theoretical concept, an extension of his earlier ‘stylistics’
(see Chiss 1986Chiss, Jean-Louis 1986 “Charles
Bally: Qu’est-ce qu’une ‘théorie de l’énonciation’?”. Histoire Épistémologie
Langage 8:2.165–176. ). As discussed in Joseph
(2022a) 2022a “The Affective, the Conceptual
and the Meaning of ‘Life’ in the Stylistics of Charles Bally”. Language &
Communication 86.60–69. , Bally’s stylistics was about the ‘affective’ dimension of a language, the elements which tie it directly to the
speaker, considerable in extent but still only a subset of the language system. With enunciation, Bally wanted to capture how a
language enunciates thought in general, as well as the relationship between speaker and hearer.
How a language enunciates, rather than how a speaker does. This is how Bally generally puts it: he does not separate languages from their speakers. Bally’s theory of enunciation is about differences in how thought is expressed in language x and language y, not between two speakers of x, just as his stylistics is not to be confused with the later stylistics that analyses the style of individual writers. Bally takes it for granted that individual speakers differ in their production of parole. How they form utterances which are clear or unclear, interesting or uninteresting to their hearers, would be something to study within the linguistique de la parole projected by Saussure.
Bally (1932Bally, Charles 1932 Linguistique
générale et linguistique française. Paris: Ernest
Leroux.: 25) specifies that the theory of enunciation is a study of
signifieds, and that “Toute énonciation de la pensée par la langue est conditionnée logiquement, psychologiquement et
linguistiquement” (“Every enunciation of thought by the language is conditioned logically, psychologically and linguistically”; ibid.,
p. 31). Enunciation is done by the language. Bally is after those structures that make languages function differently
in terms of allowing speakers to express themselves and their relations to their interlocutors, such that some languages, German for
example, are ‘speaker-oriented’, whilst others, such as French, are ‘hearer-oriented’.
Where Damourette & Pichon use language as a means of understanding psychology, Bally does the reverse. Benveniste’s conception of enunciation shares Bally’s desire to find the systematicity in the semantic; Bally however does not separate the semantic from the semiotic as Benveniste does. To reprise and continue a quote from Benveniste in § 1:
La relation du locuteur à la langue détermine les caractères linguistiques de l’énonciation. On doit l’envisager comme le fait du locuteur, qui prend la langue pour instrument, et dans les caractères linguistiques qui marquent cette relation.(Benveniste 1970a 1970a “L’appareil formel de l’énonciation”. Langages 17.12–18 (repr. in Benveniste 1974: 79–88).
The relationship of the speaker to the language determines the linguistic features of the enunciation. It must be envisaged as the fact of the speaker, who takes the language as an instrument, and in the linguistic features which mark this relationship.
For Bally, it is the linguistic features of the language that determine the relationship of the speaker to the language. The standard account of enunciation would lead us to believe that this contrasts starkly with Benveniste’s position as stated in the preceding quote, where the speaker is in control; and yet we have to take account of another statement he made in 1968:
La langue peut être envisagée à l’intérieur de la société comme un système productif: elle produit du sens […]. Elle produit aussi indéfiniment des énonciations grâce à certaines règles de transformation et d’expansion formelles; elle crée donc des formes, des schèmes de formation; elle crée des objets linguistiques qui sont introduits dans le circuit de la communication.(Benveniste 1970b 1970b “Structure de la langue et structure de la société”. Linguaggi nella società e nella tecnica (Congresso Internazionale Olivetti, Milano, 14–17 ottobre 1968), 459–460. Milano: Edizioni di Comunità. (Repr. in Benveniste 1974: 91–102.)
The language can be envisioned within the society as a productive system: it produces meaning […]. It also indefinitely produces enunciations thanks to certain formal rules of transformation and expansion; it thus creates forms, schemas of formation; it creates linguistic objects which are introduced into the circuit of communication.
It makes Benveniste’s position considerably less clear-cut when he says that the language produces enunciations – but we can hardly set this aside as a lapse without extending equal generosity to Bally.
In a similar way, Bally (1932)Bally, Charles 1932 Linguistique
générale et linguistique française. Paris: Ernest
Leroux. seems to contradict his overall position when it
introduces a distinction between modus and dictum:
Logiquement, une phrase est une représentation virtuelle actualisée par une assertion. Une représentation est virtuelle tant qu’elle n’est pas conçue comme vraie, fausse ou possible par un sujet pensant. C’est alors une simple vue de l’esprit […]. C’est l’assertion qui actualise la représentation en la localisant dans un sujet, qui devient ainsi le lieu du jugement, par le fait qu’il pose la représentation comme vraie, fausse ou possible […]. La représentation ainsi actualisée peut être appelée dictum et l’assertion qui l’actualise modus.(Bally 1932Bally, Charles 1932 Linguistique générale et linguistique française. Paris: Ernest Leroux.
Logically, a sentence is a virtual representation actualised by an assertion. A representation is virtual so long as it is not conceived as true, false or possible by a thinking subject. It is thus a simple view of the mind […]. It is the assertion that actualises the representation by localising it in a subject, who thus becomes the place of the judgement, by the fact that he posits the representation as true, false or possible […]. The representation thus actualised can be called dictum and the assertions which actualises it modus.
The brief report of the discussion following Jakobson’s 1950 talk to the Société Genevoise de Linguistique (see § 8) records that Robert Godel (1902–1984) “evokes the distinction already established by Ch. Bally between ‘dictum’ and ‘modus’”.1919. Cahiers Ferdinand de Saussure 9: 6: “M. Godel évoque la distinction déjà établie par Ch. Bally entre le ‘dictum’ et le ‘modus’”. The comment is an apt one, given that there is indeed a link between Bally’s dictum and what Jakobson says about the “3 constitutive elements of a verbal form: the participants in the process, the process, the relations between participants and process”. Jakobson’s reply to Godel suggests however that he may not have seen the link; his first comment is terminological, whilst his second seems at best tangential: “Mr Jakobson prefers a different terminology: ‘narrated event’ and ‘speech event’. He signals the importance to be accorded not only to the presence or absence of a given element in a language, but also to the place of this element”.2020.Ibid.: “M. Jakobson préfère une terminologie différente : ‘narrated event’ et ‘speech event’. Il signale l’importance à accorder non seulement à la présence ou l’absence d’un élément donné dans une langue, mais encore à la place de cet élément”. The ‘narrated event’ would appear to correspond to Bally’s ‘virtual representation’, the modus, and the ‘speech event’ to the dictum, the representation actualised in a subject. Is the ‘place’ which Jakobson signals as important the same as the ‘place’ invoked by Bally in the quote from him above – i. e., the subject, the speaker? The brevity of the report makes it impossible to determine exactly what point he was trying to make.
One other linguist of this period who merits mention is Aurélien Sauvageot (1897–1988), France’s leading specialist in
Finno-Ugric languages. Between his two identically-structured books of 1946 and 1951 on Finnish and Hungarian respectively (on which
see Joseph 2022b 2022b “Making Grammars Concrete
Again: Aurélien Sauvageot’s Esquisses of Finnish and Hungarian”. The
Architecture of Grammar: Studies in linguistic historiography in honour of Pierre Swiggers ed.
by Tim Denecker, Piet Desmet, Lieve Jooken, Peter Lauwers, Toon
Van Hal & Raf
Van Rooy, 457–470. Leuven: Peeters. ), the term énonciation and related forms double
proportionally, taking into account the greater length of the second book. Sauvageot and Benveniste had both been Meillet’s students
and were active in the same circles in Paris, such as the Société de Linguistique and the Institut de Linguistique, attending the same
papers, taking part in the same discussions. I am not suggesting that Sauvageot was a source of Benveniste’s concept of
énonciation, but it is hard to imagine that the doubling of occurrences of the term in the five years separating
Sauvageot’s two books – the topics of which are not such as to account for this change – does not indicate that the term, for whatever
reason, was being talked about more than before amongst linguists in Paris, which as the site of the Sixth International Congress of
Linguists in 1948 was also a hub for its spread.
10.Foucault
Finally, how did énonciation and especially énoncé come to be so central in Michel
Foucault’s (1926–1984) L’archéologie du savoir (1969), after scarcely figuring in his earlier work? Foucault (1966)Foucault, Michel 1966 Les
mots et les choses: Une archéologie des sciences
humaines. Paris: Gallimard., in which the history of linguistics figures prominently, appeared the same
year as Benveniste’s Problèmes de linguistique générale, which contains the papers which lay the ground for
enunciation but not the later ones in which the theory is developed. Foucault (1966)Foucault, Michel 1966 Les
mots et les choses: Une archéologie des sciences
humaines. Paris: Gallimard.
contains one instance of énoncé as a noun (p. 112), in a discussion of 18th-century grammars in which the verb is the
universal énoncé of attribution, and three as past participle; also one instance of énonciation, in
a quotation from Roch-Ambroise Cucurron, l’abbé Sicard (1741–1822) (Sicard 1808Sicard, l’Abbé 1808 Élémens
de grammaire générale, appliqués à la langue
française, tome 2. 3rd
ed. Paris: chez
Deterville.: 113),
referring to the enunciation of a thought (Foucault 1966Foucault, Michel 1966 Les
mots et les choses: Une archéologie des sciences
humaines. Paris: Gallimard.: 97). None of these has the sense
associated with Benveniste.
However the énoncé and the fonction énonciative are major themes of Foucault (1969) 1969 L’archéologie
du
savoir. Paris: Gallimard. English
version, see Foucault 1972., so much so that if any single work were to be designated the locus
classicus of enunciation theory, this is the one.2121.At the time of writing, the entry for ‘énoncé’ on French Wikipédia (https://fr.wikipedia.org/wiki/Énoncé) recognises
the important contribution made by Foucault (1969) 1969 L’archéologie
du
savoir. Paris: Gallimard. English
version, see Foucault 1972., although the entry for
‘énonciation’ does not mention it. It cites no contemporary
scholarship, hence makes no mention of Benveniste, nor of any of the other names which appear in my section titles. Again, however,
the received history is that Foucault’s approach to enunciation is based on Benveniste (Lie
2020Lie, Sulgi 2020 Towards
a Political Aesthetics of Cinema: The outside of film. Trans. by Daniel Fairfax. Amsterdam: Amsterdam
University Press. (Orig. publ. as Die Aussenseite des Films: Zur
politischen Filmästhetic, Zürich &
Berlin: Diaphanes 2012.): 26; Smith 2021Smith, Russell 2021 “Enunciation”. Oxford
Research Encyclopedias: Literature ed. by John Frow. Oxford: Oxford
University Press. ). To query that is beyond my scope; but where is the
evidence? Foucault does not appear in linguists’ discussions of enunciation, and whilst Angermuller’s (2014)Angermuller, Johannes 2014 Poststructuralist
Discourse Analysis: Subjectivity in enunciative pragmatics. Houndmills, Basingstoke & New
York: Palgrave Macmillan. French
version, Analyse du discours poststructuraliste: Les voix du sujet dans le langage chez Lacan,
Althusser, Foucault, Derrida, Sollers, trans. by Rosine Inspektor & Johannes Angermuller. Limoges: Lambert-Lucas 2013 substantial account of Foucault and his contemporaries on enunciation mentions many of the names
covered in the present article (plus some others), at no point does it enquire into Benveniste’s status as
pioneer/fondateur.
Certainly what Foucault says about énonciation is perfectly consonant with Benveniste: “Two people may say
the same thing at the same time, but since there are two people there will be two distinct enunciations. The enunciation is an
unrepeatable event; it has a situated and dated uniqueness that is irreducible” (Foucault 1972
[1969] 1972 The
Archaeology of Knowledge trans. by A. M.
Sheridan Smith. London: Tavistock.: 101). This stands in direct contrast to the “repeatable materiality” of the énoncé (ibid.,
p. 102).2222.Courtine (1981Courtine, Jean-Jacques 1981 “Quelques
problèmes théoriques et méthodologiques en analyse du discours, à propos du discours communiste adressé aux
chrétiens.” Langages 62.9–128. : 45) points out that Foucault’s characterisation of the
énoncé does not align with its use by French discourse analysts in the 1970s and after. Ducrot (1984Ducrot, Oswald 1984 Le
dire et le
dit. Paris: Minuit.: 95), for instance, takes the énonciation to be “un être
linguistique abstrait” (“an abstract linguistic being”), “une invention de cette science particulière qu’est la grammaire” (“an
invention of this particular science that is grammar”), with the énoncé “considéré comme la manifestation
particulière, comme l’occurrence hic et nunc d’une phrase” (“considered as the particular manifestation, the
hic et nunc occurrence of a sentence”), hence not repeatable. The next quotation from
Foucault suggests that he might not have been bothered by the idea that énonciation is abstract; but it does seem
to be the case that Foucault’s énoncé falls under Benveniste’s semiotic, rather than his semantic. Foucault says however that
one can, by neutralizing the moment of enunciation and the coordinates that individualize it, recognize the general form of a sentence, a meaning, a proposition. The time and place of the enunciation, and the material support that it uses, then become, very largely at least, indifferent.(Foucault 1972 [1969] 1972 The Archaeology of Knowledge trans. by A. M. Sheridan Smith. London: Tavistock.
I do not imagine Benveniste would have disputed this, since he presents his semantic realm as a complementary approach to the semiotic, and Foucault is saying that a particular speech event can be analysed from both perspectives.
Foucault wrote the book during his time in Tunisia, from 1966 to the end of 1968 (Eribon 1989Eribon, Didier 1989 Michel
Foucault. Paris: Flammarion.: 204, who notes that Foucault worked mainly in the national library in Tunis). It appeared at the start of
1969. He was not attending lectures by Benveniste. The terms énonciation and énoncé were apparently
in current use. They will also appear in Foucault’s (1976 1976 La
volonté de
savoir. Paris: Gallimard. English
version.: 26) characterisation of the
modern discourse on sex, where he argues that we have paradoxically convinced ourselves of an increasing repression in the 18th and
especially 19th century, when in reality this is precisely when sexual discourse explodes into prevalence. In the course of this
argument he acknowledges that
Without question, new rules of propriety screened out some words: there was a policing of statements [énoncés]. A control over enunciations as well: where and when it was not possible to talk about such things became much more strictly defined; in which circumstances, among which speakers, and within which social relationships. Areas were thus established, if not of utter silence, at least of tact and discretion: between parents and children, for instance, or teachers and pupils, or masters and domestic servants.(Foucault 1978 [1976] 1978 The History of Sexuality, vol. 1: An introduction, trans. by Robert Hurley. New York: Pantheon.
By this time the term and concept have been well established, not least by Foucault himself in 1969 1969 L’archéologie
du
savoir. Paris: Gallimard. English
version, see Foucault 1972., so it is no surprise to see it used here very much in the sense articulated by Benveniste (1970a), with
the identity of the speakers centrally at issue.
11.Conclusion
We know less about the emergence of enunciation as a linguistic concept than the published literature would suggest. When
teaching and writing we take shortcuts which, though obviously reductivist, are useful for getting across some piece of information
that is relevant yet tangential. If talking about enunciation, I might namedrop Benveniste just to signal that I know something about
the concept beyond its definition. Over time the reduction can settle in, and in this case, even decades of work aimed at unsettling
it need to be carried on further. Previous investigations of the origin of enunciation theory have started from an assumption that
Benveniste is its main figure, and have looked either for predecessors (as per Ono 2007Ono, Aya 2007 La
notion d’énonciation chez Émile
Benveniste. Limoges: Lambert-Lucas.) or
to assess the claim of an alternative main figure (as Šťastná n.d.Šťastná, Eva n.
d. “Words Referring to Their Own Utterance: Jakobson and Benveniste on
shifters”. https://www.academia.edu/5984378/_Words_referring_to_their_own_utterance_Jakobson_and_Benveniste_on_shifters_ does with Jakobson).
I have tried to put forward a rationale for researching the question from a different starting point: to follow the evidence suggesting that enunciation theory emerged dialogically, to use a Bakhtinian term. In Joseph (in press) In press. “Koerner, Saussure, Chomsky: An eternal upbraiding of the ‘Great man theory’”. Language & History. I discuss how a driving force in the work of this journal’s founder, E. F. K. Koerner (1939–2022), starting from his PhD dissertation, was his contesting of the ‘Great Man theory of history’ – specifically his insistence that the linguistic theories of Ferdinand de Saussure were not “the creative act of a genius who was completely independent of the intellectual atmosphere of his time” (Koerner 1971Koerner, E. F. K. 1971 “Ferdinand de Saussure: Origin and development of his linguistic theory in Western studies of language: A critical evaluation of the evolution of Saussurean principles and their relevance to contemporary linguistic theories. A dissertation submitted in partial fulfilment of the requirements for the degree of Doctor of Philosophy in General Linguistics in the Department of Modern Languages, Simon Fraser University”, https://summit.sfu.ca/_flysystem/fedora/sfu_migrate/2954/b11119603.pdf: iii). As he would do with numerous other figures revered as single-handed creators, Koerner aimed to show how their giant stature has been achieved by standing upon the shoulders of less celebrated predecessors and contemporaries.
If such was indeed the case with enunciation theory, it is all the more interesting given that it rather runs against the
assumption behind its central concept, that ‘the speaker’, or ‘the speaking subject’, is where to look for what Benveniste called the
semantic dimension of the language event. The ‘subject’ has long been a problematic concept in linguistics and in structuralism
generally (see Joseph 2023 2023 “What Structuralism Is
Not”. Structuralism as One — Structuralism as Many: Studies in
structuralisms ed. by Lorenzo Cigana & Frans Gregersen, 201–224. Copenhagen: Royal
Danish Academy of Sciences and Letters.: 208–212), and there are many reasons for tempering traditional
notions of subjecthood and agency, and above all for ‘distributing’ them.
Nor should we forget that, even if enunciation theory arose in a distributed way, it still required the scientific legitimation that comes most easily through the shortcut of a ‘father’, ideally one on whom social sanctioning has already been bestowed. The more evidence I see of how enunciation coalesced over many decades, the more I think of the conquest of Mont Blanc by that other Saussure – sorry, de Saussure.
Funding
Open Access publication of this article was funded through a Transformative Agreement with University of Edinburgh.
Acknowledgements
A preliminary version of this article was given as a paper, “Benveniste and the Origins of Enunciation”, in a session devoted to Benveniste at the Annual Colloquium of the Henry Sweet Society for the History of Linguistic Ideas, the University of Edinburgh, 5–7 September 2019. I am grateful to my fellow presenters in the session, Irène Fenoglio and Chloé Laplantine, and to others whose comments and questions in the discussion which followed helped me to clarify my views. I was also helped by the comments of the two anonymous reviewers for HL, and by James McElvenny and Frank Vonk, who directed me to relevant work concerning Bühler and Benveniste.
Notes
).: iii): “Man erforschte zwar eifrigst die
sprachen, aber viel zu wenig den sprechenden menschen”. Still, they had in mind people not
as individual wilful subjects but as the bearers of physical speech apparatus that could account for the regularity of language
change, and mental apparatus which, through a quasi-mechanical faculty of analogy, could explain exceptions to that
regularity.. and (1958 1958 “De
la subjectivité dans la langue”. Journal de
Psychologie, juillet-sept., 257–265. (Repr.
in Benveniste 1966: 258–266.)). It is surprising that, in his review of Benveniste
(1966), Winfred P. Lehmann (1916–2007) categorised these papers as “psycholinguistics” (Lehmann 1968Lehmann, Winfred
P. 1968 Rev. of Benveniste
(1966b). Language 44:1.94–96.). Equally surprising is Lehmann’s view that “If in any of his essays Benveniste discusses linguistic
theory as such, it is in the first three, which treat the development of linguistics”. In other words, for Lehmann, what
Benveniste is doing is more than mere linguistic theory, which was a compliment from the pen of a non-Chomskyan
like Lehmann in 1968Lehmann, Winfred
P. 1968 Rev. of Benveniste
(1966b). Language 44:1.94–96.., a French translation of which is included in this special
issue. Angermuller (2014Angermuller, Johannes 2014 Poststructuralist
Discourse Analysis: Subjectivity in enunciative pragmatics. Houndmills, Basingstoke & New
York: Palgrave Macmillan. French
version, Analyse du discours poststructuraliste: Les voix du sujet dans le langage chez Lacan,
Althusser, Foucault, Derrida, Sollers, trans. by Rosine Inspektor & Johannes Angermuller. Limoges: Lambert-Lucas 2013 : 7) locates “The origins of discourse analysis in France” in
“distributionalism, inspired by the American linguist Zellig Harris, and the structuralism of Ferdinand de Saussure”. is presumbably the work referred to. introduced them to a broad French reading public, who were
intrigued by his suggestion that Bakhtin had actually written the works attributed to Voloshinov and other members of their
circle, a suggestion which has since been roundly demythified. One of the anonymous reviewers of the present paper has pointed out
Voloshinov’s and Bakhtin’s use of the term высказывание [vyskazyvaniye], corresponding to ‘utterance’, and suggests that its
echoes may have resonated for Todorov and his fellow Bulgarian émigrée Julia Kristeva when they encountered
énonciation in Benveniste’s lectures. Neither mentions this in their contributions to the published version
of Benveniste’s last lectures (Kristeva 2012Kristeva, Julia 2012 “Préface:
Émile Benveniste, un linguiste qui ne dit ni ne cache, mais
signifie”. In Benveniste (2012: 13–40)., Todorov 2012 2012 “Postface:
Émile Benveniste, le destin d’un savant”. In Benveniste (2012: 183–197).), but it is a logical suggestion which merits further research.: 87) translation: “Le cas du langage employé dans
des rapports sociaux libres, sans but, mérite une considération spéciale. […] Ici la langue ne dépend pas de ce qui arrive à ce
moment, elle semble même privée de tout contexte de situation. Le sens de chaque énoncé ne peut être relié avec le
comportement du locuteur ou de l’auditeur, avec l’intention de ce qu’ils font”.: 87–88): “[L]a situation en tous ces cas est créée
par l’échange de mots, par les sentiments spécifiques qui forment la grégarité conviviale, par le va-et-vient des
propos qui composent le bavardage ordinaire. La situation entière consiste en événements linguistiques. Chaque
énonciation est un acte visant directement à lier l’auditeur au locuteur par le lien de quelque sentiment, social
ou autre”. is not sensitive to these
linguistic distinctions. The translation of Austin (1961a)Austin, J. L. 1961a Philosophical
Papers. Ed. by J. O. Urmson & G. J. Warnock. Oxford: Clarendon
Press. French version, Écrits
philosophiques, trans. by Lou Aubert & Anne-Lise Hacker, Paris: Seuil 1994. omits Austin (1961b) 1961b “Performative
Utterances”. Unscripted broadcast, BBC Third Channel, 24 Aug.
1956, corrected transcript first publ.
in Austin (1961a: 220–239).. of
translating parole as ‘speaking’.: 835n.). In the
quotes which follow, Fink sometimes translates énoncé as ‘statement’, as Foucault’s translators do (see § 10), other times as the ‘enunciated’..: 123–124]. […] The word embrayeur, which is used in technical language to translate certain of the
meanings of shift, shifter, has seemed to us appropriate to designate these units of the code
which ‘embrayent’ (engage) the message onto the situation” – as a gear shift engages the motor of an automobile.
The OED includes gear-shift as one of the senses of shifter in North American English; but when
Jespersen (1922Jespersen, Otto 1922 Language:
Its nature, development and origin. London: George Allen
& Unwin; New York: Henry
Holt.: 123) uses shifters to designate “a class of words […]
whose meaning differs according to the situation”, this implies the OED’s more general meaning of
shifter as “A person who or thing which changes or moves something; a person who or thing which moves from
one position to another”.: 169) writes that in the first edition of
Bally (1932)Bally, Charles 1932 Linguistique
générale et linguistique française. Paris: Ernest
Leroux. the section is entitled “Théorie de l’énonciation”, and that this was
changed to “Théorie générale de l’énonciation” in the second edition of 1944. In fact the first edition has “Théorie générale de
l’énonciation” both on the section title page and in the Table of Contents, but the running head has been reduced to “Théorie de
l’énonciation”, creating the confusion., although the entry for
‘énonciation’ does not mention it.: 45) points out that Foucault’s characterisation of the
énoncé does not align with its use by French discourse analysts in the 1970s and after. Ducrot (1984Ducrot, Oswald 1984 Le
dire et le
dit. Paris: Minuit.: 95), for instance, takes the énonciation to be “un être
linguistique abstrait” (“an abstract linguistic being”), “une invention de cette science particulière qu’est la grammaire” (“an
invention of this particular science that is grammar”), with the énoncé “considéré comme la manifestation
particulière, comme l’occurrence hic et nunc d’une phrase” (“considered as the particular manifestation, the
hic et nunc occurrence of a sentence”), hence not repeatable. The next quotation from
Foucault suggests that he might not have been bothered by the idea that énonciation is abstract; but it does seem
to be the case that Foucault’s énoncé falls under Benveniste’s semiotic, rather than his semantic.References