Navigation – Plan du site


Geneviève Girard-Gillet et Catherine Chauvin

Texte intégral

1Is there such a thing as an “enunciative”, or utterer-centered, approach to the lexicon? This forms part of the questions that are asked in this special issue of the e-journal Lexis. The six studies collected here all raise the question of what it means for them to ground their analyses of the lexicon, or at least, of some lexical items, in an utterer-centered approach. They all propose a reflection on what the dynamic, meaning-based enunciative approach brings to the study of the aspects of the lexicon they chose to study. Some base their approach entirely on the theory; others discuss some of its implications, and may open the way for further discussion.

What’s an utterer-centered approach?

2Although no attempt will be made to define the theory in general1, a few very concise elements can be proposed in order to bring the discussion into perspective. In the first half of the twentieth century, linguists, following Saussure (1857-1913), focused their attention on “langue” (language), as “parole” (speech) was thought too unstable to be described because of its context-based nature. Benveniste (1902-1976) turned his attention to parole: as all language use is context-based, he suggested that the role of context should be placed at the core of linguistic research. He studied the system of elements that marked subjectivity: deixis, tense/aspect, in particular, and what the analysis brought into light was the key role of the situation from which someone is speaking (the “situation of uttering2”), as it was made clear it provided one with a central anchoring point around which all locations (“repérages”) evolve. The linguists who adopt the theoretical approach call themselves “enunciativists” (énonciativistes). In the wake of these seminal studies, several theories arose; a number of researchers working on various languages base their studies on the Theory of the Predicative and Enunciative Operations brought forth by A. Culioli. On English, more specifically, other approaches were devised, like the metaoperational approach of Adamczewski, or the psycho-mechanics framework, developed in particular by Joly in the wake of Guillaume’s works.

The lexicon in utterer-centered approaches

3The following aspects could perhaps be deemed typical of the approach in relation to lexical items.

  • 3 See Pinson (this volume).
  • 4 See Paillard (this volume) and Planchon (this volume).
  • 5 Mentioned in Toupin (this volume).

4Polysemy is a definite point of interest in the theory: a number of studies have focused on the varied meanings an element (lexical or grammatical, or both) can have, and how the variation can be accounted for. Variation in meaning is taken to be central: linguistic markers normally have a number of uses, and they are generally not taken to be discrete; they are constructed in a given context as a result of interaction with other elements of the phrase, or of the utterance. A difference can be made between uses, or values, which refer to contextual effects and should not be considered to have autonomous existence (even if they are recurrent), and meaning, which is taken to be a more fundamental dimension. The way in which other approaches sometimes divide the uses of an item into clear, separate “meanings” is often rejected. Meaning is first and foremost a dynamic construction. A recurrent discussion (and, also, a moot point) is what type of meaning can be attributed to a given item, if any: if there is a form of invariance3, and of what type; variation is also taken to be (more) central, so invariance should allow for variation; the recourse to a “schematic form” (forme schématique) is, for instance, one of the ways in which the question is accounted for4. The Cognitive analyses in terms of radial categories, for instance, may be criticized, in particular because they presuppose a possible discretization of a linguistic item’s uses, and also because they tend to rely rather heavily on such notions as metaphor, which, again, supposes that there are concrete and abstract meanings, that these meanings are fixed, and that one (the “abstract” one) is just a derivation from the other. As regards the possible interaction between centering one’s attention on the utterer’s presence in language use, and discussing lexical issues (which are not necessarily directly linked), a concept which could be highlighted is that of the notional domain5. The “notion” is situated at a pre-linguistic level; it is a complex representation which includes physical and cultural constructs; it can vary from one individual to another, and even from one situation to another. Because notions are associated to a notional domain, some things are considered to belong to it clearly (the “interior”), while others are not (i.e., “exterior”); the theory also allows for an in-between zone, the “border”, as some elements may be considered to have some properties associated to the notion, but not others. This can allow for online distorsion of semantic content in relation to the utterer’s communicative needs, and may open the way for subjective and intersubjective negotiation of meaning. Lexical meaning could be truly utterer-centered and utterer-based in this sense, as it cannot be analyzed outside of a given utterer’s use of it in a given context.

  • 6 See Souesme (this volume).
  • 7 See Melis (2001) (see references)
  • 8 See Videau and Hanote (this volume).

5Utterer-centered approaches do not specifically focus on the lexicon as such; in fact, the lexicon is not necessarily opposed to grammar but is believed to be very closely linked to it. A number of studies focus on linguistic items, or “markers”, which are not necessarily lexical or grammatical; they are taken to be a representative of operations and, as such, can be of varied natures. The distinction between “the lexicon” and “syntax” can therefore be taken to be irrelevant, since it is the properties of (“lexical”) items that are conducive to their syntactic behaviour; it is the semantic properties of verbs, for instance, that lead them to have certain requirements in terms of what goes well with them and what does not: transitivity is both a syntactic and a semantic issue. Other “syntactical” behaviour can be thought to be first and foremost semantic: the compatibility with articles, for instance6, being particularly linked to countability, is also and, perhaps, first and foremost a semantic problem; it is also a matter of (inter)subjective reference, as it bears heavily upon the presence and absence of common knowledge. Another possible implication of the theory is that there might be no need to draw a clear distinction between semantics and pragmatics7. If all meaning is, in fine, contextualized, there might be no need to draw a distinction between these two levels of analysis – it might even prove impossible to do so. In the same spirit, an integrated approach can also be given of morphology, phonology and the utterer’s communicative needs, too, as the use of certain phonetic patterns8 or morphological processes can be entirely subjected to their use in context. An integrated approach of “lexical” issues may therefore turn out to be inevitable.

6A number of other aspects could be mentioned, like the discussion of the nature of examples themselves (as meaning is contextual, there can be no such things as uncontextualized analysis of language), which can have an impact on the analysis of the lexicon, although they also have a more general dimension.

7All of this shows the number of possible implications an utterer-centered approach to the lexicon may have. At a time when syntactic theories tend to be more and more rooted in semantics, when Construction Grammars focus on the interaction between form and function, and Cognitive linguistics draws attention to lexical meaning and variation again, utterer-centered theories can, obviously, shed very interesting light on a number of problems.

This volume’s papers

8The papers included in this volume are based on three languages: French, English and Khmer.

9In his paper, Denis Paillard proposes that there is no clear boundary between what is grammatical and what is lexical. He focuses upon four markers in Khmer, ʔaoy, trFv, baan and daoy. These linguistic elements have a variety of uses, some of which may have been classified as uses as “verbs”, “pragmatic markers”, “connectors”, but Denis Paillard indicates that such labels are just excessive codifications of contextual usage; he also suggests that the classification springs from an ethno-centered view of language in which (generally) Western linguists try to find the categories that have been found to exist in French or English in other, less-studied languages. He says it is a mistake, and that the range of values and uses that a given marker can have should just be taken for what it is; each marker should be analyzed for its own sake. If there is no clear opposition between what is lexical and what is grammatical, there is also no use for the concept of grammaticalization, which Denis Paillard rejects.

10Philippe Planchon, who works within Culioli’s Theory of the Predicative and Enunciative Operations, and uses D. Paillard, S. de Voguë and J.-J. Franckel’s approach to lexical meaning, proposes a semantic analysis of the French verb perdre. He tries to construe a schematic form for this verb, i.e. a formal representation of the meaning of the verb. Philippe Planchon indicates that analyses in terms of mere classification (meaning/use 1, 2, 3…) should be abandoned and be replaced by an examination of how things function contextually. His data come from the Frantext corpus. He proposes to explain the diversity of examples by endorsing the distinction between discrete, dense and compact types of functioning (discret/ dense/ compact), which, he shows, allows one to account for the collected uses in a more precise way. Context, or, in this case, co-text (what perdre is used with) also plays an important role, and he proposes to specify its role by using another distinction proposed by D. Paillard, J.-J. Franckel and S. de Vogüé between three modes of constructing relationships between co-occurring elements: nouage (lit. knotting), greffe (graft) and ajout (addition). A third, syntactic level, that of “syntactic repertoire” (cf. Paillard 2000) is added to the analysis. The paper shows that the proposed framework gives a better grasp of how the interpretations of a given utterance come to be construed.

11Mathilde Pinson focuses on the different uses that one verb in English, help, can have. Help has very diverse uses: help can convey the idea of assistance but in such phrases as I can’t help doing it, help has a very different meaning; it can also mean serve, avoid, or contribute. Besides, help is used in a number of constructions (with/ without to, for instance), which are also considered in the paper. Mathilde Pinson tries to see whether help can be said to have a primary meaning (as in a sample of 100 occurrences of help found in the COCA the meaning of assistance is found 95 times, the meaning seems to be dominant), and, to explain the different uses, she combines notions such as Rosch’s prototypes and enunciative concepts such as “domaine notionnel”. She also supposes that meaning emerges contextually, and also examines the question from a synchronic point of view to see how the various meanings and/ or constructions appeared. She summarizes the different meanings in a final table and proposes to account for them with a restricted set of features: action/ shared action/ change/ euphoricity/ resolution of problem/ argument structure.

12Jean-Claude Souesme examines have + a + “N” combinations of the John had a quick look (at something) vs Mary has a strange look type. In such phrasings, have has often been analyzed as being either static or dynamic, and what is introduced by a is said to be either a common noun, or a nominalized verb. Jean-Claude Souesme calls this binary analysis into question: he asserts that meaning emerges contextually, and that the differences are a consequence of contextual interpretation. Other elements that are to be found in the utterance help one to form a given interpretation: adjectives, for instance, can guide the reader/listener, as he had a quick look tends to be eventive, vs He had a strange look, which tends to assign a property to the subject’s referent. A succession of imperatives will generally bring a process to mind: Come on! Have a smile!, but a simple past or present verb form often leads one to think in terms of properties, as in Ian doesn’t like his teacher because he has a severe look, or You need a tonic. I have a strong drink in my cupboard. The paper also includes a discussion of the typology of nominalized predicates, which sometimes are homonymous with the so-called common noun (cf. drink or look), and sometimes are not (cf. think vs. thought); the analysis allows the author to discuss the role of a (“indefinite article”?) in such constructions, too. His conclusion is that the well-established idea that have expresses the notion of possession should be rejected in favour of a more comprehensive notion, « l’état de l’ayant » (cf. Benveniste; lit. “state of having”): have attributes a quality to the subject of the sentence, and the speaker focuses on the qualitative aspect of the relation anyway, even in so-called “eventive” uses.

13In a study that spans eight centuries, as the data range from Old English to Present-day English, Fabienne Toupin studies two types of plural morphology, -s plurals or zero-plurals, that are used on nouns that refer to animals, more particularly game animals. Her main source is the OED, to which are added examples taken from five novels and non-fiction publications. The fact that she seeks to explain is that some nouns referring to animals never take the –s suffix, like dear, whereas others, like herring, can take it occasionally. Fabienne Toupin suggests that what matters most is the speakers’ viewpoints and conceptualizations of what surrounds them; her hypothesis is that the animal names form a lexical category in Guiraud’s sense, and that the morphology of these nouns is linked to the cultural status the animals have, in relation, in particular, to their edibility and individuality. She discusses how some of the concepts developed within the enunciative school can account for this phenomenon.

14In their paper entitled “Pronunciation of Prefixed Words in Speech: The Importance of Semantic and Intersubjective Parameters”, Nicolas Videau and Sylvie Hanote examine the effect that the speaker’s communicative intentions have on word stress in English. Lionel Guierre has long since proposed that the morphological structure of a word has an impact on its pronunciation in English: a separable prefix normally bears some level of stress, and an inseparable prefix tends to be unstressed. This turns out to be insufficient for conversational speech. The analysis is based on the Corpus Parole (80,000 words), an oral corpus compiled at the University of Poitiers from recordings of BBC Radio 4 for British English, and National Public Radio (NPR) for American English. The examples show that the context of use explains the prominence or lack of prominence of the prefixes. If there is explicit or implicit contrast with something that was mentioned before, the prefix may bear word stress. Another suggested factor is clarification of meaning when a coined word is used: when a new verb is coined in speech, such as demothing or unfriend, the speaker tends to make the prefix more prominent to help the co-utterer to construe the meaning. These parameters account for the variation that can be observed, and the data show that it is important to take into account the intentions of a speaker towards his/her addressee to explain the stress patterns that are to be found in a given context.

Haut de page


1 Some references are included in each of the papers and in the final selected bibliography.


3 See Pinson (this volume).

4 See Paillard (this volume) and Planchon (this volume).

5 Mentioned in Toupin (this volume).

6 See Souesme (this volume).

7 See Melis (2001) (see references)

8 See Videau and Hanote (this volume).

Haut de page

Pour citer cet article

Référence électronique

Geneviève Girard-Gillet et Catherine Chauvin, « Introduction », Lexis [En ligne], 9 | 2015, mis en ligne le 13 mai 2015, consulté le 28 mai 2017. URL :

Haut de page


Geneviève Girard-Gillet

Université Sorbonne Nouvelle – Paris 3

Articles du même auteur

Catherine Chauvin

Université de Nancy 2

Articles du même auteur

Haut de page
  • Logo Centre d’études linguistiques
  • Logo Université Jean Moulin – Lyon 3