THE THEORY OF ARGUMENTATION IN LANGUAGE AND THE EXPLANATION OF THE SENSE OF DISCOURSE

ABSTRACT

The studies on Semantics as a discipline have their mark in the publication, by Michel Bréal, article Les lois intellectuelles du langage: fragment de sémantique, in 1883. Since then, many theorists have elaborated their research to develop a semantics whose approaches explore different facets of the meaning of language. One of these surveys is the Semantic Linguistics, or Argumentative Semantics, whose principle that guides are that argument, or sense, is entered in the language. Faced with this scenario, this paper aims to present a panorama on the foundations of The Theory of Argumentation in Language, developed by Oswald Ducrot and collaborators, and discuss its main concepts and its epistemological basis. For this, we revisited the main works of the semanticist in order to offer a rereading on the concepts defended by the author, as well as to develop the conception of language from which the TAL investigates the meaning. Thus, this article is characterized as a bibliographical review and is presented as an introduction to the Theory of Argumentation in Language.

Semantics; Argumentation in Language; Epistemology

RESUMO

Os estudos acerca da Semântica enquanto disciplina têm seu marco na publicação, por Michel Bréal, artigo Les lois intellectuelles du langage: fragmente de sémantique, em 1883. Desde então, muitos teóricos têm elaborado suas pesquisas com vistas ao desenvolvimento de uma semântica cujas abordagens explorem diferentes facetas do sentido da linguagem. Uma dessas pesquisas é a Semântica Linguística, ou Semântica Argumentativa, cujo princípio que a norteia é o de que a argumentação, ou o sentido, está inscrita na língua. Diante desse cenário, este trabalho tem como objetivo apresentar um panorama sobre os fundamentos da Teoria da Argumentação na Língua, desenvolvida por Oswald Ducrot e colaboradores. Para isso, foram revisitadas as principais obras do semanticista com vistas a oferecer uma releitura dos conceitos defendidos pelo autor, bem como a desenvolver a concepção de linguagem a partir da qual a ANL investiga o sentido. Destarte, este artigo caracteriza-se como uma revisão bibliográfica cujo objetivo reside em apresentar uma introdução à Teoria da Argumentação na Língua.

Semântica; Argumentação na Língua; Introdução

Introduction

The denomination Semantics for the field of language studies that makes sense by object of study came into existence only in 1883, with the publication by Michel Bréal of the article Les lois intellectuelles du langage: fragmente de sémantique. Later, in a more developed version, Bréal publishes his Semantic Essay, guided by the idea that language finds its place in man and its study should therefore integrate the historical sciences.

Rodolfo Ilari and João Wanderley Geraldi, in their work Semantics (2006), point out that despite the designation “science” for semantic studies, there are still several difficulties in establishing boundaries and definitions in this field: “semantics is an investigation domain of movable limits.”¹ (ILARI, GERALDI, 2006, p. 06, free translation) These difficulties are shown not only in relation to the definition of concepts, which place researchers of this area in a revolving movement (for example, the definition of “Semantics” as the study of meaning, and then, the delimitation of what “Meaning” is), but also to the numerous theoretical approaches that take the study of meaning as the object of research.²

This difficulty signalized by Ilari and Geraldi (2006) was not unnoticed by Bréal (BRÉAL, 1992, p. 19, free translation) where, in the introduction that he wrote himself to his own work, he reports the difficulty of dealing with the issues related to meaning: “This book, which has been started and interrupted many times, and from which [...] I have published some parts, and in many of these resuming periods, today I decide to release it to the public. Many times, discouraged by the difficulties of my subject, I promised not to return to it!”³

Irène Tamba (2006) comments that the late delimitation of an investigative field in relation to meanings is due to the difficulty in distinguishing meaning as it is commonly perceived in everyday language use and as object of study. The author also states that, besides the denomination persists to the present day, the works developed by linguists in contemporary times have little to do with that science of meanings described and proposed by Bréal. Part of the explanation for this evidence is since the assumptions that guide language studies have changed over time, which eventually provided new hypotheses for certain linguistic phenomena used for analysis. This fact is verified in the historical course traced by the author, which divides semantic studies into four major periods: a) the period of comparative linguistics, in which historical and lexical semantics predominated; b) the structural period, characterized by lexical but synchronic semantics; c) the period of formal grammars, in which sentence and discourse-oriented semantics developed; and d) the period of the cognitive sciences, in which the study of meaning appears related to the cognitive dimension of language (TAMBA, 2006, p. 13).

The third period of the development of the studies of meaning emerged not only to the necessity of semantics linked to Chomsky’s (1957)Generative-Transformational Grammar, but also to other theorists who sought to relate semantics to logical questions of language. Along with the Speech Act Theory, first developed by John Austin (1962) and later by John Searle (1969, 1979), emerged during this period, according to the label by Émile Benveniste (1989, 2005), and Argumentative Semantics 1717, the object of studies of these pages, first developed by Jean-Claude Anscombre and Oswald Ducrot (1983), and more recently by Carel and Ducrot (2005), Ducrot and Carel (2008), and Ducrot (2016).

Argumentative Semantics, in the beginning, was also described as a Pragmatic Semantics (DUCROT, 1987a), since it considered the use of language as the expression of speech acts, an inherited aspect from its initial affiliation with the Philosophy of Language. Over time, concepts previously integrated into the theory were reformulated (such as polyphony, for example), excluded (such as the topoi, which were part of the second phase of the theory), or added (such as the notion of semantic block, one of the concepts that characterizes one of the most current versions).

Despite the various reformulations, some concepts remain until today and guide the development of research in Linguistic Semantics. In view of this, we seek, through this paper, to raise a discussion about the main principles of the Theory of Language Argumentation, in particular, of its internal hypotheses, that is, of the concepts suggested by the authors to comply with the proposal of creation of a theory capable of explaining how meaning is constructed in the use of language and how enunciations are interpreted by the speakers of this language.

This discussion is justified for different reasons. The first concerns to the presentation of a work that can be taken as an introduction to the studies of Argumentative Semantics, especially to those researchers who look for a first approach to this theoretical construct. Secondly, we consider this work pertinent because of the difficulty to find some works published by the author, whether or not they were translated or were not issued in Brazil (such as Les mots du discours and Les échelles argumentatives, published in 1980, the work considered to be the milestone of Theory of Language Argumentation - La argumentacion en la lengua, originally published in 1983 in co-authorship with Jean-Claude Anscombre, and the collection of conferences given in Colombia, in 1988, called Polifonía Y argumentación) or that already have their editions exhausted (such is the case, for example, of the work Principles of Linguistic Semantics: say and not to say (1977)). In addition to a review text, this article finds its originality in the theoretical framework it proposes, in view of the author’s vast bibliography, in the presentation of concepts and epistemological foundations, in the proposition of a course of study that interprets them in a context of elaboration of a theory that goes beyond three decades, and the unique analysis it offers to its readers.

To accomplish our goal, we start by reading several texts in which Oswald Ducrot and his collaborators describe and explain the way the Theory of Argumentation in Language was conceived. In this paper, we initially present the simulation method according to which the TAL was idealized. We then address the main internal hypotheses of the theory. Lastly, our final considerations.

The simulation method

For Ducrot (1980, 1987a), all scientific activity, from Descartes to the present day, is governed by a mode of conception called simulation method, which considers the artificial reproduction of the facts that are the object of analysis. The simulation method consists in organizing the scientific research into two stages that must follow: an empirical stage and a reproductive stage of the phenomena chosen for the study.

In the first of these two stages, the researcher observes and isolates the phenomena intended to be analyzed and that are produced spontaneously in nature. The second stage is the construction - or idealization - of a machine capable of reproducing facts in an analogous way that happens freely in nature. For the author, this imitation of the real aims at offering the researcher the hypotheses that would explain the process through which facts are produced.

Ducrot (1980) describes this mechanism in detail. For the author, the facts F, spontaneously produced by a natural mechanism M, must be identically reproduced by an artificial mechanism M’, of which, the results are facts F’. The theoretical model M’ is defined by Ducrot (1980, p. 20, free translation) as “a body of hypotheses expressed by artificial language”⁴ . M’ will be considered effective if it has been able to simulate F’ in an analogous manner to which the natural mechanism M produces F. In addition, the M’ mechanism should be able to reproduce other phenomena of/from the same nature as F, such as A, B, C, resulting in A’, B’, C’. In the case of Argumentative Semantics, facts A, B, C are meaningful discourses produced in a certain discursive situation, which receive from the linguist a certain interpretation. For the author, “explaining these facts is to look for the mechanism responsible for this interpretation, considered from a linguistic point of view as part of the data” (DUCROT, 1980, p. 20, free translation)⁵ .

The mechanism described above, from which Argumentative Semantics was conceived, however, it is not developed randomly. It has its creation guided by two hypotheses types, which must be distinguished: the external hypotheses and the internal hypotheses. The first hypotheses are related to the observation phase and they imply that, since the first stages of the research, the observed facts have already been subjected to certain concepts, extracting from the analyzed phenomena only what can be considered as pertinent. In the case of Argumentative Semantics, the external hypothesis, that is, what is taken as the object of observation, is formulated in the following terms: “the manner in which enunciation [sentences] are interpreted in the particular situations in which they are employed” (DUCROT, 1987a, p. 52, emphasized by the author). The formulation of this external hypothesis, as the author himself states, is guided especially by theoretical assumptions in Ferdinand de Saussure’s General Linguistics Course, as well as from the ideas of other authors such as Émile Benveniste and John Austin (DALL CORTIVO-LEBLER, 2014).

To understand the way Argumentative Semantics cuts its object, we start from an article by Benveniste, entitled The Formal Apparatus of Enunciation (1989), in which the author presents three ways of studying the great phenomenon of enunciation. The first of these concerns to the vocal performance of language, more precisely, sounds. The second is about the conversion of language into discourse, whose analysis focuses on defining how meaning is formed from words, which leads to sign theory and meaning analysis, questioning the semantization of a language. The third aspect pointed out by him is the enunciation within the formal framework of its performance. The object of study taken by Benveniste is the third aspect pointed out, being the second one that we attribute to the Theory of Language Argumentation (TAL), since it studies the semantization of language without detaching it from the enunciative framework in which it appears. Unlike Émile Benveniste, whose object of study is the process - the enunciation - Oswald Ducrot chooses as an object the enunciation, the product of the enunciation, which carries the impressions of its production - the meaning of the enunciation reflects, mirrors its enunciation (DUCROT, 1987). This is what we will discuss in the following explanation.

The Internal Hypotheses of the Theory of Language Argumentation

To the comply with the objective of studying the relations between language and discourse, the creation of internal hypotheses in the TAL aims to explain the relationship between three elements: the physical reality of a statement (speech or writing), its semantic value (its meaning) and its employment situation (the enunciation). The internal hypotheses (IH), defined as those that command the creation of the mechanism that simulates the facts to be studied, start from the linguistic materiality of the enunciation, which is seen as the performance of the phrases of language, theoretical construct, integrating a) alterity, b) Saussure’s theoretical precepts and c) enunciative theories. The junction between the theoretical construct and the given one will be, as we will see in detail, by introducing aspects related to enunciation in the theoretical construct, language.

For Ducrot (1980, 1987a), The IH (Internal Hypothesis) are divided into four major groups. The first one contains the internal hypotheses that refer to the way the linguist conceives his sentences - the set of abstractions that make up the language - as well as the correspondence between a determined sentence and a determined enunciation. The second group of internal hypotheses concerns the enunciative aspects of the theory, which specify the way situational representations act upon the meanings of sentences to produce certain semantic values at the same level of enunciation. The third group proposes that the meaning of the enunciation derives from the syntagmatic combination of the terms that compose it and not merely from the sum of their meanings. Lastly, the fourth group of internal hypothesis postulates that, for each word of the language studied, a description of its meaning must be constructed. Let’s look at each of these hypotheses in detail.

In the first set of internal hypotheses, according to which the language is formed by a conjoint of sentences, we find the elements underlying the statement in theoretical state, that is, the sentences/phrases. For the authors of the theory, language consists of a finite number of sentences, that is, a series of words combined according to syntax rules and taken out of the discursive situation. They are constructions of the linguist to explain the infinity of enunciations and their meaning, which is the result of an abstraction based on the observation of several occurrences. In opposition to the theoretical construct, Ducrot (1980, 1987a, 1998) defines his observable as a set of enunciations, which are the performances of the sentence. They consist of the transformation of language into discourse through the enunciative action of a speech subject, a speaker. Therefore, each enunciation contains an allusion to its utterance, which leaves on it the impressions of person, time and space, which make it an unrepeatable occurrence of a certain sentence. For the author, the semantic value of a sentence happens because of its meaning, a concept that aims to explain the occurrences of enunciations, composed of instructions or guidelines that indicate what work should be done by the enunciation interpretant to reach its semantic value - its meaning.

The meaning of an enunciation is to what I consider, a theoretical object: what justifies resorting to it is its explanatory value, the fact that a certain regularity, a certain systematicity is possible, in predicting the meaning of enunciation. (DUCROT, 1992, p. 228, free translation)⁶ .

“meaning”, or even “literal meaning” [...] could not be confused with the “real object”. It is exactly an “object of knowledge,” or “theoretical object.” (DUCROT, 1992, p. 233, free translation)⁷ .

Ducrot (1987a) warns that the passage from sentences to enunciation is not characterized as a sum between meaning and some elements of the situation that would be added to it. For him, it is, in fact, a transformation of language into discourse, something much more complex than the addition of a rhetorical component to the linguistic component. Therefore, he explains that there is a crucial difference of nature and quantity between enunciation and sentence.

The quantitative difference exists because the enunciation says much more than the sentence. Let’s imagine that Paulo, meeting his colleague Pedro, enunciates (1) I have just met with my advisor, I have a lot of work to do. In this enunciation, the object of Paulo to Pedro is the meeting with his advisor, which took place prior to the moment of enunciation. Now let’s imagine Paulo meets his girlfriend the next day and tells her (2) Yesterday I met my advisor; I have a lot of work to do. The object of the enunciation is still Paulo meeting his advisor; however, the enunciation occurred at another time and had another interlocutor. From the point of view of language, the object of Paulo enunciation is only one sentence; however, from the point of view of language in use, we have two different enunciations. This is because the meaning of the enunciation is the representation of its utterance, which, in turns, contains the indicators of person, time and space, that is, the enunciation is overdetermined in relation to the sentence, because it contains the enunciation (DUCROT, 1990).

The difference in nature occurs between the components sense and meaning. The sentence is associated with a meaning, consisting of instructions which provide indications of how to interpret the enunciation. On the other hand, the enunciation has a meaning, which is its semantic value. Thus, the meaning of the sentence, open, constitutes a kind of employment, which suggests indications of how the meaning of the enunciation should be understood. This is one of the reasons why the TAL rejects the notion of literal meaning: if it were considered as part of the meaning, the meaning of the enunciations that compound a certain sentence would have a common trait to which some characteristics regarding to the situation used would be added. For TAL, meaning is quite another thing, it must explain and not describe the meaning of the statement.

The instruction, component of meaning, indicates the work that must be done to understand the enunciation, which must consider the point of view that the speaker of the speech that produced the enunciation wished to express. This way, indications about speech activity are inscribed in the language itself, which introduces speech into the language and enunciation in the sentence (DUCROT, 1987b).

This formulation makes sense as the description of the linguistic system is deduced from the use of the language, the observation of the behavior of infinite enunciations, and that language is the structure that enables its execution. In this sense, the meaning of the sentence must be explanatory and not descriptive, since it must provide the indications to those who interpret a statement, based on the discourse situation, for the search of the precise meaning sought by the speaker, which implies that the meaning is different depending on the discourse situation.

To get at the instruction of a sentence or a word means, therefore, to observe its behavior in innumerable enunciations. From this observation, an open rule is constructed, which will always and only refer to the use of language. Instruction is the bridge that connects the language system and speech.

[…] structural research in linguistic semantics [...] would consist of taking as a domain of studies, as an empirical field, the set of enunciations [sentences]⁸ of a language and defining each of them, from the semantic point of view, from their co-occurring relationships with others in real discourses that this language is the medium. The meaning of an enunciation in the eyes of linguistics would then consist of a sort of condensed representation of the associations of which it is susceptible to use. (DUCROT, 1987b, p. 69, free translation)⁹ .

In addition to the instructional conception of meaning, enunciative aspects also exclude literal meaning from the literal meaning of sentences, since the subjectivation of the speaker in producing language is crucial for the construction of meaning. Thus, there is no predefined meaning, it will depend on the situation used, and what the speaker intends to express.

Ducrot (1990) does not include in his semantic research descriptions in which language is seen as a means of transmitting information about reality, and this information can be evaluated in terms of truth and falsehood. In the understanding of the variative conception of language, to which the author is opposed, the meaning of the words would be constituted by the capacity that these words would have to convey information about what they are talking about, making the objective character predominate, giving it the property to describe the facts as they stand.

The informativeness as a characteristic of language is present in theories that conceive language as a mirror of reality, a concept contested by many theorists - and especially by Ducrot (1990)- because it disregards that between reality and language there is a speaker. In order to refute the notions of objectivity and informativeness supposedly present in language, Ducrot (1990) calls upon to intra-linguistic analysis, guidance rendered by the affiliation of his theory to Saussurian studies of language, as the author himself states: “[…] attributing to the linguistic order an irreducible character, forbidding it to be founded on another level of reality, Saussure establishes at the same time the legitimacy and necessity of seeking in itself the principle of its rationality […]”¹⁰ (DUCROT, 1987b, p. 68, free translation).

In this type of analysis, the study of the linguistic elements is only based on the enunciations already produced, excluding the external context that motivated their production, and contextual research, when necessary, continues to be guided by linguistic markings present in the enunciation. For the author, the notion of truth or falsehood that enunciations appear to have can be explained by the fact that the speaker put different enunciations into his enunciators and assimilated them to this or that person: “[…] what linguists call reality is simply what, according to them, is the object of other sciences on which the obligation to prove is discharged […]”¹¹ (DUCROT, 2004, p. 370, free translation). This means that, according to the author’s point of view, language cannot be given the task of proving something about reality because of the difference in nature between both domains: from language emanates subjectivity, whereas the description of reality claims objectivity.

Considering language as a way of constructing discourses and not as a means of giving information about reality, the linguist also intends to oppose the traditional conception of argumentation. This conception argues that the enunciations of discourse are based on the facts they convey, that is, their linguistic structures have as semantic function the factual description, and this semantic function is merely informative. Let’s look at the scheme discussed by Ducrot (1990).

In the figure above, an argumentative discourse must satisfy some conditions, which are, to have two segments, one named argument (A) and another conclusion (C). A must indicate a fact F, that gives the argument A the possibility to be false or true. The conclusion C can be inferred from the fact F. The author argues that, in this argumentative conception, the role played by language is very small, since the tie between A and C has nothing to do with language, only with the fact that it spreads. It is characterized by its exteriority to language, as it seeks in the facts of the world the motivation for conclusions, which, in turn, are likely to be false or true. For Ducrot (1990), this conception invalidates the role of language, which becomes only an instrument that subjects use to talk about what is external to it.

This position adopted by the linguist arose when he started to observe that some statements derived from the same fact, however, presented different conclusions, especially by the introduction of argumentative operators like little and a little. Imagine a situation in which a person is sick and, in order to be better, this individual needs to eat. In this case, we can have two enunciations: (3) Pedro ate little and (4) Pedro ate a little.

In this enunciation (3), the conclusion that can be drawn is that Pedro ate little, so his health will not be better, and, for the enunciation (4), the conclusion would be the opposite, Peter ate a little, so his health will be better. It is important to notice that both statements translate the same fact, that is, Pedro ate a small amount of food, however, what suggests the conclusion not to be the same, for both enunciations the words little and a little, are named by the author as argumentative expressions. These expressions have the function to determine which segments can be related to the first one, since for the theorist there is no possibility of discourse segments expressing meaning if they are not related to one another.

In this case, the first segment of the enunciation only has its defined meaning when related to what follows, the meaning being set out by the interrelation between them. Other examples that question the notion of relationship, as well as the argumentative orientation, are as follows: (3) Pedro ate little and (4) Pedro ate a little. In (3), there is the expression of a finding, made by observing the amount of food ingested by Pedro, which could lead to the continuation. Pedro ate little, soon he will be hungry again. But in (5), there is the indication of a habit, which leads to the continuation Pedro eats little, he is concerned about his health. The argumentative orientation, in this case, is determined by the inflection of the verb, which, along with the other segments, permits certain continuations and prohibits others.

The argumentative orientation of a term, by its characteristic of pointing out the possible relationships that may exist between one term and another, places in the paradigm of choices a series of elements that can be linked as continuations in a discourse. So, the value of the term, or its meaning, will be determined by crossing the syntagmatic axes - of combinations - and paradigmatic - of possibilities:

We can understand why I call such a conception structuralist: the act of enunciation [utterance] is defined in it, from the semantic point of view, by its relation to what is beyond it, more precisely, by its relation to the future it projects. : its continuation as a constitutive of its being. (DUCROT, 1992, p. 221, free translation)¹² .

I would like to point out this property, characteristic in my view, of all speech, of being definable only in relation to a sequence from which it is intended to begin, so that its meaning is the very meaning in which it intends to lead the discourse situation. (DUCROT, 1992, p. 223, free translation)¹³ .

The definition of the meaning of a term, or an enunciation, based on the relationships it establishes with other language terms, motivated Ducrot (1990) and collaborators to reject the traditional argumentation proposal, since for them, the segment A of the figure 1, which refers to a fact, cannot be judged in terms of truth or falsehood, nor can it be taken separately, since words cannot have a complete meaning without relating them to others, which, in the same way, occurs with the segments of the enunciations. Moreover, it is not only the facts that have the ability to argue in a discourse, it is not only the facts that determine the argumentative power of the enunciation, but mainly the linguistic form that this fact assumes. Facts alone tell us nothing, language translates them, and this is only possible through the speech of a speaker who expresses their position through linguistic markings, which gives language the power to dialogue.

Figure 1
– Traditional Conception of Argumentation

As mentioned above, the traditional conception of meaning has always seen in language an objective aspect, which gives it the property to speak of the outside world to it with complete impartiality. In addition to this aspect, two other aspects are part of this description of meaning: the subjective aspect, concerning the position that the speaker assumes in relation to what this person is dealing with, and the intersubjective aspect, which is constituted by the relation that the speaker maintains with their interlocutor, a distinction used especially by the German Karl Bühler (apudDUCROT, 1990).

Among these three indications of meaning, Ducrot (1990) rejects the one that states that language has an objective aspect, because he believes that it does not give access to reality directly, but that reality consists only in a theme for debate between two individuals (DUCROT, 1990). The author also believes that only the subjective and intersubjective aspects are constitutive of meaning and considers that together as they form the argumentative value; the argumentative value is constituted by the set of possibilities of continuation that can be given to a discourse and considered as the fundamental level of description of meaning, which explains the illusion that words have a real factual value (DUCROT, 1990).

There are never phrase-level informational values. Not only are there no purely informative sentences, but there is not even an informative component in the meaning of sentences, which does not mean that there are no informative uses of sentences. [...] Such (pseudo)informative uses are derived from a purely argumentative ‘deeper’ component. (ANSCOMBRE; DUCROT, 1994, p. 214, free translation)¹⁴ .

The concepts of argumentative orientation and argumentative value seem very to be very similar, however, their difference lies in the enunciation. The first one concerns to the possibilities of linking one segment to another, while the second one takes into account the subjectivation of the speaker and the appeal this individual makes to their interlocutor. The argumentative value would be close to the meaning, since it contains the enunciation, while the argumentative orientation would be close to the meaning(signification), because it contains, in a certain way, indications from a linguistic nature.

The second group of internal hypotheses is part of the idea that the enunciation is the historical event responsible for the emergence of the enunciation. For Ducrot (1980, 1987a), the enunciation has three meanings: the first one is characterized by being the psycho-physiological activity involved in the production of the enunciation. The second one is the product of the speaking activity from the subject, which is the enunciation itself. The third one, and the considered by the author as adequate to the semantic description he intends to make, is defined as the historical event that constitutes the appearance of the enunciation. Thus, the enunciation is an event that culminates in the emergence of the utterance, that is, there is a moment when the enunciation does not exist yet and another when it does not exist anymore. This interval of time is the enunciation.

Another reason that inserts the TAL in the enunciative framework is the fact that it considers that any and all linguistic manifestations are the product of a speech from a speaker activity, which is to an interlocutor. These figures are essential in the TAL, because, besides not having the possibility of language in use without the presence of these elements, the meaning constructed by the use of language is the result of the position adopted by the speaker about the things spoken by this speaker.

Although the object of analysis is not the enunciation itself, but its product, Ducrot considers inseparable enunciation and utterance, because “[…] the meaning of the utterance is a description, a representation that it brings from its enunciation, an image of the historical event constituted by the emergence of the enunciation […]”¹⁵ (DUCROT, 1980, p. 34, free translation). Every enunciation must mention its utterance, as it is crucial in explaining the meaning constructed by the speaker.

The observation of the enunciation without considering a speech situation makes it identical to all others; it is necessary to mention the act which produced the enunciation at different points of time and space. Therefore, to consider the enunciation as the mirror of its own utterance is, first of all, to admit that it presents itself as produced by a speaker who addresses a statement to an interlocutor.

Explaining the meaning of enunciation by constructing rules, or generalities containing indications of the meaning to be sought, there must inevitably be indications concerning the speech activity. Many enunciations run the risk of not receiving a satisfactory description if there is not an allusion to the fact of being used for a particular purpose: to interact through enunciation. Thus, it is reaffirmed that the enunciation must have its properties defined without any allusion to social, historical or psychological elements, but they should be considered in relation to the act of being employed. Therefore, it is not possible to make a description of the language without alluding to the speech activity, because the language itself contains elements that refer to its use and these two instances cannot be taken as independent.

The third group of IHs deals with the combinatorial principles according to which the meanings of words merge to give meaning to sentences, that is, - the meaning in the enunciation is the result of the lexical combination. What the author points out is that the calculation of meaning must take segments wider than a word, and that, instead of considering words in relation to each other, considers them in relation to discourses, which makes their work a structuralism of discourse (MOURA, 1998). Quoting from Ducrot’s words clarifies the meaning of this group of IHs:

The word, conceived as an abstract linguistic entity, only collaborates in the meaning of the enunciation in an indirect way: it begins to link with the other words to constitute the meaning of the sentence, and it is, considered the discourse situation, the production of meaning of the enunciation. (DUCROT, 1980, p. 09, free translation)¹⁶

In this group of internal hypotheses, the keyword is relation. It is through the relationship between the segments of an utterance that its value, i.e. its meaning, is defined. In this case, as for Saussure’s sign, there is no definition of meaning (or a definition of what a sign is) without first determining what kind of relation the terms have to each other. According to Ducrot (2009, p. 11), “[…] a linguistic entity (an enunciation, for example) cannot define itself independently of its use in a dialogue. [...] The enunciation will then be defined by the possibilities of response that it opens and of those that it closes.”

Lastly, the fourth group of internal hypotheses proposes that, for each word of the language studied, a description of its meaning must be constructed. Regarding to it, Ducrot devotes the study of some words like mais, décidément, eh bien! d’ailleurs in his book Les mots du discours (1980), which he calls them operators. They represent a class of words responsible for promoting the articulation of enunciation and they manifest formal characteristics. The most common case is the use of P but Q, to which Ducrot devotes several articles¹⁷ , and certainly motivated him to develop various theoretical tools.¹⁸

Thus, we close the presentation of the internal hypotheses of the TAL, which constitute its main body from which the concepts that operationalize the semantic analysis were formulated.

Final considerations

From the explanation given in the above pages, it is possible to state that the external hypothesis of The Theory of Language Argumentation, “the manner in which enunciations [statements] are interpreted in particular situations in which they are employed” ²¹(DUCROT, 1987a, p. 52, free translation, author’s highlights) based on theorists like Saussure (1975), Platão (1969), Benveniste (1989 9, 2005) and Austin (1962), is present in the creation of the internal hypotheses. The relationship between IH and EH is inherent, as IH must always be in accordance with EH. Those, unlike these, may suffer changes: the changes in the IHs made during the theoretical course of Ducrot (1980, 1990, 2008) and his collaborators, portrayed in the different phases by which the theory was formulated and reformulated (alteration of the concept of polyphony, exclusion of the notion of tops and creation of concepts in Semantic Block Theory), in no way they compromise the functioning of the mechanism created, on the contrary, the adjustments made were created with the intention of adjusting them to be even more faithful to the cause they embraced.

Thus, the attribution of a semantic value to the sentences of the language is not an external hypothesis, since it does not originate from the observation, but refers to the process performed by the mechanism M’, whose work was to produce an explanation based on an artificial language for the fact F taken at the beginning of the observation. The attribution of a semantic value to the phrases in a language is the revelation of the mechanism of fact production that underlies the whole process, so it is artificial and therefore theoretical, metalinguistic.

If it is the point of view that creates the object, as stated by Saussure (1975), we will always be stuck in creating an object, its pre-definition in relation to reality, to the data. From this stems the originality in Saussure`s statement in which language is form and not substance and cannot be given in advance.

REFERÊNCIAS

• ANSCOMBRE, J.-C.; DUCROT, O. La argumentación en la lengua Madrid: Gredos, 1994.
• ANSCOMBRE, J.-C.; DUCROT, O. L’Argumentation dans la langue Bruxelas: Mardaga, 1983.
• AUSTIN, J. L. How to do things with words Oxford: Oxford University Press, 1962.
• BENVENISTE, É. Problemas de linguística geral I Campinas: Pontes, 2005.
• BENVENISTE, É. Problemas de linguística geral II Campinas: Pontes, 1989.
• BRÉAL, M. Ensaio de semântica São Paulo: Pontes, 1992.
• CAREL, M.; DUCROT, O. La semántica argumentativa: una introducción a la teoría de los bloques semânticos. Buenos Aires: Colihue, 2005.
• CHOMSKY, N. Syntactic structures Berlin: Mouton de Gruyter, 1957.
• DALL’CORTIVO-LEBLER, C. A teoria da argumentação na língua e sua relação com Platão, Saussure e Benveniste: breve discussão epistemológica. Filologia e Linguística Portuguesa, São Paulo, v.16, n.2, p. 331-364, 2014.
• DUCROT, O. Présentation de la théorie des blocs sémantiques. Verbum: Revue de Linguistique, Nancy, v.38, n. 1-2, p. 53-65, 2016.
• DUCROT, O. Argumentação retórica e argumentação linguística. Letras de Hoje, Porto Alegre, v. 44, n. 1, p. 20-25, jan./mar. 2009.
• DUCROT, O.; CAREL, M. Descrição argumentativa e descrição polifônica: o caso da negação. Letras de Hoje, Porto Alegre, v. 43, n.1, p. 7-18, jan./mar. 2008.
• DUCROT, O. Sentido y argumentación. In: ARNOUX, E. N.; NEGRONI, M. M. Homenaje a Oswald Ducrot Buenos Aires: Eudeba, 2004. p.359-370.
• DUCROT, O. L’interprétation en sémantique linguistique: un point de départ imaginaire. In: DUCROT, O. Dire et ne pas dire: principes de sémantique linguistique. Paris: Hernann, 1998. p. 307-323.
• DUCROT, O. Nota sobre a pressuposição e o sentido literal (posfácio). In: HENRY, P. A ferramenta imperfeita Campinas, SP: Ed. da Unicamp, 1992. p. 203-238.
• DUCROT, O. Polifonía y argumentación: conferencias del seminario Teoría de la Argumentación y Análisis del Discurso. Cali: Feriva, 1990.
• DUCROT, O. A descrição semântica em linguística. In: DUCROT, O. O dizer e o dito Campinas: Pontes, 1987a. p. 45-62.
• DUCROT, O. Esboço de uma teoria polifônica da enunciação. In: DUCROT, O. O dizer e o dito Campinas: Pontes, 1987b. p. 161-218.
• DUCROT, O. Analyse de textes et linguistique de l’énonciation. In: DUCROT, O. et al. Les mots du discours Paris: Minuit, 1980. p. 07-56.
• DUCROT, O. Mais occupe-toi d’Amelie. Actes de la Recherche en Sciences Sociales, [S.l.], n. 6, p. 47-62, 1976.
• DUCROT, O.; VOGT, C. De magis a mais: une hypothèse sémantique. Revue de Linguistique Romane, Lyon/Strasbourg, tomo 43, p.171-172, 1979.
• FERRAREZI JR., C.; BASSO, R. Semântica, semânticas São Paulo: Contexto, 2013.
• ILARI, R.; GERALDI, J. W. Semântica São Paulo: Ática, 2006.
• MOURA, H. M. de M. Semântica e argumentação: diálogo com Oswald Ducrot. DELTA, São Paulo, v.14, n.1, p.169-183, fev. 1998.
• PLATÃO. Sophiste, politique, philèbe, timée, critias Édition établie par Émile Chambry. Paris: Garnier-Flammarion, 1969.
• SAUSSURE, F. de. Curso de linguística geral Organizado por Charles Bally e Albert Séchehaye. Tradução: Antônio Chelini, José Paulo Paes e Izidoro Blikstein. 7. ed. São Paulo: Cultrix, 1975.
• SEARLE, J. R. Expression and meaning. Cambridge: Cambridge University Press, 1979.
• SEARLE, J. R. Speech acts. Cambridge: Cambridge University Press, 1969.
• TAMBA, I. A semântica São Paulo: Parábola, 2006.

• 1
  Original: “a semântica é um domínio de investigações de limites movediços.” (ILARI, GERALDI, 2006, p. 06).
• 2
  In Brazil, Editora Contexto published in 2013 the work Semântica, Semânticas (semantics), organized by Celso Ferrarezi Junior and Renato Basso, with the purpose of presenting an overview of the different “semantics” developed by Brazilian linguists. This work shows not only the diversity of theoretical approaches but also the richness of those researches produced.
• 3
  Original: “Este livro, iniciado e interrompido muitas vezes, e do qual [...] fiz publicar algumas partes, em diversas retomadas, decido hoje liberá-lo ao público. Quantas vezes, desanimado pelas dificuldades de meu assunto, eu me prometi não mais retornar a ele!” (BRÉAL, 1992, p. 19).
• 4
  Original: “un ensemble d’hypothèses exprimées par un langage artificiel” (DUCROT, 1980, p. 20).
• 5
  Original: “Expliquer ces faits, c’est chercher quel est le mécanisme M responsable de cette interprétation, interprétation considérée, du point de vue linguistique, comme partie du donné.” (DUCROT, 1980, p. 20).
• 6
  Original: “A significação de um enunciado é para mim um objeto teórico: o que justifica recorrer a ela é seu valor explicativo, o fato de que ela torna possível uma certa regularidade, uma certa sistematicidade, na previsão do sentido das enunciações [enunciados].” (DUCROT, 1992, p. 228).
• 7
  Original: “‘ significação’, ou ainda ‘sentido literal’, [...] não poderia ser confundido com o ‘objeto real’. Trata-se exatamente de um ‘objeto de conhecimento’, ou ‘objeto teórico’.” (DUCROT, 1992, p. 2).
• 8
  Throughout Ducrot’s work, some concepts underwent through changes, especially in relation to their denomination. One of them is the concept of sentence, which on the date the article from which we cited this quote was written, Ducrot, in many passages, took it by synonym of enunciation. However, we suggest a correction, in brackets, to still be true to the original text without any conceptual damage to our work.
• 9
  Original: “[…] a pesquisa estrutural em semântica linguística [...] consistiria em tomar como domínio de estudos, como campo empírico, o conjunto de enunciados [frases] de uma língua e definir cada um deles, do ponto de vista semântico, a partir das suas relações de co-ocorrência com outros nos discursos reais de que essa língua é o meio. A significação de um enunciado [frase], aos olhos da linguística, consistiria, então, numa espécie de representação condensada das associações de que ele é suscetível no uso” (DUCROT, 1987b, p. 69).
• 10
  Original: “[…] atribuindo à ordem linguística um caráter irredutível, proibindo-a de fundar-se num outro nível de realidade, Saussure estabelece, a um só tempo, a legitimidade e a necessidade de procurar nela própria o princípio da sua racionalidade […]” (DUCROT, 1987b, p. 68).
• 11
  Original: “Lo que los lingüistas llaman ‘realidad’ es simplemente aquello que, según ellos, es el objeto de otras ciencias sobre las que se descarga la obligación de probar” (DUCROT, 2004, p. 370).
• 12
  Original: “Vê-se por que chamo de estruturalista tal concepção: o ato de enunciação [enunciado] é nela definido, do ponto de vista semântico, por sua relação com aquilo que está além dele, mais precisamente, por sua relação com o futuro que ele projeta: sua continuação aparecendo como constitutivo do seu ser.” (DUCROT, 1992, p. 221).
• 13
  Original: “Gostaria de apontar essa propriedade, característica a meu ver, de toda fala, de ser definível somente com relação a uma sequência da qual ela se pretende o começo, de modo que seu sentido é o próprio sentido no qual ela pretende orientar a situação de discurso.” (DUCROT, 1992, p. 223).
• 14
  Original: “Nunca hay valores informativos en el nivel de la frase. (There are never informative values at sentence level.) No solo no hay frases puramente informativas, sino que ni siquiera hay, en la significación de las frases, componente informativo, lo que no significa que no hay usos informativos de las frases. […] tales usos (pseudos)informativos son derivados de un componente más ‘profundo’ puramente argumentativo.” (ANSCOMBRE; DUCROT, 1994, p. 214).
• 15
  Original: « Le sens de l’énoncé, c’est, pour moi, une description, une représentation qu’il apporte de son énonciation, une image de l’événement historique constitué par l’apparition de l’énoncé. » (DUCROT, 1980, p. 34).
• 16
  Original: « Le mot, conçu comme entité linguistique abstraite, ne collabore au sens de l´énoncé que d’une façon indirecte: il commence par se combiner aux autres mots pour constituer la signification de la phrase, et c’est celle-ci qui, vu la situation de discours, produit le sens de l´énoncé. » (DUCROT, 1980, p. 09).
• 17
  A totalidade de hipóteses internas é apontada em Ducrot (1980, p. 21). Ducrot and Vogt (1979) and Ducrot (1976) retomado em L’Argumentation dans la langue, em colaboração com Jean-Claude Anscombre (1983) e em La semántica argumentativa: una introducción a la teoría de los bloques semánticos por Marion Carel (2005).
• 18
  The totality of internal hypotheses is pointed out in Ducrot (1980, p. 21).
• 19
  Original: “a maneira pela qual os enunciados [frases] são interpretadas nas situações particulares em que são empregadas” (DUCROT, 1987a, p. 52).

Publication Dates

• Publication in this collection
  24 June 2020
• Date of issue
  2020

History

• Received
  10 June 2018
• Accepted
  03 Apr 2019