This chapter presents the theoretical background for the comparative analysis conducted here. It has a clear historical and epistemological dimension, as it seeks to confront two major traditions in the research field of discourse marking. Its scope is rather modest, though, since it does not provide an exhaustive overview of the scholarly literature on DMs. In line with the English-French contrastive view adopted in the present book, I more modestly compare two prominent traditions in the study of DMs—on the one hand, the pragmatic tradition dominant in the English-speaking countries; on the other hand, the enunciative tradition, which has been influential only in France and Switzerland. An overview of the literature on DMs in French and English linguistics actually highlights major epistemological differences between the two linguistic traditions. Therefore, comparing these two linguistic approaches sheds light on crucial categorisation issues concerning the definition of the class of DMs itself and the theoretical principles underlying the study of specific DMs. The questions that we need to address are the following: What is the unifying factor behind this supposed “class” of markers? What is the part played by semantics, pragmatics and syntax, respectively, in this definition? How do we account for the development of DMs and the links between the original non-discursive uses and the newly acquired discursive uses? The two traditions compared do not necessarily provide the same answers, hence the need to examine these various issues at closer quarters.
In the English-speaking literature, DMs are mainly analysed as functional units in line with the pragmatic tradition initiated by Schiffrin (1987) in this research field. On the other hand, the French tradition, strongly influenced by speaker-centred theories known as theories of “énonciation” (Ducrot et al. 1980, Ducrot 1985; Culioli 1990, 1999a, b), sees DMs as traces of subjectivity reflecting various attitudinal stances. Moreover, the pragmatic tradition has never paid much attention to DMs based on a verbum dicendi (with the exception of Craig and Sanusi 2000; Brinton 2005, 2008: 73–110), whereas the French enunciative framework has given rise to many studies on such DMs (Saunier 2012; Péroz 2013; Gómez-Jordana Ferary and Anscombre 2015; Rouanne and Anscombre 2016 to name only the most recent ones). Theories of “énonciation” have been mainstream theories in France and Switzerland for the last forty years. Outside the French-speaking community, however, they remain largely unknown, part of the reason for this being that few publications are in English (with the notable exception of Ranger 2018, who provides a full-fledged presentation of Culioli’s theory, the Theory of Enunciative and Predicative Operations). The two goals pursued in this chapter are to familiarise scholars with enunciative linguistic approaches and to assess the convergences and divergences between these theories and better-known pragmatics-based theories.
Section 1.1 lays out the main tenets underlying the study of DMs within a pragmatic framework, successively examining several aspects of this linguistic tradition. I shall first briefly retrace the rise of DMs as a “non-syntactic” but functional class in the 1980s (Sect. 1.1.1). I shall then shed light on the functionalist view defended by pragmatics, at the expense of a semantics-based analysis (Sect. 1.1.2). Subsection 1.1.3 explores the different types of functions that DMs may serve, and Subsection 1.1.4 discusses pragmaticalisation and other diachronic processes that may lead to the development of DMs. Concluding remarks are presented in Sect. 1.1.5. In Sect. 1.2, I move on to the definition of DMs within the French-speaking tradition of “énonciation”. I shall start with a general presentation of “énonciation” as a speaker-centred theory (Sect. 1.2.1). The next Subsection (1.2.2) is devoted to the definition of DMs as “markers of operations” in enunciative frameworks, while Sect. 1.2.3 concentrates on the unitary approach to DMs advocated in Culioli’s Theory of Enunciative and Predicative Operations. Subsection 1.2.4 sheds light on the specificity of DMs of saying as reflexive markers commenting on speaker commitment. Concluding remarks on enunciative approaches to DMs are provided in Sect. 1.2.5.
1.1 Defining DMs in the Pragmatic Tradition
1.1.1 Historical Landmarks: The Rise of a “Non-syntactic” Functional Class
Historically speaking, the class of DMs was first defined in the 1980s within pragmatic studies that took an interest in items such as well or I mean used in oral interaction in English, as in Schiffrin’s (1987) pioneering study. Originally, “discourse” thus referred to oral interaction and responded to the crucial need to set up a new referential frame that went beyond sentences or clauses and could accommodate linguistic markers that were hard to analyse through well-established syntactic categories (adverb, conjunct, etc.). Ranger (2018: 23) notes that the term “discourse” may refer either to an extra-sentential level of analysis, or to language use (“discourse” being opposed to abstract language structures). In original pragmatics-based works, it corresponds to the former definition with an additional emphasis on dialogical spoken interaction. In enunciative approaches, discourse more simply refers to any language activity and does not have a specific meaning in terms of genre (see Sect. 1.2).
The emergence of this research field may be seen as an important epistemological turn in linguistics, as it opened up new research paths in areas that had been overlooked by the dominant linguistic theories of the time, i.e. structuralism and generative grammar (Celle and Huart 2007: 1–2). Coining a new term—“discourse marker”—was a way to expose the inadequacy of syntax to account for such linguistic items. The class of DMs is thus intrinsically “non-syntactic”, and this has two major consequences. First, from a semasiological viewpoint, the “non-syntactic” approach explains the heterogeneity of the members of this class (Dostie and Pusch 2007; Lewis 2006; Beeching 2016), which gathers very different markers that developed through a decategorisation/recategorisation process (i.e. adverbs, interjections, clauses that came to be recategorised as DMs). The four DMs compared in the present study went through that very process: they started out as full clauses and gradually acquired a new status as DMs, possibly as “comment clauses” (Brinton 2008). I shall leave this question open for now: the syntactic behaviour of the four DMs of saying under scrutiny and their possible recategorisation as specific DMs called “comment clauses” (or “reduced parenthetical clauses” in other theoretical categorisations, see Schneider 2007) will be discussed in relation to the corpus findings in Chapters 4 and 5.
It should be stressed that there exists a variety of terms to refer to these specific markers: DMs, but also pragmatic markers, pragmatic particles, etc. (Beeching 2016: 3). These various terms are not neutral and tend to reflect specific theoretical positions. For instance, the term “fillers” implies that these markers are devoid of any meaning, a belief that is nowadays criticised by most researchers, hence the decline of the term itself in the scholarly literature (Dostie and Putsch 2007: 6). “Discourse marker” is undoubtedly the most widespread and theoretically neutral term (Paillard 2017).
Secondly, the “non-syntactic” nature of this new linguistic class has led researchers to resort to other types of analysis, mainly pragmatic within the English-speaking research community. It should be noted, however, that some linguists have attempted to establish purely syntactic criteria to define DMs, especially in terms of initial position, optionality and loose connection with the rest of the clause (Schourup 1999: 230–232; Brinton 2008: 1). As stressed by Fischer (2006), such attempts have failed, since there is no consensus whatsoever regarding syntactic criteria. For inst...