Discourse Markers in Romance Languages. Crosslinguistic Approaches in Romance and Beyond

1. Introduction¹

This paper focuses on the semantic and pragmatic meanings of the Italian adverb davvero ‘really’, showing that under peculiar textual circumstances it triggers specific pragmatic enrichments. In particular, we describe the pragmatic meanings conveyed by davvero in a range of text types, including spontaneous spoken language, spoken political discourse, political tweets and written news report. The data collected from different corpora are analysed both quantitatively and qualitatively, to provide a description of the role of the adverb in the organization of discourse, focusing on its pragmatic use(s) in different text types.

The paper is organized as follows: in Section 2 the state of art is presented; in Section 3 the reference corpora and the methodology adopted are introduced; in Section 4 data are analysed; in Section 5 we focus on the manipulative potential of the structures introduced in Section 4; in Section 6 a discussion of the findings is provided; in Section 7 conclusions are drawn.

2. State of the Art

At least since Traugott (1995), adverbs are acknowledged to be a class of words which easily get additional pragmatic meanings². The pragmatic enrichment triggered by Italian davvero in some specific linguistic configurations has been firstly described by Cimmino and Cominetti (2023a), while previous studies on the adverb have focused on its syntactic, semantic and textual features, with no specific reference to the pragmatic meanings (De Cesare, 2000, 2002a, 2002b, 2003; Ricca & Visconti, 2014). Cross-linguistically, the pragmatic enrichment triggered by adverbs whose meaning is comparable to that of davvero has been studied. In particular, the work of Paradis (2003) and Romero and Han (2004) on the English adverb really, and the work of Yuan (2015) on Chinese zhende ‘really’ show pragmatic behaviours for some aspects similar to those of davvero.

In this work, we wish to describe some of the pragmatic meanings of davvero as implicitly conveyed contents: in particular, we will show that davvero – under some specific textual circumstances – is able to trigger either a conversational implicature, or both a conversational implicature and a presupposition. In fact, the pragmatic enrichment triggered by discourse or pragmatic markers can be analysed not only as a speech act marker or context marker, but also as implicit content (Lagerwerf, 1998; Karagjosova, 2001; Zeevat 2004; Yuan, 2015).

Since we will refer to the notions of presupposition and implicature to describe the pragmatic meanings triggered by davvero, let us clarify to which definitions of these categories we are relying on. For presupposition, we are adopting the definition by Stalnaker (2002: 701): ‘to presuppose something is to take it for granted, or at least to act as if one takes it for granted, as background information – as common ground among the participants in the conversation’. Presupposition is considered a heterogeneous category (Levinson, 1983: 217; Sbisà, 2007). A long tradition has described many different linguistic triggers, including definite descriptions (Frege, 1892), change of state predicates (Sellars, 1954; Karttunen, 1973), factives (Kiparsky & Kiparsky, 1971; Karttunen, 1971), iteratives (Levinson, 1983: 182), and more. Recently, particles, conjunctions and discourse markers have also been described as presupposition triggers (Lagerwerf, 1998; Karagjosova, 2001; Zeevat, 2004; Yuan, 2015). Typically, presupposed content refers to the existence and/or unicity of some entity, the being in place of some state of affairs, or the happening of an action. However, presupposed content may also include expectations and pre-requirements.

As for implicatures, we are adopting the milestone definition by Grice (1975): implicatures are propositions that in specific contexts can be communicated through an utterance without being explicitly said. According to Grice, conversational implicatures arise as a consequence of the speakers’ compliance to the cooperative principle: ‘Make your contribution such as is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged’. The apparent violation in discourse of one of the four maxims of conversation, which jointly express the cooperative principle, can also lead to the activation of an implicature.

3. Corpora and Methodology

In order to investigate the role of davvero in the organization of discourse, we considered different text types, using four corpora of Italian as a source of data: the IMPAQTS corpus of political monologic discourse (Cominetti et al., 2022), the Kiparla corpus of spontaneous spoken conversation (Mauri et al., 2019), the Repubblica corpus of written journalistic prose (Baroni et al., 2004); finally, we considered an ad hoc corpus of political tweets. The textual variation of this collection of corpora was conceived to allow disentangling the pragmatic use(s) of davvero from specific discourse traits. In particular, we want to gauge the impact of diamesic and diaphasic variation, as well as of the different communicative aims of the text. For this reason, not only are the selected corpora varied diamesically (spoken vs written) and diaphasically (spontaneous vs planned), but they are generally expected to represent a more (IMPAQTS and tweets corpora) or less (Repubblica and Kiparla) persuasive text type³.

The IMPAQTS corpus of Italian political discourse, currently under construction, includes monologues uttered by 150 Italian politicians in different public contexts (chambers of the Parliament, meetings, rallies, press conferences, etc.) from 1946 to present. For the purposes of the present work, we interrogated the available section of the corpus, amounting to 763 speeches and 1,470,144 tokens. To obtain a comparable amount of data representative of other text types, we considered the entire Kiparla corpus (1,125,977 tokens) and a section of Repubblica corpus, choosing the portion of data produced in the year 2000 in the ‘news’ section⁴, amounting to 1,421,721 tokens. Finally, we considered a corpus of around 382,000 words, composed of 11,450 tweets twitted by the official accounts of four Italian politicians⁵. The corpus was collected in the month of February 2022 by automatically downloading the 3000 most recent tweets published by each of the four considered people: accordingly, the tweets cover a time span from the beginning of 2013 to February 2022.

The four corpora were automatically searched for occurrences of the adverb davvero. The retrieved occurrences were then manually analysed. More precisely, we annotated: the type of sentence in which the adverb occurs; the semantic reading of the adverb (cf. Section 4.2.1); whether the adverb triggers a pragmatic enrichment; in case the adverb has a pragmatic meaning, what this meaning is.

4. Data Analysis