Functions of Language - Volume 30, Issue 1, 2023

This paper examines the linguistic realization of continuative discourse relations in British English written discourse comparing narrative and argumentative dyadically edited texts. The data comprise 18 co-edited texts and metadata documenting the editing process (keystroke logs and transcripts of the dyads negotiating discursive well-formedness). The focus of analysis lies on the linguistic realization of coordinating continuation and narration, which keep the discourse on the same level, and on the linguistic realization of subordinating elaboration and explanation, which introduce a deeper level in the discourse hierarchy. Special attention is paid to contexts in which the discourse relations are encoded in intra-clausal coherence strands, and to contexts in which they are additionally signalled in the peripheries. The quantitative analysis of the signalling of continuative discourse relations shows genre-specific preferences for the signalling of continuation and elaboration in the argumentative data, and continuation, narration and explanation in the narrative data. Both the products of the edited data, the co-edited texts, and the metadata show that the linguistic realization and interpretation of continuative discourse relations are – to varying degrees – subject to recontextualization. We suggest that this variation provides evidence for (1) discourse relations as constitutive parts of discourse grammar, and (2) genre as a blueprint which constrains their linguistic realization.

Continuity and discontinuity (maintaining or shifting deictic centres across segments) are important aspects of discourse relations. Yet they have been attributed to these relations in very different ways. This calls for an analysis of individual instances of discourse relations with respect to their continuity dimensions. To this end, we operationalise Givón’s (1993) continuity dimensions (time, space, reference, action, perspective, modality, and speech act), decomposing them into distinctive features that allow a consistent and accurate classification of the continuity dimensions in discourse relation tokens. This inventory was applied to five representative relation types (causal, contrastive, conditional, elaboration, and temporal) from the RST Discourse Treebank (Carlson & Marcu 2001). We found that relations can simultaneously be more continuous for some dimensions but more discontinuous for others. What is more, discourse relations typically vary widely in different continuity dimensions and thus cannot be described as fully continuous or discontinuous, neither on the level of the entire relation type nor for one of its particular dimensions. Using examples of causal, conditional, and contrastive relations, we also illustrate how the results of our analysis can be used to verify hypotheses about correlations between continuity and discourse relations.

Proposals such as continuity and causality-by-default relate the level of expectedness of a relation to its linguistic marking as an explicit or implicit relation. We investigate these two proposals with regard to the English transcripts of six TED Talks and their Lithuanian, Portuguese and Turkish translations in the TED-Multilingual Discourse Bank (TED-MDB), annotated for discourse relations, following the Penn Discourse Treebank style of annotation. Our data shows that the discontinuous relations contrast and concession are indeed frequently explicit in all languages. But continuous relations show differences per relation and language. For instance, cause is frequently conveyed implicitly in English and Portuguese, but not in Lithuanian and Turkish. We explore temporal continuity by analysing whether the forward-order sense result is more frequently implicit than the backward-order reason. The hypothesis is confirmed by English and Portuguese, but not Lithuanian and Turkish. However, in Turkish, the arguments of the backward-order relation reason are frequently presented by the reversed order of arguments, retaining the linear order of events even in the presence of the connective. The causality-by-default hypothesis is not confirmed, as cause is not the most frequent implicit relation in the four languages.

In research on information structure and discourse cohesion, contrast has been defined in different ways, depending on the pragmatic/semantic relation established between the propositions involved in the contrast, on the text types and on other discourse conditions. As a whole, despite – or possibly because of – its vagueness, contrast has proved to be a useful heuristic tool for characterizing discourse cohesion phenomena. This paper focuses on results from our research concerning cohesion phenomena in elicited discourse in Romance (Italian, French) and Germanic (German, Dutch) languages and aims to offer a more precise characterization of contrast against several variation parameters. We take into consideration earlier work on three tasks (Finite Story, Polarity-Switch Dialogues, Map Task) and add a new one (Spot the Difference). The comparison between the results allows us to disentangle the following variables: information units involved in the contrast relation; discourse conditions (monologue vs. dialogue); speakers’ access to information (shared vs. non-shared); effect of contrast on information in the common ground (alternative maintained vs. rejected). The aim is to achieve a more fine-grained definition of contrast relations, which allows us to identify and characterize the divergent behavior of Romance and Germanic languages, and to relate intra- and crosslinguistic differences revealed by speakers’ preferences in speech with structural specificities of the two language groups.

This contribution aims to set out the effects of discourse marking on processing. On the basis of examples from Spanish, we try to show the principles governing the interplay between the procedural meaning of discourse markers (connectives) and the conceptual meaning of the discourse segments linked by them. To determine these principles, two types of experiments were performed: one comparing marked and unmarked utterances, and a second one comparing utterances that activate mental representations that pragmatically match or clash with the instruction encoded by the connective. Evidence shows that (a) discourse marking by means of a connective generates a new route for accessing information; (b) procedural meaning is a definitory feature of connectives; (c) the procedural meaning of connectives introduces asymmetry and rigidity into discourse as to conceptual meanings; (d) in mismatches between the assumption activated by the instructions of a connective and mind-stored assumptions, accommodation processes may take place, which are effortful but seek to guarantee the retrieval of cognitive effects from the utterance.