Modelling stance adverbs in grammatical theory: tackling heterogeneity with Functional Discourse Grammar

Highlights

• There is no such thing as a well-defined class of stance adverbs in linguistic theory.

• Different types of stance adverbs can be distinguished on the basis of clear criteria.

• FDG allows us to tackle the heterogeneity of stance adverbs by offering a unified analysis.

Abstract

In many linguistic approaches and theories, a distinction is made between adverbs that are propositional (representational, ideational, referential) and adverbs that are non-propositional (interpersonal, (inter)subjective, evaluative, parenthetical); i.e. between what is said (the proposition expressed) and a speaker’s stance on what is said (e.g. the speaker’s attitude towards, evaluation of, or commitment to the message conveyed). As is well-known, however, the latter group, including such diverse adverbs as briefly, frankly, sadly, allegedly, hopefully, probably, obviously and cleverly, differ substantially in terms of their discourse-pragmatic or semantic function, truth-conditionality and syntactic and prosodic behaviour. This paper addresses the question of whether, given this functional and formal heterogeneity, it is helpful, or even possible, to recognize a category of stance adverbs. It is argued that our understanding of these adverbs will benefit from a stricter application of very specific functional and formal criteria, allowing for the distinction of a number of well-defined subclasses of adverbs with partly overlapping, but nevertheless defining functional and formal properties. Subsequently, it is demonstrated how the theory of Functional Discourse Grammar (FDG), a theory characterized by a “form-oriented function-to-form” approach (Hengeveld and Mackenzie, 2008: 38–39), making use of different levels and layers of linguistic analysis, can be used to bring out both the similarities and the differences between the various subclasses of stance adverbs without having to resort to a single dichotomy (e.g. ideational vs. interpersonal; truth-conditional vs. non-truth-conditional, propositional vs. non-propositional; parenthetical vs. non-parenthetical). In addition, it is shown how important concepts and distinctions in the literature on (inter)subjectivity, such as the distinction between semantic and pragmatic subjectivity, and between subjectivity and intersubjectivity, can be dealt with within the framework of FDG.

Introduction

In many linguistic approaches and theories, particularly those that are functionally or cognitively oriented, a distinction is made between adverbs that are propositional (representational, ideational, referential, adjunctive), and adverbs that are non-propositional (interpersonal, (inter)subjective, evaluative, parenthetical, disjunctive); i.e. between what is said (the proposition expressed) and a speaker’s comments on what is said (e.g. the speaker’s stance on, attitude towards, evaluation of, or commitment to the message conveyed). As is well-known, however, the group of stance adverbs, including such diverse adverbs as briefly, frankly, sadly, allegedly, hopefully, probably, obviously and cleverly, is far from homogeneous, not only when it comes to their exact discourse-pragmatic or semantic function, but also in terms of their truth-conditionality and their syntactic and prosodic features.

This paper will address the question of whether, given this functional and formal heterogeneity, it is helpful, or even possible, to define a category of stance adverbs (and, by extension, of stance expressions in general).¹ It will be argued that rather than relying on ill-defined notions like stance adverbs, our understanding of the adverbs in question will benefit from a stricter application of a number of well-defined functional and formal criteria. This, it will be shown, will result in the distinction of a number of well-defined subclasses of adverbs with partly overlapping, but nevertheless distinctive functional and formal properties, which together make up ‘a complicated network of similarities overlapping and criss-crossing’ (Wittgenstein, 1953, Section 66).

For the purpose of this classification, use will be made of the theory of Functional Discourse Grammar (FDG), a theory that is characterized by a ‘form-oriented function-to-form’ approach (Hengeveld and Mackenzie, 2008, Keizer, 2015) and which makes use of different, interactive levels and layers of linguistic analysis. On the basis of a number of specific functional and formal criteria, each subclass of stance adverbs will be assigned to a particular layer of representation. Such an analysis, it will be argued, may help to bring out both the similarities and the differences between them without having to resort to a single dichotomy (representational (ideational) vs. interpersonal; truth-conditional vs. non-truth-conditional, propositional vs. non-propositional; syntactically parenthetical vs. syntactically non-parenthetical). In addition, it will be made clear how important concepts and distinctions in the literature on (inter)subjectivity, such as the distinction between semantic and pragmatic subjectivity, and between subjectivity and intersubjectivity, are dealt with within the framework of FDG.

This paper is organized as follows. Section 2 is devoted to the notions stance and subjectivity in general, and the heterogeneity involved in these notions. Section 3 provides a brief introduction to Functional Discourse Grammar; particular attention is paid to those aspects of the theory relevant to the analysis of stance adverbs, such as the different levels and layers of representation. Subsequently, Section 4 introduces and applies the various criteria (discourse-pragmatic, semantic, syntactic and prosodic) used for assigning a particular subgroup of adverbs to a particular level and layer of analysis. In this section, use is made of data retrieved from two corpora of contemporary English (the British National Corpus (Davies, 2004) and the Corpus of Contemporary American English (Davies, 2008)), with occasional examples from the NOW Corpus (Davies, 2016) and the LDC Fisher Corpus (Cieri et al., 2004). Section 5 demonstrates how the different subclasses of stance verbs are analysed within FDG. Section 6 concludes the paper.