Revisiting Sentence Adverbials and Relevance is a most solid, insightful and timely piece of research. It revises a previous, often-quoted relevance-theoretic analysis of five types of sentence adverbials: illocutionary or speech-act (e.g. frankly, confidentially), attitudinal (e.g. (un)fortunately, sadly), evidential (e.g. obviously, seemingly), hearsay (e.g. reportedly, allegedly) and epistemic (e.g. certainly, probably). This analysis is chiefly due to Ifantidou (Reference Ifantidou2001) and, to a lesser extent, other relevance theorists.
Based on embedding tests with conjunctions like if and because, the analysis assessed the (non-)truth-conditional nature of said adverbials, thus attempting to unravel their contribution to utterance content and, ultimately, to comprehension. It argued that evidential, hearsay and epistemic adverbials impact the truth conditions of the propositions expressed by the utterances in which they occur. In contrast, illocutionary and attitudinal adverbials are non-truth-conditional and contribute to what in relevance-theoretic pragmatics is known as higher-level explicatures – that is, the superordinate propositional forms that the mind inferentially creates in order to capture the speech acts that speakers are thought to perform when saying something or their attitude towards what they say. This differing contribution was explained to be a consequence of the former type of adverbials making up independent syntactic and semantic units giving rise to distinct utterances and separate assertive speech acts. Finally, the analysis claimed that the semantics of evidential and epistemic adverbials modulates the strength of the propositions expressed by indicating whether speakers commit themselves to them. In contrast, that of hearsay adverbials causes hearers to take what is said as an interpretation of the beliefs or words of someone other than the speakers, thus reducing or removing their commitment to what is said.
Pandarova makes a rather critical and theoretically informed appraisal of this analysis, which enables her to detect a series of limitations, weaknesses and problems, the most notable of which concerns the mixed results yielded by the tests for truth-conditionality on which it is based. Relying on a good number of bibliographical references, which come not only from the field of pragmatics, but also from those of syntax, semantics and information structure, the author seeks to find answers for the following questions:
(i) Could evidential, hearsay and epistemic adverbials have both truth-conditional and non-truth-conditional interpretations? Would this also be the case of attitudinal adverbials?
(ii) How would (non-)truth-conditional interpretations arise? Are they determined by syntax, semantics or pragmatics?
(iii) If sentence adverbials could be both truth-conditional and non-truth-conditional, what disambiguates them in context?
Responses to these questions are given on the grounds of recent syntactic, information-structural and pragmatic approaches, as well as through a thorough examination of data from English and German. Pandarova proposes that tight integration into clause structure accounts for truth-conditional readings of sentence adverbials. Their non-truth-conditional interpretations are suggested to result either from the loose integration of these adverbials into clause structure as covert parenthetical clauses or from invisibility to c-command relations. Since syntactically parenthetical adverbials may sometimes contribute to the truth-conditional content of utterances or they may not, the author resorts to pragmatics with a view to positing a different explanation from that given by Ifantidou (Reference Ifantidou2001). Succinctly, five factors determine (non-)truth-conditional interpretations of sentence adverbials: their linear position within the sentential structure, the intonation with which they are produced – most importantly, comma intonation – whether they express epistemic, evidential, hearsay or attitudinal qualifications, their information-structural properties and the wider discourse context of the utterances where they appear.
The book is a most comprehensive and well-organised volume. It is structured in seven chapters (pp. 1–12). The first of them is an introduction that presents the origins of the work, its goals and structure. The last one is the conclusion, where the author summarises her main claims and the answers suggested for the problems posed by sentence adverbials (pp. 226–32). As for the bulk of the work, the contents of its other five chapters are detailed in what follows.
Chapter 2, ‘Relevance theory’ (pp. 13–54), very clearly and didactically summarises the main postulates of Sperber & Wilson's (Reference Sperber and Wilson1995) most influential cognitive-pragmatic model, the framework inspiring most of the author's proposals. With a variety of original examples, the first half of the chapter addresses the role of relevance in cognition, the relevance-theoretic view of communication as an ostensive-inferential activity and the notions of manifestness, mutual manifestness and context. Its second half centres on the relevance-theoretic model of verbal communication. It firstly discusses the inferential processes resulting in lower-level explicatures, or the fully propositional forms that the mind constructs by inferentially enriching the logical form obtained through decoding: saturation, free enrichment, ad hoc concept construction and disambiguation. The last of these will subsequently be particularly important in order to understand how the ambivalence of sentence adverbials is resolved. Then, the discussion focuses on higher-level explicatures and the linguistic elements determining their content, which include sentence adverbials. To conclude, the chapter briefly revises the relevance-theoretic conception of implicatures and their types.
The third chapter, ‘Sentence adverbials in relevance theory’ (pp. 55–86), is dedicated to an in-depth review of the previous relevance-theoretic analysis of sentence adverbials. It starts by revising their contribution to truth-conditional content and explicatures. Next, it examines the role of hearsay adverbials as indicators of interpretive use and of evidential adverbials as markers of epistemic strength. It finishes by addressing the double-utterance/speech-act hypothesis and presenting the research questions to which the alternative relevance-theoretic approach that Pandarova will propose must respond.
Chapter 4, ‘An alternative relevance-theoretic account of attitudinal, evidential, hearsay and epistemic adverbials’ (pp. 87–131), puts forth the intended new account. After introducing some initial assumptions and terminological distinctions pertaining to the proposition expressed by utterances containing sentence adverbials, the first part of the chapter examines a series of examples consisting of naturally occurring data. These reveal the unfeasibility of distinguishing between truth-conditional and non-truth-conditional categories of sentence adverbials, and suggest that it is more plausible that they may have truth-conditional and non-truth-conditional uses and interpretations. Why these interpretations originate is addressed in the second part of the chapter, where the author argues that the semantics and pragmatics of epistemic adverbials facilitate representations of epistemic judgements about the truth of propositions, those of evidential adverbials enable representations of the evidence supporting it, those of hearsay adverbials trigger representations about the nature of that evidence, and those of attitudinal adverbials result in representations of the speaker's evaluative attitude towards the truth of propositions. Additionally, she shows that such representations may be used interpretively and concludes that pragmatics is more determinant than semantics as regards the truth-conditional or non-truth-conditional readings of sentence adverbials.
‘Adverbial syntax and (non-)truth-conditionality’ (pp. 132–52) is the fifth chapter and it delves into the role that syntax plays in the possible opposite readings of sentence adverbials. Their truth-conditional interpretations are accounted for on the grounds of syntactic integration in host clauses, while their non-truth-conditional ones are explained as a result of syntactic orphanhood or invisibility to c-command relations. Focusing on their syntactically parenthetical uses, the chapter closes by suggesting that adverbials thus employed do not directly relate to the host clause, but are integrated in some covert parenthetical clause.
The bulk of the work is wrapped up by chapter 6, ‘Syntax and beyond: Explaining (non-)truth-conditional interpretations’ (pp. 153–225). This chapter pursues three goals. The first one is to overcome the weakness of the account presented in the previous chapter: its lack of constraints. Adopting a parsimonious approach, Pandarova suggests that, while syntactically integrated adverbials clearly contribute to the truth-conditional content of the host clause, adverbials occurring in other positions would be ambiguous between truth-conditional and non-truth-conditional readings, an ambiguity that would be resolved at the pragmatic rather than the syntactic level. The second goal is to explain why evidential, epistemic and hearsay adverbials have truth-conditional readings in scope tests, whereas attitudinal adverbials do not. The author relies on information-structural properties like pointhood and at-issueness in order to solve this enigma. The third goal is to unravel what ultimately determines the (non-)truth-conditional interpretations of sentence adverbials. In this regard, she suggests the five aforementioned factors: syntactic placement, prosody, the type of qualification expressed, information-structural properties and context of use.
Overall, this most illuminating, inspiring and valuable contribution to the fields of pragmatics, syntax and information structure is well written and reads smoothly, despite the density of the issues that it deals with. Its author continuously and skilfully signposts throughout it, repeatedly recalls the research questions that are progressively being tackled and efficiently cross-refers to specific sections and chapters, thus adequately guiding the reader and facilitating understanding of her arguments and proposals. In addition to the knowledgeability of specialised terminology, as well as the precision when using it, her dissections and easy-to-follow discussions of the examples illustrating her discussion and claims are most helpful.
By bringing together pragmatic, syntactic and information-structural approaches, Pandarova certainly manages to demonstrate the powerfulness and fruitfulness of collaboration between these three camps. Quite undeniably, this cutting-edge work is mostly theoretical; therefore, it does not include a chapter or section detailing methodological issues such as the data-collection process. Although a brief clarification is made to the effect that the data selected for the examples come from various corpora, at certain points more detailed information about their geographical provenance or the age, gender, education level or socioeconomic class of their producers would have been welcomed. This omission might cause readers to wonder whether the data are solely representative of specific varieties of pluricentric languages like English and German, and, hence, whether the argumentation and the claims made throughout the book would only apply to those varieties. Relatedly, an issue that is not duly clarified is the extent to which the adduced examples evince correct standard usage of sentence adverbials or (highly) idiosyncratic practices.
Although the monograph quite satisfactorily succeeds in ascertaining the effect of the syntactic features and information-structural properties of sentence adverbials on their (non-)truth-conditionality, it unfortunately does not consider two issues that may reveal further functions in these adverbials and lend additional support to the claims made about their (non-)truth-conditionality. The first of such issues concerns their relationship with the so-called epistemic vigilance (e.g. Sperber et al. Reference Sperber, Clément, Heintz, Mascaro, Mercier, Origgi and Wilson2010). The second issue pertains to the potential role of these adverbials as indicators of lexical pragmatic processes (e.g. Wilson Reference Wilson2003).
More than a decade ago, Dan Sperber, Deirdre Wilson and some of their colleagues started to explore the role of vigilance mechanisms in communication. These mechanisms have been argued to assess the reliability of informers and the trustworthiness of information by monitoring and scrutinising a wide variety of linguistic, paralinguistic and nonverbal features determining trust allocation. These include eye gaze, gesturality, hesitation, seeming nervousness, intonation, rephrasing or lexical and/or syntactic mistakes, to name but some. As a result, such mechanisms have been claimed to trigger a critical attitude – namely, epistemic vigilance – which prevents hearers from being blindly gullible.
Ever since then, various linguistic elements occurring in the linguistic input have also been shown to contribute to the activation and tasks of such mechanisms. Among them are evidential participles (Padilla Cruz Reference Padilla Cruz and Piskorska2020); these are verbal participles like alleged, presumed or suspected, which pre-modify nominal heads but lack purely adjectival functions, as illustrated in the following examples:
(1) Alleged killer arrested.
(2) Judge values presumed fraud in 855 millions.
(3) Boy suspected kidnapper dead.
Perhaps epistemic, evidential and hearsay adverbials also fulfil a similar function. In other words, they could alert or assist the tasks of epistemic vigilance mechanisms in a similar fashion to the exemplified participles. If this was the case, the (non-)truth-conditionality of sentence adverbials might also be a consequence of, or be indicated by, their insertion in the linguistic input in order to raise the activation of such mechanisms or assist their tasks. Should this be a plausible hypothesis, it would be interesting to elucidate whether such insertion and function are furthermore accompanied by certain prosodic features, like lexical stress or a particular pitch, which make sentence adverbials more salient, thus contributing to the disambiguation of their possible readings.
As for the second issue, the discussion in chapter 2 regarding the pragmatic processes performed during comprehension deals with ad hoc concept construction. Summing it up, relevance-theoretic pragmatics posits that mental concepts are not static, but malleable entities which are amenable to sense modulation or specification. Accordingly, their denotation may be inferentially adjusted on every occasion of use of the lexical items triggering their activation through two distinct processes: narrowing, or restriction of their denotation, and broadening, or loosening thereof. These processes yield an occasion-specific, perhaps one-off, conceptual representation, which is labelled ‘ad hoc’. Pandarova could have explored whether some of the adverbs heading sentence adverbials might trigger these lexical pragmatic processes.
Indeed, sentence adverbials might behave in a similar manner to the expletives in the following examples:
(4) This damned/bleeding/fucking computer has stopped working again!
(5) This cake is fucking delicious.
In the first case, the expletive does not provide information about the qualities of the computer in question, but rather seems to be added in order to voice the speaker's irritated, angry, fierce and/or scornful attitude towards it. In the second case, in turn, the expletive seems to intensify the quality denoted by the adjective delicious, while adding emotional or attitudinal overtones of surprise, delight, amazement, etc., or an admixture thereof. In both cases, the outcome of the use of these expletives would be ad hoc concepts. In the first case, the expletive would enact the modulation of the encoded concept computer – the small caps conventionally indicating a concept in relevance-theoretic pragmatics – as computer* – the star being conventionally used in relevance-theoretic pragmatics in order to mark an inferentially specified concept – which could be paraphrased as ‘computer towards which the speaker feels scorn’, ‘computer that makes the speaker angry’ or even ‘computer that drives the speaker mad’. In the second case, the expletive would finetune the concept delicious in such a way that it refers to a type of deliciousness that surprises the speaker or exceeds their expectations (Padilla Cruz Reference Padilla Cruz2022).
Conceptual adjustments like these would be possible because of the assumptions that the expletives would make manifest, which could be added to the encyclopaedic entries of the concepts encoded by the lexical items to which they are appended. Moreover, these adjustments could even be triggered by specific paralinguistic and kinesic features (Padilla Cruz Reference Padilla Cruz2023). Similarly, some of the sentence adverbials in the examples that Pandarova discusses could be thought to finetune conceptual components of the decoded logical form, thus giving rise to ad hoc concept construction. If this was the case, enactment of this lexical pragmatic process would reveal the contribution of these adverbials to the truth-conditional content of the utterances where they appear and further contribute to disambiguating their semantics.
Notwithstanding this, Pandarova has undeniably done an excellent job. Quite convincingly, she has suggested a series of criteria enabling the disambiguation of sentence adverbials. Moreover, she has lent support to repeatedly voiced claims concerning the need to integrate approaches from different fields in order to gain better and deeper insights into the interpretation of sentential components. However, the issues connected with sentence adverbials do not appear to be settled at all; rather, there still lies much work ahead to be done. Surely, future efforts by either Pandarova herself or other researchers will contribute to solving them, thus enriching our understanding thereof.
