By means of the intra-clausal and inter-clausal positions of MSA, the diversity of stance involved is known.Hunston & Francis (2000）indicate that patterns add value to the creation of meaning.Goldberg (1995）approaches patterns through construction grammar.Hoey (1991）explores patterns of lexis as bonds of nets in texts.Halliday & Matthiessen (2004: 43）also acknowledge the significance of patterns.The intra-clausal and inter-clausal modality supplementing patterns through MSAs demonstrate that modal devices can be combined flexibly in relation to clausal positions to realize the stance of discourse.Figure 5.2 sums up the realization of stance by means of intra-clausal and inter-clausal positions of MSAs.
Figure 5.2 Realization of stance by the intra-clausal and inter-clausal positions of MSAs
5.2.2 Ways of modality supplementing
To a large extent, MSAs co-occur with MVs, depending on the stance conveyed by the latter.According to Coates (1983: 26), central MVs can be used both deontically and epistemically, but the percentages differ somehow.Will is mostly used for prediction, and secondarily for volition.Thus, will is normally epistemic, expressing attitudes or judgments.Shall is mostly used for prediction and secondarily for volition.Figure 5.3 sketches Coates’ argument.
For Coates (ibid), each central MV has a primary use, and even a secondary use and an infrequent use.This means that MVs are polysemous in many cases.Coates (1983: 14-16）mentions that there are three types of indeterminacy as related to MVs: (a）gradience; (b）ambiguity); (c）merger.Unfortunately, she has not explored how such indeterminacy can be reduced.Chapter 4 mentions that MSAs can perform the function of disambiguating modal vagueness.Coates (ibid）seems to think about the basic meanings of central MVs more than the effects of contextual factors on the meanings of MVs.
In fact, the distributions of MVs are related to a number of contextual factors and the stance conveyed.Yang (2006a）studies the distributions of central MVs in contemporary academic writings and legal documents, with a corpus of 182,902 words for the former genre and a corpus of 236,378 words for the latter genre.Yang’s central MVs are based on Biber et al.(1999: 73), where there are nine central MVs, i.e.can, could, may, might, must, shall, should, will and would.
Yang (ibid）argues that the MV will is used frequently in academic writings in order to express the writer’s or speaker’s intentions about the research, and predictions about the future, and that in legal documents shall tends to denote liabilities, regulations and forbiddance, and may generally indicates rights or conveys forbiddance when negated.Hence, the same MVs (e.g.shall and may）can have different distributions in the two genres.
Given the different modal meanings that each central MV may represent in different genres, it is reasonable to state that the combinations between MSAs and MVs vary from genre to genre and show differences in the stance involved.Gotti (2004: 328-329）studies will and shall as means of prediction in early modern English, and finds that will and shall denoting assurance are often accompanied by adverbs as content-oriented boosters or emphasizers of the certainty of the interlocutor’s acceptance of the validity of the proposition, e.g.will verily, shall indeed, will surely, and shall sooner than perhaps.
The following part will demonstrate the stance revealed by the combinations between MSAs and MVs in two ways, namely, the double-leveled and multiple-leveled modality, and overt and covert inter-participant modality supplementing.
5.2.2.1 Double-leveled and multiple-leveled modality
Sometimes there is only an MSA for the MV involved.At other times, there are two or even more MSAs for a single MV.The former and latter cases may be termed as double-leveled modality and multiple-leveled modality respectively.The stance involved varies in this sense.
Hyland (2000: 111）studies metadiscourse in academic writing, holding that metadiscourse can perform a variety of functions.In fact, Hyland’s list of metadiscourse includes some adverbs with a likelihood of becoming MSAs.Table 5.1 shows a variety of functions that potential MSAs can display as interpersonal metadiscourse in academic writing.
Table 5.1 Functions of potential MSAs as interpersonal metadiscourse in academic writing
(after Hyland, 2000: 111)
CategoryFunctionExamples
HedgesWithhold writer’s full commitment to statementsperhaps, possibly
BoostersEmphasize force or writer’s certainty in messagein fact, definitely
Attitude markersExpress writer’s attitude to propositional markersunfortunately
Relational markersExplicitly refer to or build relationship with readerfrankly
Table 5.1 shows that MSAs can function in many ways, for instance, withholding writer’s full commitment to statements.Sometimes, in a clause, besides the interpersonal meaning conveyed by the MV, an MSA is needed for another level of interpersonal meaning, be the two interpersonal meanings the same or different.Then, there appear double levels of modality.
Consider the following:
(193）I can’t actually confirm the precise details.
(194）I honestly can’t remember where it was.
In Examples (193）and (194), the negated MV can’t conveys an inability of acting or dynamic modality.The MSAs actually and honestly are comment adjuncts, conveying the addressers’ attitudes to the propositions involved, and emphasizing an inability of acting or dynamic modality.Seen in this way, the interpersonal functions of the MVs and MSAs in the two examples are the same in nature but reinforce each other.The stance in these two examples is that the addressers convey an inability of doing something.The MSAs strengthen the stance revealed by the MVs.