Abstract
In this introductory paper we outline the aims of the special issue on epistemic and evidential expressions in English across time, the main areas of research within this field and give a summary of the papers included in the special issue, while also highlighting how they contribute to the overarching research aim and how they intersect.
Keywords
1. Introduction
Language is constantly changing, and often these changes are attributed to changes in society (e.g., Michael 2014). This is true of all fields and functions of language, but may be particularly true for the field of interest in the present special issue, namely epistemic and evidential modal marking, as they are often intricately connected to social contexts, participants’ roles, and conventions in a given community of practice. It has been argued that “in general basic structural features [such as indefinite pronouns] tend to have greater stability than pragmatically sensitive features like evidentials, address terms and politeness phenomena” (Trudgill 2020:2, with reference to Wichmann & Holman 2009). For both epistemic and evidential modal marking, changing social conceptualizations of participant relations are likely to be closely related to evolving language practices, and connections between micro-perspectives on social interaction and macro-perspectives on large-scale change seem particularly fruitful, bringing together insights from historical linguistics, diachronic linguistics, and diachronic interactional sociolinguistics.
To understand the reasons behind this proposed close connection between epistemic/evidential marking and social processes, we need to recall the fundamental functions of epistemic modal and evidential expressions. While epistemic modal markers indicate the degree of certainty with which a participant assumes a proposition to be true (cf. e.g., Palmer 2001:8; Van der Auwera, Schalley & Nuyts 2005:201), evidential modal marking indicates the perceptual or epistemological base that a participant asserts to have for the truth of the proposition (cf. e.g., Cornillie 2009) or “the speaker’s/writer’s basis of knowledge as something seen, heard, inferred, or told” (Bednarek 2006:636). Around a quarter of the world’s languages have grammaticalized systems of evidentiality (Aikhenvald & LaPolla 2007:2), which often have an obligatory marking of for example, visual, sensory, and/or reported evidence (cf. Aikhenvald 2004, 2006; Aikhenvald & LaPolla 2007). In such languages, the connection between evidentiality and epistemic modality may not be a very close one, as they function as distinct grammatical-semantic categories. In languages such as English, however, where the expression of evidential marking and of epistemic marking may be understood as optional from a grammatical perspective, a close connection between evidentiality and epistemic modality can be observed in actual usage, as the respective communicative needs, that is, to mark something as uncertain or as based on a less than certain epistemological source, often overlap. In both cases, the user of epistemic/evidential markers can be assumed to use them either based on a lack of complete knowledge, that is, for factual reasons (e.g., She seems to have left/maybe she has left) or for reasons of communicative caution, that is, in order to avoid potential face threats (e.g., There seems to be/may be a problem with your report; cf. also Graefen 2000). Whether or not a speaker/writer feels the need to exercise this kind of communicative caution clearly depends, among other factors, on the conceptualization of their role vis-à-vis their interlocutor, for example, power relations between them. This connection makes the use of these markers very susceptible to social change.
This broader understanding of evidentiality goes back to Chafe and Nichols’ (1986) seminal work who understand evidentiality as “any linguistic expression of attitudes toward knowledge” (Chafe 1986:271). This approach is concerned with expressions through which participants claim how they came to know something (evidentiality), how confident they are about their knowledge (epistemic modality), or if they treat new knowledge as unexpected (mirativity; see also Bednarek 2006). Past text linguistic, pragmatic, and interactional research has often cut across both narrow and broad approaches to evidentiality (e.g., the contributions to Diewald & Smirnova 2010b; Fetzer & Oishi 2014). While in “non-evidential” languages, a participant’s claims of evidential stance are optional from a grammatical point of view, evidential expressions may be obligatory from a discourse-functional perspective, signaling a participant’s “footing” (Goffman 1979). Evidential marking contributes to coherence in discourse (Herlyn 1999) and has been shown to be typical of argumentative contexts where the accountability of the participant is involved (Wooffitt & Alliston 2008). Evidential expressions may serve multiple functions in discourse, for example “index the speaker’s claim to authority, responsibility, and entitlement for a statement” (Fox 2008:187). Speech representation, which includes reported speech and quotations, has textual, discursive, pragmatic, as well as social functions (Brinton 2023). Even in languages with grammatical evidentiality, the choice of evidential markers may be constrained by discourse types and stylistic choices (Aikhenvald 2006).
The forms and functions of epistemic markers (and possible co-occurrences with evidentials) have been analyzed across types of discourse (e.g., Hyland 1994). In conversation analysis, the analytic interest has centered on the linguistic formatting of “knowledge claims that interactants assert, contest and defend in and through turns-at-talk and sequences of interaction” (Heritage 2013:370; cf. also, Raymond & Heritage 2006). For example, a study by Curl and Drew (2008) finds structural differences in requests across types of interactional settings, uncovering different claims of entitlement, with speakers showing that they expect a request will be fulfilled when addressing it to a family member or friend, while patients in telephone calls to doctors mark a lack of such expectations.
Concerning diachronic approaches, the evolution and use of epistemic modal markers have attracted considerable attention in research on the history of English in all periods (e.g., Kytö 1991; Krug 2003; Leech & Smith 2006), as have grammaticalization processes in present day English (PDE) from an interactional, discourse-functional perspective (e.g., Kärkkäinen 2003). Markers of evidentiality, which are less grammaticalized in English, have been tackled to a lesser extent in diachronic studies (but cf. e.g., Simon-Vandenbergen & Aijmer 2007; Aijmer 2009), but recent interactional and typological work has suggested ongoing grammaticalization processes in PDE (Reber 2021; Mélac 2022).
We think that both strands of research addressed here—functional-oriented diachronic linguistics as well as interactional sociolinguistic approaches concerned with recent and ongoing change—are interested in similar research questions, but rarely interact and share their different perspectives. The aim of the special issue is to bring together insights from historical linguistics into the emergence and historical usage of evidential and epistemic markers and insights from interactional, sociolinguistic studies focusing on current processes and recent innovations.
The remainder of the introduction is structured as followed: in section 2, we address the use of epistemic and evidential markers from the perspective of historical linguistics and grammaticalization studies, while section 3 is concerned with interactional approaches to recent change in usage. We end with a summary of the contributions to the special issue (4) and the list of contributions.
2. Emergence and Historical Usage of Evidential and Epistemic Markers
Epistemic and evidential expressions have attracted a lot of attention in English historical linguistics. Originally preterite-present verbs, the modals form a recognizable class of their own already in Old English, and their development from this time period on has been traced in a number of works (e.g., Denison 1993:292-339 or Fischer 2007:159-209). Notably, their loss of full-verb properties, such as the ability to occur with a direct object, and their semantic shifts from lexical to more grammatical meaning have been viewed in the framework of grammaticalization, and the tendency for modal meanings to develop from dynamic modality to deontic modality and (then) to epistemic modality has been understood as evidence of the tendency for subjectification to accompany grammaticalization (Traugott 1989, 1995; Traugott & Dasher 2002), although typological research has since shown that the development of epistemic meanings from deontic ones is cross-linguistically actually rather rare (Narrog 2012). Alongside this focus of research within the framework of grammaticalization on large-scale diachronic trends, numerous studies have focused more on small-scale phenomena, with regard to synchronic historical usage patterns as well as from a diachronic perspective. The modals have perhaps attracted the most attention, for example, in Kytö’s (1991) work on Early Modern English, on the usage of may, might, can, and could. Work on more changes from Early Modern English to present-day English often compares the usage patterns of the modals to those of the semi-modals that emerged in Early Modern English, such as have to (e.g., Krug 2003), and discuss frequency changes, where the modals tend to lose ground to the semi-modals (cf. e.g., Leech & Smith 2006), with epistemic uses, however, declining less drastically than deontic ones (Kranich 2021). Modal adverbs have also been studied in depth, and the emergence of epistemic meanings have been highlighted here as well as results of processes of grammaticalization: for instance, Lenker (2010:117) argues that the Middle English epistemic adverb truly has undergone the cline “clause‑internal adverbial > sentence adverbial > discourse particle” proposed by Elizabeth Closs Traugott (e.g., Traugott & Dasher 2002; see also Brinton 2023).
Concerning evidential meanings, historical linguistics again presents interesting insights from the framework of grammaticalization. Discourse markers represent typical forms of new grammatical evidentials across languages (see Aikhenvald 2004:140-141; Keevallik 2008; Deutscher 2011). Based on the observation that evidential markers can undergo grammaticalization (see, e.g., Diewald & Smirnova 2010a on German), it makes sense to envisage clines of not only more or less grammaticalized individual evidential markers, but also more or less grammaticalized evidentiality as a category. Thus, Reber (2021) suggests that evidential and non-evidential languages should be conceptualized as forming two extremes on a continuum, with English being positioned in between: while at one extreme end of the cline there are languages that have fully-fledged, obligatory evidential marking, such as the ones discussed in Aikhenvald’s work, at the other end, we find languages with purely lexical marking of evidentiality. Languages in the middle part of the cline have some grammaticalized options for marking evidentiality and certain contexts, such as hear-say, where discursively the marking of evidentiality is not fully optional.
3. Interactional Approaches to Recent Change in Epistemic and Evidential Markers and the Impact of Social Change
Fostered by the availability of historical recordings (Reber & Jucker 2023a, 2023b), a research focus on more recent changes has been on the evolution of epistemic and evidential practices in situated spoken interaction (Reber 2021; Aijmer, 2026; Põldvere, 2026). Noticeably, these studies provide evidence for changes in language use within a relatively short time span. A widely observed phenomenon is the emergence of epistemic and evidential pragmatic markers. In their now-classic paper, Thompson and Mulac (1991) demonstrate the grammaticalization of the epistemic matrix construction I think that X (I think that we’re definitely moving toward being more technological) to an epistemic phrase, that is, a parenthetical, I think in conversation. The loss of the complementizer that (I think Ø exercise is really beneficial, to anybody) and the mobility of the construction within the clause (It’s just your point of view what you like to do in your spare time I think.) point to the reanalysis of the verbal construction in terms of a phrase similar to the adverbial usage of maybe. In contrast to the epistemic verbal construction, the meaning of the parenthetical indexes “degree of speaker commitment” (Thompson & Mulac 1991:13).
Recent research in diachronic interactional sociolinguistics has provided evidence of evolving epistemic and evidential markers within decades. In this strand of research, we can identify two perspectives on exploring change in language use and social factors: first, corpus-based approaches, where data from different periods of a representative corpus, for example, the BNC, are compared. The distribution of linguistic structures over time is correlated with social variables of the participants, for example, gender and age. Second, evolving language use is explored within a “community of practice” (Lave & Wenger 1991). Here social processes and other contextual factors are discussed to provide a possible explanation for the changes identified (rather than correlated). The pioneering study in this field is Reber’s (2021) analysis of evolving forms and functions of reported speech with the quotative verb say in British Prime Minister’s Questions during two time periods (1978-1988 and 2003-2013). The results show that there is a sharp increase of reported speech with SAY between the 1978-1988 and 2003-2013 data sets, most prominently in the question-answer sequences between the Leader of the Opposition and Prime Minister, which correlates tightly with a changed composition of participation in the community of practice from 1978 to 2013. Against this backdrop, the analysis provides evidence for qualitative and quantitative changes, ranging from the micro level to more global levels between the 1978-1988 and 2003-2013 data sets: these include (i) changes in the exemplar representations (Bybee & Eddington 2006; Bybee 2013) of reporting clauses and grammaticalization processes in reporting clauses; (ii) changes in the forms, frequencies, and distribution of reported clauses (e.g., an increase and diversification of forms of direct speech) over turn types and participation roles; (iii) changes in the packaging and complexity of rhetorical structures (e.g., lists, contrasts, puzzle-solution) in which reported speech is organized; and (iv) changes in the frequency and use of quotations as building-blocks for action formation in the courses of action between the leader of the opposition and prime minister, which become visible in a sedimentation of recurrent sequences of actions, that is, enticing sequences (cf. Reynolds 2015) and trading-quotes sequences. In more general terms, these findings point, to varying extents, to an increase in a visualized literalization of quotations, in a credibility enhancement, a more interpersonal style, more audience involvement as well as a heightened polarization through speakers’ quoting practices between the 1978-1988 and 2003-2013 periods.
Another vibrant area in current research on modality in English concerns the connection between recent change in usage patterns and a general tendency toward democratization in societies (e.g., Mair 2006; Hiltunen & Loureiro Porto 2020b; and the papers in Hiltunen & Loureiro Porto 2020a and in Kranich, Knewitz, Pirazzini and Bruns 2026). Democratization can be understood from a linguistic perspective as a “rise of more congenial, less face-threatening alternatives in a society apparently more egalitarian, democratic, and antiauthoritarian” and the connected tendency “to avoid unequal and face threatening modes of interaction” (Farrelly & Seoane 2012:393). That means societies undergoing democratization, as many English-speaking societies have, are likely to exhibit changes in language use in many areas, for example, how face-threatening acts are performed and mitigated, but also how facts are communicated, how legal reasoning is performed in court opinions (Reber, 2026), who is believed to have the right to assume the role of expert and to influence other people’s important decisions, and how directly one may communicate these (this latter point being closely connected to the contribution by Põldvere 2026). Kranich, Hampel, and Bruns (2020) and Kranich (2021) have shown how these changes affect modal usage, with more hierarchical modals such as must and may, becoming less frequent in their deontic senses, while Kranich, Bruns, and Hampel (2021), focusing on the use of direct or indirect strategies in requests, found that the assumption of a power position tends to have less impact on younger than on older participants. These findings seem to support the notion that democratization in language, in addition to the general greater attention to face wants mentioned by Farrelly and Seoane (2012:393), furthermore affects the conceptualization of social hierarchies and leads to a more equal treatment of all interlocutors, regardless of whether the interlocutor has more, equal, or less power than a co-participant in a given scenario. Since epistemic and evidential markers bear a close association with interpersonal functions, they are prime candidates for being affected by these sorts of changes. The contributions united in this special issue thus pay particular attention to historical context, communities of practice, and the way the markers are used in specific constructions, which collocations and which common contexts of use they occur in (cf. also Cappelle & Depraetere 2016 for the importance of close attention to individual modal constructions). In addition, the characteristics of the interactions and discourse contexts in which the markers are evidenced will be focused on in the qualitative parts of many of the presented analyses, thus allowing for a better grasp on communicative functions as well as participant roles and their assumptions about the social context, which in turn allows a more in-depth understanding of the functions individual epistemic and evidential markers have in their concrete social and historical contexts. The approaches in the individual studies thus combine insights from the two fields of historical linguistics on the one hand and interactional sociolinguistics on the other. While historical linguistics, in its early days, rather tends to focus on changes in the language system, in particular such that are easier to reconstruct on the basis of only written data, in more recent times, efforts have increasingly seen the light to use data that allow to reconstruct changes in participant interactions, such as evidence from trials (cf. e.g., Grund 2012, 2026; Claridge, Jonsson & Kytö 2024), in order to pay attention to the interactional nature of language use and language change, as also visible in the increasing number of studies in historical pragmatics and politeness (cf. e.g., Jucker 2013, 2020; Brinton 2023). Interactional (socio-)linguistics, on the other hand, originally focused on synchronic research questions (e.g., Gumperz 1982; Schiffrin 1994), as their core interest in the nature of conversational interactions made the use of every-day discourse from the present the most plausible material for their investigations. More recently, however, interactional (socio-)linguistic concepts and methods have also successfully been applied to diachronic data in order to understand better how new constructions arise through interactions (cf. e.g., Reber 2021). It thus seems to us like the right time to bring researchers working within these two fields closer together as their research interests—understanding how language use happens in interaction, how it is shaped by its social contexts and brings about new constructions in interaction—increasingly converge, and this special issue can be seen as a step in this direction.
From a methodological perspective, previous research has hardly tackled the question of how to provide empirical proof for the relation between a particular type of linguistic change and social factors. So far, most linguistic research will simply rely on plausibility (cf., e.g., Leech, Hundt, Mair & Smith 2009:214, who, after stating that one “can only speculate,” refer to their explanation based on social change as “one obvious suggestion”; cf. also Culpeper & Nevala 2012), which does not seem satisfactory. It is, indubitably, a complex problem, as it will always be easier to state with certainty that a social process, such as democratization, and a linguistic change, such as an emergence of a new modal construction, occur in roughly parallel timelines and that a connection seems plausible, than to furnish solid proof of one being the cause of the other. We will not be able to present a final solution to this problem here, either, but we do think that the contributions in this special issue provide advancements into the right direction, through the benefits of combining insights from historical linguistics (with its close attention to historical context and cotext) and from diachronic interactional sociolinguistics (with its close attention to the nature of interactions). Furthermore, the effort to combine more global quantitative results with more qualitative micro-perspectives on specific changes and specific constructions, allows for close scrutiny of context, cotext, communities of practice, and the broader historical background.
4. The Contributions to the Issue
The contributing articles share a common interest in either of the following two major themes:
i. Evolution and usage of epistemic and evidential markers
ii. The impact of social change on epistemic and evidential markers
i. Evolution and Usage of Epistemic and Evidential Markers
Epistemic and evidential expressions are clearly shaped by their interactional and discursive uses. The formal and functional evolution of epistemic and evidential stance markers across specific historical periods is crucial in understanding the emergence and subsequent evolution of these markers. The contributions to this theme share an interest in epistemic and evidential expressions as resources for stance-taking in Early and Late Modern English.
Peter Grund explores evidential resources used in Early and Late Modern English legal records, focusing on the modulation (i.e., negation, degree modification, evidential combinations, and justifications) of seven verbs that mark source of information in English:
Innovative insights on the way epistemic and evidential stance is expressed in legal proceedings also comes from Daniela Landert’s contribution. Her study focuses on adjectives conveying epistemic stance, which, as a word class, are often neglected in the area of modality studies. She shows that while less common than for example, modal verbs, epistemic modal adjectives are nevertheless a very worthwhile area of investigation. Investigating the formal contexts in which modal adjectives occur, the study underlines the limitations of automatic retrieval techniques, as the adjectives in question do not occur in specific formal contexts, which leads to the conclusion that one has to rely on manual retrieval in order to distinguish epistemic from other uses. In the second part of her study, Landert focuses on genre, comparing uses in medical treatises, political tracts, and trial proceedings in Early Modern English, and showing clearly different usage patterns. In trials, for instance, the importance of participant roles in interactions are particularly pertinent, as for example, court officials, witnesses, or defendants exhibit clear differences in their use of stance adjectives.
Elisabeth Reber traces the development of inferential adverbial constructions in the form of adverbs (therefore, hence, consequently), prepositional phrases (for . . . reason(s) . . .) and clausal structures (that(’s) is why, this is why) in the Corpus of US Supreme Court Opinions from Late Modern English to Present Day English. She finds noticeable differences between the forms: for instance, therefore dominates across all periods in the corpus but shows a steady decline. Although much less common in use,
ii. The Impact of Social Change on Epistemic and Evidential Markers
The situated, context-specific emergence of epistemic and evidential markers is highly dependent on social factors, such as the ways in which we visualize the relationship between ourselves and our interlocutors when we speak/write. In this vein, the interrelationship between social change and social interactions and thus the changing choices and usage of epistemic and evidential markers is another key interest of the proposed special issue. The studies subsumed under this theme are characterized by a shared focus on more recent change in spoken British English from the twentieth century to today.
Karin Aijmer’s investigation of obviously focuses on recent change in the use of this marker. Comparing frequencies and meanings of obviously in the spoken components of the original BNC and BNC2014, she outlines the impact of the sociolinguistic variables age and gender on the usage in the two time frames. It becomes evident from her analysis that while the inference meaning was the main one in the earlier time frame, the main function of obviously in the later time frame is that of strong engagement, that is, as a competitor to of course. Her results provide evidence that it is young speakers, in particular female young speakers, who drive this change preferring obviously in this recent function to the more established of course, which, as a result, has sharply declined in frequency in the more recent data. Aijmer explains this change as likely due to a different preferred style in conversations that is characterized by more speaker involvement and where obviously often fulfills the role of highlighting common ground and thus marking solidarity between speaker and hearer.
Finally, Nele Põldvere’s contribution takes a novel approach to changes in the use of the English modals by singling out one speech act, namely advice giving, in order to shed more light on the question whether democratization and the flattening of social hierarchies plays a role in the changes visible in the modal domain, as previously hypothesized (e.g., by Myhill 1997). She shows that in her spoken British English data, it is indeed the more equal and informal advice-giving strategies (such as constructions with
5. List of Contributions
Peter Grund: Negotiating Knowledge: Evidential Verbs and Stance in Early and Late Modern English
Daniela Landert: Stance Adjectives in Early Modern English
Elisabeth Reber: On the Evolution of Inferential Practices in Late Modern English and Present Day English Court Opinions
Karin Aijmer: ‘Yeah obvs. That’s awesome.’ A Short-Term Semantic-Pragmatic Change in Present-Day English
Nele Põldvere: Advice-giving in Spoken English: A Targeted Study of Modality Shifts in one Speech Act
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
