1.1 Beebe, Takahashi, and Uliss-Weltz (1990)Beebe, Leslie, Tomoko Takahashi, and Robin Uliss-Weltz 1990 “Pragmatic Transfer in ESL Refusals.” In Developing Communicative Competence in a Second Language, ed. by Robin C. Scarcella, Elaine Slosberg Andersen, and Stephen Krashen, 55–73. New York: Harper & Row. – Taxonomy and impact

The most influential study on refusals was conducted by Beebe, Takahashi, and Uliss-Weltz (1990)Beebe, Leslie, Tomoko Takahashi, and Robin Uliss-Weltz 1990 “Pragmatic Transfer in ESL Refusals.” In Developing Communicative Competence in a Second Language, ed. by Robin C. Scarcella, Elaine Slosberg Andersen, and Stephen Krashen, 55–73. New York: Harper & Row. (henceforth BTU), who looked at pragmatic transfer of Japanese speakers of English. Using discourse completion tasks (henceforth DCT), they elicit refusals of four types of SAs: ‘requests’, ‘invitations’, ‘offer’, and ‘suggestions’. Based on their findings, they established a taxonomy of refusal strategies or ‘semantic formulae’ (Figure 1), which has since been frequently used. Their taxonomy distinguishes between a direct and an indirect super-category, each consisting of multiple categories and sub-categories. In addition, they include adjuncts to refusals which do not function as refusals themselves. Their choice to differentiate between direct and indirect realizations can be explained with the theoretical environment at the time: Brown and Levinson’s (1987)Brown, Penelope, and Stephen C. Levinson 1987 Politeness. Some Universals in Language Usage. 2nd ed. Cambridge: Cambridge University Press. politeness theory was the most dominant account of politeness, postulating that more indirect SA realizations are more polite than direct ones. More recent approaches to im/politeness and interpersonal pragmatics (see e.g. Haugh, Kádár, and Mills 2013Haugh, Michael, Dániel Z. Kádár, and Sara Mills 2013 “Interpersonal Pragmatics. Issues and Debates.” Journal of Pragmatics 58: 1–11. ; Locher and Graham 2010Locher, Miriam A., and Sage L. Graham eds. 2010 Interpersonal Pragmatics. Berlin: De Gruyter. for trends and current debates) have moved away from this equation and look at a variety of linguistic means for interpersonal effects (see e.g. Part II in Locher and Graham 2010Locher, Miriam A., and Sage L. Graham eds. 2010 Interpersonal Pragmatics. Berlin: De Gruyter. ).

Not only BTU’s classification is still frequently used in analyses of refusals (e.g. Allami and Naeimi 2011Allami, Hamid, and Amin Naeimi 2011 “A Cross-linguistic Study of Refusals: An Analysis of Pragmatic Competence Development in Iranian EFL Learners.” Journal of Pragmatics 43 (1): 385–406. ; Siebold and Busch 2015Siebold, Kathrin, and Hannah Busch 2015 “(No) Need for Clarity – Facework in Spanish and German Refusals.” Journal of Pragmatics 75: 53–68. ; Bella 2014Bella, Spyridoula 2014 “Developing the Ability to Refuse: A Cross-Sectional Study of Greek FL Refusals.” Journal of Pragmatics 61: 35–62. ; Lee 2013Lee, Hanjung 2013 “The Influence of Social Situations on Fluency Difficulty in Korean EFL Learners’ Oral Refusals.” Journal of Pragmatics 50 (1): 168–186. ; Babai Shishavan and Sharifian 2016Babai Shishavan, Homa, and Farzad Sharifian 2016 “The Refusal Speech Act in a Cross-Cultural Perspective: A Study of Iranian English-Language Learners and Anglo-Australian Speakers.” Language & Communication 47: 75–88. ; Guo 2012Guo, Yinling 2012 “Chinese and American Refusal Strategy: A Cross-cultural Approach.” Theory and Practice in Language Studies 2 (2). ; Nikmehr and Jahedi 2014Nikmehr, Armin, and Fateme Jahedi 2014 “The Effect of Status on Refusal Strategies Used by American Native Speakers of English and Iranian EFL University Students.” International Journal of Applied Linguistics & English Literature 3 (5): 91–99.; Sattar, Lah, and Suleiman 2011Sattar, Hiba Qusay Abdul, Salasiah Che Lah, and Raja Rozina Raja Suleiman 2011 “Refusal Strategies in English by Malay University Students.” GEMA Online Journal of Language Studies 11 (3): 69–81.; Félix-Brasdefer 2003Félix-Brasdefer, J. César 2003 “Declining an Invitation. A Cross-Cultural Study of Pragmatic Strategies in American English and Latin American Spanish.” Multilingua – Journal of Cross-Cultural and Interlanguage Communication 22 (3): 225–255. , 2008Félix-Brasdefer, J. César 2008 Politeness in Mexico and the United States. A Contrastive Study of the Realization and Perception of Refusals. Amsterdam: Benjamins. ); their study design itself has been replicated (somewhat modified) in numerous papers, for instance in Chen, Ye, and Zhang (1995)Chen, Xing, Lei Ye, and Yanyin Zhang 1995 “Refusing in Chinese.” In Pragmatics of Chinese as Native and Target Langugage, ed. by Gabriele Kasper, 119–164. Second Language Teaching and Curriculum Center, Honululu: University of Hawai’i Press., Yamagashira (2001)Yamagashira, Hisako 2001 “Pragmatic Transfer in Japanese ESL Refusals.” Retrieved from http://​www​.k​-junshin​.ac​.jp​/juntan​/libhome​/bulletin​/No31​/kiyo​_31​_15​.pdf; last access: 9.6.2015: 259–275., Chang (2011)Chang, Yuh-Fang 2011 “Refusing in a Foreign Language: An Investigation of Problems Encountered by Chinese Learners of English.” Multilingua – Journal of Cross-Cultural and Interlanguage Communication 30 (1): 71–98. , Kwon (2004)Kwon, Jihyun 2004 “Expressing Refusals in Korean and in American English.” Multilingua – Journal of Cross-Cultural and Interlanguage Communication 23 (4): 339–364. , Nelson et al. (2002)Nelson, Gayle L., Joan Carson, Mahmoud Al Batal, and Waguida El Bakary 2002 “Cross-Cultural Pragmatics. Strategy Use in Egyptian Arabic and American English Refusals.” Applied Linguistics 23 (2): 163–189. , Hashemian (2012)Hashemian, Mahmood 2012 “Cross-Cultural Differences and Pragmatic Transfer in English and Persian Refusals.” Journal of Teaching and Language 3 (4): 23–46., and Abed (2011)Abed, Ahmed Qadoury 2011 “Pragmatic Transfer in Iraqi EFL Learners’ Refusals.” International Journal of English Linguistics 1 (2): 166–185. . All of these studies, however, employ either DCTs or role plays.

Despite (or maybe because of) this taxonomy’s popularity, only few modifications and additions have been suggested over the years (e.g. Félix-Brasdefer 2003Félix-Brasdefer, J. César 2003 “Declining an Invitation. A Cross-Cultural Study of Pragmatic Strategies in American English and Latin American Spanish.” Multilingua – Journal of Cross-Cultural and Interlanguage Communication 22 (3): 225–255. ; Salazar Campillo 2009Salazar Campillo, Patricia 2009 “Refusal Strategies. A Proposal from a Sociopragmatic Approach.” Revista Electrónica de Lingüística Aplicada 139–150.) – none of them, however, being substantial.

Since BTU’s taxonomy was designed with the objective of applying it to various languages, it was assumed that it is equally suitable for the analysis of different language varieties. Their categories were consequently taken as the starting point for this study.

1.2Historical development of commissives and directives

Although BTU’s data differs from the one examined here, their findings are a good starting point. Their results are, however, not straight-forward to read. The authors focus on pragmatic transfer, comparing Japanese speakers of English with American native speakers of English, and do not provide a full account of strategy frequencies. They do state that refusals of requests and invitations by American native speakers typically include the strategies ‘regret’ and ‘excuse’, as well as adjuncts – not taken into account in this paper – expressing a positive opinion or gratitude. In the case of offers, ‘no’ and letting the interlocutor off the hook are common. With the exception of ‘no’, these strategies are classified as indirect ones.

While no diachronic studies of refusals have hitherto been conducted, linguists have looked at the development of related SAs: other commissives and directives. What both types have in common is what Searle calls their direction of fit, which expresses “how [the propositional] content is supposed to relate to the world” (1976Searle, John R. 1976 “A Classification of Illocutionary Acts.” Language in Society 5 (1): 1–23. , 4). Commissives and directives have a world-to-word direction of fit: they are attempts to get the world to match their content by demanding, asking for, promising, etc. some future action.

Researchers have mostly grouped strategies in terms of their level of (in)directness. Findings suggests that so called ‘conventionally indirect’ realizations (Blum-Kulka and Olshtain 1984Blum-Kulka, Shoshana, and Elite Olshtain 1984 “Requests and Apologies. A Cross-Cultural Studie of Speech Act Realization Patterns (CCSARP).” Applied Linguistics 5 (3): 196–213. , 201) are the most frequent type for directives in Present Day English (henceforth PDE) (e.g. Aijmer 1996Aijmer, Karin 1996 Conversational Routines in English. Convention and Creativity. London: Longman.; Blum-Kulka and House 1989Blum-Kulka, Shoshana, and Juliane House 1989 “Cross-cultural and Situational Variation in Requesting Behavior.” In Cross-Cultural Pragmatics. Requests and Apologies, ed. by Shoshana Blum-Kulka, Juliane House, and Gabriele Kasper, 123–154. Norwood, NJ: Ablex.) and promises (which are commissives) in 19th century business letters (Del Lungo Camiciotti 2008Del Lungo Camiciotti, Gabriella 2008 “Two Polite Speech Acts from a Diachronic Perspective. Aspects of the Realisation of Requesting and Undertaking Commitments in the Nineteenth Century Commercial Community.” In Speech Acts in the History of English, ed. by Andreas H. Jucker, and Irma Taavitsainen, 115–131. Amsterdam: John Benjamins. ). At earlier periods of the English language, ‘direct strategies’ (Blum-Kulka and Olshtain 1984Blum-Kulka, Shoshana, and Elite Olshtain 1984 “Requests and Apologies. A Cross-Cultural Studie of Speech Act Realization Patterns (CCSARP).” Applied Linguistics 5 (3): 196–213. , 201) appear to have been the most common ones, for instance in the case of binding promises in Middle English (Pakkala-Weckström 2008Pakkala-Weckström, Mari 2008 ““No botmeles bihestes”. Various Ways of Making Binding Promises in Middle English.” In Speech Acts in the History of English, ed. by Andreas H. Jucker, and Irma Taavitsainen, 133–162. Amsterdam: John Benjamins. ) and directives in EModE (Culpeper and Archer 2008Culpeper, Jonathan, and Dawn Archer 2008 “Requests and Directives in Early Modern English Trial Proceedings and Play Texts, 1640–1760.” In Speech Acts in the History of English, ed. by Andreas H. Jucker, and Irma Taavitsainen, 27–44. Amsterdam: John Benjamins. ). Culpeper and Archer (2008Culpeper, Jonathan, and Dawn Archer 2008 “Requests and Directives in Early Modern English Trial Proceedings and Play Texts, 1640–1760.” In Speech Acts in the History of English, ed. by Andreas H. Jucker, and Irma Taavitsainen, 27–44. Amsterdam: John Benjamins. , 72) found that, in the EModE drama texts and trial proceedings they analysed, ¾ of all directives were realized directly by imperatives. What is interesting is that BTU’s categories ‘excuse’ and ‘hook’ are actually non-conventionally rather than conventionally indirect (Blum-Kulka and Olshtain 1984Blum-Kulka, Shoshana, and Elite Olshtain 1984 “Requests and Apologies. A Cross-Cultural Studie of Speech Act Realization Patterns (CCSARP).” Applied Linguistics 5 (3): 196–213. , 201) strategies. This raises the question why the results for refusals differ from the overall preference of conventional realizations for commissives and directives in PDE and where EModE refusals fit in.

1.3Defining refusals

A wide range of SAs can elicit refusals. The first methodological step in the present study was to determine which utterances were to be included. Since it has a diachronic perspective, the tertium comparationis had to be independent of language varieties. Relying on SA labels, or more specifically SA verbs, might have been problematic. All the refusal-eliciting SAs used in previous studies have one thing in common: their world-to-word direction of fit. This provided the basis for delineating what counts as a refusal. The below definition (Reichl 2015Reichl, Isabella 2015 “Refusals. Dispreferred Responses to Pre-Cooperative Illocutionary Acts in Early Modern English Drama.” MA thesis, University of Vienna.) is as broad as necessary in order for it to be applicable to (a) corpus data and (b) data sets from different periods:

1. Conditions applying to the refusal-eliciting SA

   1. the refusal-eliciting SA performed by a speaker 1 (S1) has a world-to-word direction of fit;

   2. the propositional content condition of the eliciting SA is a future action X to be performed by S1 and/or the addressee/speaker 2 (S2) and/or third parties (S3);

   3. the performance of action X (by S1 and/or S2 and/or S3) as put forth by S1 depends on S2’s cooperation;

   4. action X can be a physical, verbal, or mental action;

   5. the intended perlocutionary effect of the eliciting SA hence consists of two parts: first, the cooperation of S2 and, second, the performance of X. The two could, in some cases, happen simultaneously, e.g. if S2 (truthfully) responds to a question, he/she cooperates and, at the same time, performs X;

2. Conditions applying to the refusal

   1. a refusal, performed by S2 is a second-pair part in response to a SA performed by S1;

   2. the response performed by S2 is a denegation of the SA performed by S1;

This definition has a number of advantages: first, the distinction between ‘orders’ and ‘requests’ (Searle and Vanderveken 1985Searle, John R., and Daniel Vanderveken 1985 Foundations of Illocutionary Logic. Cambridge: Cambridge University Press.) is obsolete; if there is a refusal, it is irrelevant whether S1’s eliciting SA intends to give the option to refuse or not; second, questions, i.e. requests for information (Searle 1969, 66; Searle and Vanderveken 1985Searle, John R., and Daniel Vanderveken 1985 Foundations of Illocutionary Logic. Cambridge: Cambridge University Press., 199), are included, while other types of questions such as rhetorical questions or ‘outlouds’ (Stivers and Enfield 2010Stivers, Tanja, and N. J. Enfield 2010 “A Coding Scheme for Question-Response Sequences in Conversation.” Journal of Pragmatics 42: 2620–2626. , 2623) are not. The definition also excludes a number of SAs. Utterances which represent, as Culpeper and Archer (2008Culpeper, Jonathan, and Dawn Archer 2008 “Requests and Directives in Early Modern English Trial Proceedings and Play Texts, 1640–1760.” In Speech Acts in the History of English, ed. by Andreas H. Jucker, and Irma Taavitsainen, 27–44. Amsterdam: John Benjamins. , 59) put it, “[i]ndirect speech acts in which the requestive force is the direct force and not the indirect force”, such as apologising by begging someone pardon, were not considered. SAs in which S1 informs interlocutors about a future action (e.g. holiday plans) are also excluded. Mediated third person SAs, i.e. utterances in which S1 speaks for an S3 but not him/herself, were not included either.

3.1Classification of strategies

In the following subsections, the different super-strategies of the proposed revised classification shall be introduced. In addition, the provided examples will also illustrate the dimension of intrinsically expressed stance.

3.1.1Definite refusals

Group 1 subsumes strategies expressing a definite refusal and consists of two further sub-sets. Strategies in the first sub-group simply express – implicitly or explicitly – that X will not happen. It consists of six strategies, four of which are neutral in terms of stance and two of which express a positive stance.

Example (1) illustrates two neutral categories: [no], consisting of the one-word responses ‘no’ or ‘nay’, and [negA/F] a modified version of BTU’s ‘negative willingness and ability’. In this classification, it includes negations of ability, as shown above, as well as negations of the future act X happening. Neither of the strategies in (1) expresses a stance. The emphasized utterance in (2), by contrast, does. Here, the speaker expresses their hypothetical wish for X to happen, implying it will not. In doing so, they convey a positive stance towards X.

Realizations of the second sub-group, by contrast, provide a reason – implicitly or explicitly – why X will not happen. This subset is the largest one, consisting of 13 strategies: six neutral, six negative, and one positive.

In (3), we see two strategies expressing a positive stance and no stance, respectively. Initially, the speaker states their theoretical willingness to comply under different, hypothetical circumstances ([cond>hyp]). This hypothetical willingness conveys a positive stance. The second strategy constitutes a plain excuse or reason ([exc]) why X will not happen without the speaker taking a stance. An instance of a strategy communicating a reason and a negative stance is given in (4). Using sarcasm ([sarc]), the speaker indirectly expresses a negative stance towards X, simultaneously implying the reason for the refusal: they do not want X to happen.

3.1.2Attempts at dissuasion

The strategies in this second group differ from the previous one in that, rather than expressing a definitive refusal to cooperate, they constitute attempts at dissuading S1. Three negative, two neutral, and one positive strategy form this group.

A neutral dissuasive strategy is exemplified in (5). This new category includes offers of compensation and bribery ([comp]). In the above example, the speaker attempts to dissuade his wife from her request that he stays home, suggesting the decisive power is with the listener rather than the speaker. The strategy is neutral, because it does not reveal a positive or negative feeling towards X or S1. In (6), by contrast, the refuser takes a positive stance towards X by proposing an alternative time and location ([alt]), conveying his overall willingness for X to happen, if in a somewhat modified form. The strategy ‘guilt trip’ ([guilt]), on the other hand, is inherently negative regarding its expressed stance: it consists of utterances intended to make S1 feel guilty in order for them to drop the proposal, as illustrated in Example (7).

3.1.3Attempts at deflection

The final group comprises strategies which constitute refusals by means of deflection, i.e. neither a clear acceptance nor refusal is performed. None of these strategies consequently expresses a stance. The strategy ‘avoidance’ ([avoid]), for instants represents utterances in which S2 evades S1’s proposal. In (8), S2 deliberately builds his response on a word play and so avoids performing an acceptance or refusal.

(8)

All strategies found in the data, and those retained from BTU, can be categorized in terms of one of these super-strategies. The benefits of clearly defined functional sub-classes will be addressed in the subsequent section.

3.2An alternative classification – Differences and advantages

The classification proposed here (Figure 2) differs considerably from BTU’s. The second and third super-strategy are based on their categories ‘attempts to dissuade interlocutor’ and ‘avoidance’, respectively. In contrast to BTU’s dissuasion strategies, however, Group 2 does not include criticism, requests for help, letting the interlocutor of the hook, and self-defence55.This strategy was subsumed under ‘plain excuse’. – all of which are treated here as definite refusals rather than attempts at dissuasion. In addition, proposals of alternatives and lack of enthusiasm were subsumed under Group 2, unlike in BTU, and two new categories were added: compensation and dissuasion. Group 3 differs from BTU’s group ‘avoidance’ as well: first, no non-verbal responses were included; second, jokes and repetitions were subsumed under the new category avoid; finally, indefinite responses were considered deflection as well and the new class ‘questioning the proposal’ was established.

All remaining of BTU’s strategies, if retained, were classified as ‘definite’ refusals – either providing or not providing a reason. Newly created definite strategies are: oppositions and counters, preconditions, exclamations, requests to stop, sarcasm, and hypothetical conditions – split off BTU’s conditions for future or past acceptance, since the former is neutral whereas the latter expresses a positive stance.

Establishing hypothetical conditions as a category on its own was consequently necessitated by the introduction of stance as a classification dimension. The number of sub-categories which are completely new suggests considerable differences between BTU’s data and those analysed here. In terms of super-strategies, their taxonomy proved inconsistent and not fully thought through. They furthermore did not consider that some strategies intrinsically express stance – a shortcoming probably owing to the theoretical environment of their study. This analysis dimension may prove a more useful tool in the analysis of interpersonal effects than in/directness. Similarly, systematically grouping realizations in functional super-strategies may also aid in the analysis of im/politeness or identity construction.

3.3Most frequent strategies – Comparison to BTU’s findings

The usefulness of the classification dimensions proposed in the previous section can be further corroborated by quantitative analyses. In total, the IRs in P1 contained 283 instances of refusal strategies, those in P2 216. An additional 74 and 137 locutions were found in the CRs in P1 and P2, respectively. Overall, excuses are the most common strategy in my data, followed by expressions of criticism (Table 2). The former do, however, increase over time, while the latter decreases – in IRs as well as CRs. A slight decline can also be seen in the frequencies of noes and expressions of oppositions, whereas the percentage of exclamations rises. Interesting, none of these top strategies expresses a positive stance; they are all either neutral or even negative.

As already mentioned, BTU found that ‘regret’, ‘excuse’, ‘no’, and ‘letting the interlocutor off the hook’ are commonly used in Present Day American English refusals. As Table 2 shows, ‘no’ and excuses are common in the drama texts under investigation as well. Most significant, however, is the fact that BTU attest the frequent use of expressions of regret or apologies, which only appear a handful of times in the EModE data. This strategy is considered to express a positive stance in the revised classification – a feature, as mentioned above, not prominent in my findings. Rather, expressions of criticism, which express a negative stance, are the second most common strategy in the drama texts, whereas they do not seem to emerge in BTU’s data. This may, however, not (merely) be a difference in terms of language change over time, but could well have to do with the types of data.

3.4Differences in the use of super-strategies and stance as time and interactions progress

A number of trends with regard to the super-strategies used and stance expressed can be detected when we compare (a) the two periods and (b) the initial refusals with the CRs.

Definite refusals which express a reason make up the majority of utterances. In both the IR and the CRs of both P1 and P2, this super-strategy is used in more than 60% and up to ¾ of the realizations, as Table 3 shows.66.Chi-square yielded a p-value of 0.0004 Looking at the frequencies in the IR compared to the CRs, we see that this super-strategy is more common in CRs than in IRs in both periods. Definite refusals not expressing a reason and deflecting strategies, by contrast, are overall more frequently used in initial refusals than the CRs. For dissuading strategies, no general trend can be detected when comparing IRs and CRs, due to the low number of occurrences. As an interaction progresses, then, strategies conveying a reason for the refusal are used more frequently whereas the number of realizations expressing a definitive refusal without a reason as well as attempts at deflecting decrease compared to the IRs.

While the use of definite strategies expressing a reason remained consistent over time, there is a noteworthy decrease of those not expressing a reason and an increase of deflecting realizations. This partially matches the general trend towards increased indirectness in other commissives and directives, as discussed in the introduction. Where the above results, as well as those of BTU, differ, however, is that non-conventionally indirect rather than conventionally indirect realizations feature most prominently in the case of refusals and – as far as the EModE data is concerned – increased over time. One possible reason for this may be that refusals are second-pair parts, whereas the directives and commissives in the above studies are first-pair parts. There is, however, a methodological aspect to this discrepancy as well. Kohnen (2007, 139) points out that non-conventionally indirect realizations are often ‘hidden’ – and hence not analysed – in corpus studies, because they are not automatically retrievable. Although Culpeper and Archer (2008)Culpeper, Jonathan, and Dawn Archer 2008 “Requests and Directives in Early Modern English Trial Proceedings and Play Texts, 1640–1760.” In Speech Acts in the History of English, ed. by Andreas H. Jucker, and Irma Taavitsainen, 27–44. Amsterdam: John Benjamins. do take into account non-conventionally indirect directives, their differentiation between head acts and support acts means that non-conventionally strategies are treated as ‘second-class realizations’ and often categorized as mere supporting moves rather than directives proper when they occur together with more direct forms. In the approach by BTU, which we followed here, by contrast, any locution expressing a refusal, no matter how (in)directly, was classified as such.77.Adopting the methodology of strategy sequences to the refusal-eliciting SAs, the data revealed that the majority of those realizations are in fact of the so called non-conventionally indirect type, as Table 5 in the appendix shows. In addition, non-conventional strategies are, by far, the predominant group in S1’s continued attempts at persuading S2, making up 74.1% and 90% in P1 and P2, respectively. This points to another question regarding methodology: while it is apparent from BTU’s study design that they elicited only an initial refusal, but no continued attempts, it is unclear whether CPs were taken into account in previous (historical) studies on directives and commissives such as Culpeper and Archer’s (2008)Culpeper, Jonathan, and Dawn Archer 2008 “Requests and Directives in Early Modern English Trial Proceedings and Play Texts, 1640–1760.” In Speech Acts in the History of English, ed. by Andreas H. Jucker, and Irma Taavitsainen, 27–44. Amsterdam: John Benjamins. . If they were, the authors do not explicitly state so and do not compare IPs with CPs.

Regarding the stance expressed in refusal realizations, neutral strategies are overall the most common, as shown in Table 4.88.p-value: 0.506. They amount to roughly 2/3 of all locutions, followed by negative strategies, which make up the final third. Interestingly, strategies expressing a positive stance are hardly used at all in either period. As already addressed briefly, however, the reason for the high amount of negative strategies may lie in the text type under investigation: conflicts are, after all, a central part of drama (2008)‘drama’ 2008 In The Oxford Dictionary of Literary Terms ed. by Chris Baldick. Oxford: Oxford University Press.. A comparison with refusal strategies in other text types of the same periods would hence be highly interesting, but unfortunately available and suitable data is scarce.

The diachronic comparison of expressed stances in refusal realizations does not reveal any substantial developments. Although there is a minimal decrease of neutral and negative strategies and a corresponding increase in positive ones, both in the IRs and CRs, evaluating the meaningfulness of this change is impossible due to the small number of positive strategy tokens. We can, however, detect differences between IRs and CRs once more: as interactions progress, neutral strategies are used more frequently and negative ones less so. The expression of (negative) stance is hence somewhat more common in initial refusals rather than at later stages of an interaction.