3.1Methodology

We investigate the following research questions:

Following our three research questions, we conducted a tripartite analysis.

In working towards a definition of flattery (RQ1) it is important to consider how flattery relates to interaction ritual. Following Goffman (1967)Goffman, Erving 1967 Interaction Ritual: Essays on Face-to-Face Behavior. New York: Pantheon., we define ritual as a phenomenon which

• operates with conventionalised pragmatic patterns;

• triggers a frame and self-displaying behaviour; and

• triggers intensive and often exaggerated interactional involvement and chains of speech acts.

We break down interactional manifestations of flattery in our corpus into replicable speech acts. We rely on a finite and minimalist speech act typology (Edmondson 1981Edmondson, Willis 1981 Spoken Discourse: A Model for Analysis. London: Longman.; Edmondson and House 1981Edmondson, Willis, and Juliane House 1981 Let’s Talk and Talk About It: A Pedagogic Interactional Grammar of English. Munich: Urban & Schwarzenberg.; Edmondson et al. 2023Edmondson, Willis, Juliane House, and Dániel Z. Kádár 2023 Expressions, Speech Acts and Discourse: A Pedagogic Interactional Grammar of English. Cambridge: Cambridge University Press.), displayed by Figure 1:

Having a finite typology allows us to rigorously study and compare speech acts through which flattery is realised. Ever since Austin and Searle, the idea that speech act categories need to be finite has been present in pragmatics (e.g. Kissine 2013Kissine, Mikhail 2013 From Utterances to Speech Acts. Cambridge: Cambridge University Press. ; Levinson 2017Levinson, Stephen 2017 “Speech Acts.” In Oxford Handbook of Pragmatics, ed. by Huang Yan, 199–216. Oxford: Oxford University Press.). The idea of finiteness precludes ‘discovering’ new and culturally-specific speech acts. The speech acts we propose are such simple and basic constituents of language use that they can easily be replicated in the study of interaction across languages and datatypes (see House and Kádár 2023House, Juliane, and Dániel Z. Kádár 2023 “Speech Acts and Interaction in Second Language Pragmatics: A Position Paper.” Language Teaching 57: 1–12. ). We tested the cross-cultural applicability of the speech act categories with the aid of various corpora, including e.g. English, German, Chinese and Japanese data (see Edmondson et al. 2023Edmondson, Willis, Juliane House, and Dániel Z. Kádár 2023 Expressions, Speech Acts and Discourse: A Pedagogic Interactional Grammar of English. Cambridge: Cambridge University Press.). We can therefore argue that this typology is suitable for the study of speech acts in various linguacultures. This system has two features which preclude defining flattery as a speech act:

1. The rationale behind using a minimalist speech act system is to avoid creating additional speech acts ad libitum. In our view, flattery is a specific interactional phenomenon, and defining it as a speech act would imply that an infinite number of pragmatic phenomena could also be defined as speech acts, which would in turn compromise the replicability of our speech act approach.

2. Different from various conversation analysts (e.g. González-Lloret 2010González-Lloret, Marta 2010 “Conversation Analysis and Speech Act Performance.” In Speech Act Performance: Theoretical, Empirical and Methodological Issues, ed. by Alicia Martínez-Flor, and Esther Usó-Juan, 57–74. Amsterdam: John Benjamins. ), we approach speech act as a unit realised on the utterance level, rather than a phenomenon co-constructed across various turns. Our study will show that historical Chinese flattery tends to be realised through different speech acts and often through a cluster of interactional moves, i.e. it is not an utterance-level phenomenon.

We break down moves into speech acts through which flattery is initiated and responded to. In so doing, we examine the interactional relationship between speech acts, using Edmondson’s system (1981Edmondson, Willis 1981 Spoken Discourse: A Model for Analysis. London: Longman.). According to this system, a speech act (or speech act-sequence) realised as an Initiating interactional move may be Satisfied, ‘Countered’ or ‘Contra-ed’. Satisfying refers to speech acts through which the Initiating speech act is accepted — e.g. in our case Satisfy would be if a speech act through which flattery is Initiated gets bluntly accepted by the recipient. Countering refers to speech acts through which the Initiating speech is objected to but not entirely rejected, whereas if an Initiating utterance is turned down entirely it is ‘Contra-ed’. For example, if an Initiating flattery is only ostensibly refused it is Countered, while if it is genuinely refuted (e.g. “Don’t flatter me!”) it is Contra-ed.

3.2Data

Our corpus consists of 66 excerpts featuring flattering interaction, drawn from the novel Guanchang Xianxing Ji 官场现形记 (‘Officialdom Unmasked’). This satirical novel — written by Li Baojia between 1901 and 1905 — depicts the life of officials in late imperial China. It is often referred to as one of the Wan-Qing Sida Qianze Xiaoshuo 晚清四大谴责小说 (‘Four Great Satirical Novels of the Late Qing Dynasty’), and as such it has been thoroughly studied in Sinology (see Hu’s 2016Hu, Ying 2016 “Late Qing Fiction.” In The Columbia Companion to Modern Chinese Literature, ed. by Kirk A. Denton, 104–110. New York: Columbia University Press. overview). Guanchang Xianxing Ji provides an excellent source for studying flattery because most of the characters in the novel are officials and their family members who aim to get promoted, and in so doing they often engage in flattery (Holoch 1980Holoch, Donald 1980 “A Novel of Setting: ‘The Bureaucrats.’” In The Chinese Novel at the Turn of the Century, ed. by Milena Doleželová-Velingerová, 76–115. Toronto: University of Toronto Press. ).

In Guanchang Xianxing Ji we found six metapragmatic expression-clusters through which the source describes flattery, at Table 1 below shows. The metapragmatic richness of these expressions is noteworthy because it indicates that flattery was in fact an important practice in historical Chinese. Flattery is also mentioned in descriptions of the recipient’s (normally positive) reaction to the other’s flattery (e.g. Extract [1]).

We annotated extracts by using our speech act typology, and later we also translated relevant parts into English. We provided our own translation instead of using the standard Columbia University Press translation.22. https://​cup​.columbia​.edu​/book​/officialdom​-unmasked​/9789622095427 In interpreting speech acts, we relied on interrater analysis, and as part of this we also analysed the ostensible versus genuine nature of flattery in our data. Our transcripts are arranged in tables, displaying the Chinese original and its Pinyin romanisation, the English translation and the speech act annotation in the same line.

We were unavoidably faced with the standard historical pragmatic problem that what we have on hand is literary data, however closely it may reflect spoken language. We agree with Culpeper and Kytö (2000)Culpeper, Jonathan, and Merja Kytö 2000 “Data in Historical Pragmatics: Spoken Interaction (Re)Cast as Writing.” Journal of Historical Pragmatics 1 (2): 175–199. that it is not only valid but also important for the pragmatician to study such historical sources because they provide insight into the operation of historical interactional phenomena. A related issue is that our corpus only represents flattery in historical Chinese during the Qing Dynasty (1644–1911). While it would have been fruitful to conduct a diachronic comparison of sources from various dynasties to examine the historical development of flattery, a problem with conducting such an analysis is that satirical novels like Guanchang Xianxing Ji only occurred during the Qing Dynasty, even though satire itself had appeared much earlier in Chinese (for an overview see Knetchges 1970Knechtges, David 1970 “Wit, Humor, and Satire in Early Chinese Literature (to A.D. 220).” Monumenta Serica 29 (1): 79–98. ).

4.1Defining flattery in historical Chinese

In order to define flattery in historical Chinese, we need to first consider how flattery relates to elevation/denigration and complimenting, by examining speech acts through which these phenomena tend to be realised.

As regards elevation/denigration in Chinese, Kádár et al. (2022)Kádár, Dániel Z., Juliane House, Fengguang Liu, and Lin Jiao 2022 “Self-Denigration in Chinese: An Interactional Speech Act Approach.” Language & Communication 88 (3): 1–15. pointed out with the aid of multiple corpora that elevation/denigration is not simply a Chinese ‘equivalent’ for complimenting behaviour. Rather, the default aim of elevation/denigration is to indicate the relationship between the interactants, and so the use of elevation/denigration is not specific to any particular speech act or cluster of speech acts. The situation is different with flattery. Table 2 summarises the types and frequency of speech acts in our corpus:

Table 2 represents the total number of speech acts in our corpus, with flattery being most often realised by the following three speech acts:

1. Opine: This speech act through which information is presented in a ‘subjective’ tone is by far the most frequent illocution in our corpus. The utterance “Professor Wang, you are the most knowledgeable person I’ve ever known” is a typical flattering Opine because it is formulated in a subjective way.

2. Tell: This speech act through which information is presented in a matter-of-fact way is occasionally used to realise flattery (e.g. Extract [8]). However, such uses are rare. A typical Tell is a ‘neutral’ illocution which gains a flattering function as the speaker describes a situation or achievement which is flattering to the other in a factual manner, and then realises an Opine or Congratulate through which he connects the flattered person with the flatterable. For example, “Professor Wang, only a few scholars are on the forefront of academia. I believe you are one of them.” is a typical Tell→Opine sequence through which flattery is realised.

3. Congratulate: This speech act is called for when the addressee has achieved some success, goal or good fortune, or when the addressee celebrates some event. Congratulate realisations are much rarer than Opines in our corpus. A typical flattering Congratulate would be “Professor Wang, it is simply fantastic that you have become one of the best academics worldwide.”

Other speech acts in Table 2 most often represent moves through which flattery is elicited/responded to.

Extract (1) illustrates a typical flattery realisation:

As Extract (1) shows, flattery is realised via Opines by default. In the interaction above, most utterances relate to an imaginary future event and/or express information the speaker has no way to know, and as such they reflect the speaker’s opinion relating to the future rather than expressing congratulation for something in the here-and-now. Due to the prevalence of Opines, flattery is strongly anchored in one particular speech act, unlike elevation/denigration which can occur in any speech act. Elevation/denigration is nevertheless clearly important in the realisation of flattery: as Extract (1) shows, Prefect Bu uses the elevating address form daren 大人 (‘Your Honour’) four times in flattering Tong Ziliang.

Similar to elevation/denigration, flattery overlaps with complimenting. For us, complimenting is an interactional phenomenon realised through the following three speech act types (see Edmondson et al. 2023Edmondson, Willis, Juliane House, and Dániel Z. Kádár 2023 Expressions, Speech Acts and Discourse: A Pedagogic Interactional Grammar of English. Cambridge: Cambridge University Press.):

1. Opine, as in the case of “You look fantastic today”.

2. Remark (a Phatic speech act), such as “It’s great to see you”, which is a typical compliment realised in Small Talk.

3. Congratulate, such as “It’s amazing how well you performed at your A-levels”.

Phatic Remarks are completely absent from our data. Congratulates are relatively rare (see Table 2) and they tend to be realised together with Opines, whenever the flatterer congratulates the other for an event in the here-and-now and then realises a Congratulate as a departure for bombarding the other with Opines (see Extract [2]). In summary, it is clear that while flattery and complimenting in historical Chinese have an important overlap, complimenting is a more diverse pragmatic behaviour than flattering in terms of speech act realisation.

The question therefore emerges: exactly what is flattery in historical Chinese? To answer this question, it is relevant here to consider two interrelated pragmatic characteristics of flattery, also reflected by Extract (1):

1. Instances of flattery represent exaggerated pragmatic behaviour. As Extract (1) shows, the Opines through which flattery is realised are not warranted by the praisable. For example, here the interactants discuss a trivial administrative matter, and so the Opine through which the recipient is compared to ancient sages is clearly overdone and ostensible. Further, manifestations of flattery often include hypothetical scenarios which are disproportionate with the context: e.g. in Extract (1), the speaker who realises flattery claims that the behaviour of the recipient will trigger the admiration of the imperial court, which is far-fetching since only administrative matters are discussed.

2. Instances of flattery represent intensive interactional engagement. In our corpus, flattery is often realised by chains of speech acts through which the flattered person gets ‘bombarded’ with praise.

As Kádár (2024)Kádár, Dániel Z. 2024 Ritual and Language. Cambridge: Cambridge University Press. argues, in many Chinese interaction rituals — spanning food- and drink-offering to exchanging gifts — exaggerated behaviour and speech act-chaining are preferred, and flattery therefore has archetypical characteristics of ritual. The fact that flattery is not simply an ad hoc phenomenon but rather ritual behaviour in historical Chinese is also often confirmed by the narrative of Guanchang Xianxing Ji. For example, in Extract (1) the text not only explicitly describes Bu’s words as a form of gongwei 恭维 (‘flattery’) but also states that Tong Ziliang was “impressed by such a storm of” praising, implying that Bu’s behaviour represented a conventionalised ritual.

Based on the above, we define flattery as follows: Flattery is a form of ritual in historical Chinese which always operates with exaggerated features and triggers intensive interactional engagement. While some exaggeration characterises flattery in any linguacultural setting, what is special to the context under investigation is the ritual nature of flattery, implying that exaggeration and related intensive interactional involvement are conventionalised and expected in many contexts. This implies that, while flattery may be negatively evaluated by an external observer, for the participants it is actually normative, and this is also why flattery is at the heart of a satirical novel.

We now turn to our second research question, investigating conventional realisation patterns of flattery in our data.

4.2Realisation patterns of flattery

Since flattery is a ritual, it can be expected to follow conventionalised realisation patterns — i.e. types of strategies — similar to other rituals. The study of flattery realisation patterns in our corpus has resulted in the following typology, displayed by Table 3:

As Table 3 shows, strategies of flattery can be divided into two clusters: flattery is either realised as an Initiating move by a single participant (unidirectional flattery) or by both participants (bidirectional flattery), or as a Satisfying move in an ‘elicited’ way. It is noteworthy that unidirectional flattery as an Initiating move only represents 36.3% of flattery realisations in our corpus. This tendency shows that flattery in historical Chinese is often a more elaborate interaction ritual phenomenon than simply A flattering B.

In the following, we provide an overview of strategies in the typology above.

4.3Responses to flattery

We now turn to our third research question, i.e. whether flattery in our corpus triggers conventionalised response patterns — and, if yes, what do these patterns look like? Here we again compare flattery with both complimenting and elevation/denigration, except that we now focus on response patterns to these phenomena.

As Table 4 shows, flattery triggers the following response patterns in our corpus:

It is significant that flattery in the vast majority of cases is simply accepted by the flattered person. This tendency is different from how both complimenting and other-elevation is conventionally responded to in Chinese (see Section 2). While examples like (1) above have already provided cases where flattery is accepted through a Satisfying move, Extract (7) illustrates this behaviour in more detail:

Here a newly rich person Liu Hou, who essentially purchased an official rank for money, is being flattered by a group represented by a paid mediator Fat Aunty Huang. The members of this group want Liu to use his position to assist them in their promotion. Initially, Fat Aunty Huang realises flattery through a chain of Opines, praising Liu as a person who could easily help others to get promoted, and then she realises a Suggest to one of the group members who paid her to arrange a meeting with Liu, advising him to put himself at the care of Liu. Through the Suggest, Fat Aunty Huang boosts the flattery realised by Opines, attributing authority to Liu over the others. In response, Liu Satisfies the flattery by realising Tells and Opines through which he outlines that he actually can provide the requestable, exaggerating his influence on higher decision makers, i.e. he bluntly accepts the flattery. In response, Fat Aunty Huang re-Initiates flattering Liu who appears to be very pleased by the flattery.

The reason why flattery is often accepted in our corpus can be explained by the fact that it does not represent a merely phatic but rather a goal-oriented form of ritual interaction. This interpretation is supported by the fact that whenever flattery is modestly responded to in our corpus, such responses tend to be ostensible

1. either in the sense that they pave the way for counter-flattery, as in Extract (3) above,

2. or the modesty in the response is simply false.

Extract (8) illustrates this second type of ostensible flattery-response:

The interaction above represents a conversation between a high-ranking lady in a rich family, Lady Bao, and Madam Qu who comes from a lower-ranking family and wants Lady Bao to help her husband. Madam Qu realises flattery through a Tell and an Opine, representing an exaggerated compliment: Qu is much older than Bao but she claims that she would like to become Bao’s daughter. In response, Bao makes an ostensible Counter move, arguing that she is willing to accept Qu as her daughter although age difference between them would make this unusual. What makes this responsive move ostensible is that Bao accepts the unrealistic element of Qu’s flattery, i.e. that Qu would become her daughter. In response, Qu Counters the Counter move, first by realising a ritual Complain, then continuing to flatter Bao, and finally realising a Willing (Offer).

As Table 4 shows, there are only few cases in our corpus where the recipient of flattery switches to another topic. While topic-switch can indicate embarrassment in many contexts, it does not have such a meaning in our corpus. Rather, as Extract (9) shows, topic-switches often simply indicate that the flattered person accepts the flattery and then moves on to another topic:

Here the flattered person Governor-General Wen elicits flattery through what we defined as the strategy of presenting/requesting confirmation of the flatterable. His subordinate realises flattering through Opines, and in response Wen switches topic, i.e. he does not modestly Counter the flattery but simply accepts it by Re-Initiating the interaction through a topic-switch.

In summary, our analysis of flattery-response patterns shows that flattery in our corpus represents ritual behaviour. Such behaviour is different from both conventional compliment-responses and other-elevation in Chinese which are expected to be responded in a modest/self-denigrating way. While flattery may be responded to with (often ostensible) modesty, such responses are often used to elicit further flattery. For instance, in Extract (3) the initial seemingly ‘innocent’ exchange leads to intensive flattering, ending with the Request realisations featured in Extract (10):