3.1Methodology

Our main analysis focuses on face-to-face language use in a corpus of TikTok videos, studied through the lens of ritual, speech acts and interaction. We first focus on the behaviour of older participants in our data, and then turn to the behaviour of younger participants.

Our discourse analytic investigation is anchored in interaction ritual (see Kádár 2017Kádár, Dániel Z. 2017 Politeness, Impoliteness and Ritual: Maintaining the Moral Order in Interpersonal Interaction. Cambridge: Cambridge University Press.Kádár, Dániel Z. 2017 Politeness, Impoliteness and Ritual: Maintaining the Moral Order in Interpersonal Interaction. Cambridge: Cambridge University Press., 2024Kádár, Dániel Z. 2024 Ritual and Language. Cambridge University Press: Cambridge. Kádár, Dániel Z. 2024 Ritual and Language. Cambridge University Press: Cambridge. ) : we approach advising in Chinese family settings as a ritual which is not ceremonial and scripted but rather interactionally co-constructed. This approach originates in the seminal work of Goffman (1967)Goffman, Erving 1967 Interaction Ritual: Essays on Face-to-Face Behavior. Garden City: Doubleday.Goffman, Erving 1967 Interaction Ritual: Essays on Face-to-Face Behavior. Garden City: Doubleday. who used ritual to describe interactional order in a variety of practices in daily exchanges. Accordingly, we approach Chinese New Year family advising as a conventionalised and interactionally co-constructed ritual practice. We consider various representative features of interaction ritual:

• Interaction rituals evolve in a frame (Goffman 1974Goffman, Erving 1974 Frame Analysis: An Essay on the Organization of Experience. Cambridge: Harvard University Press.Goffman, Erving 1974 Frame Analysis: An Essay on the Organization of Experience. Cambridge: Harvard University Press.);

• they tend to trigger self-display (Bax 2010Bax, Marcell 2010 “Rituals.” In Historical Pragmatics, ed. by Andreas Jucker, and Irma Taavitsainen, 483–520. Amsterdam: John Benjamins. Bax, Marcell 2010 “Rituals.” In Historical Pragmatics, ed. by Andreas Jucker, and Irma Taavitsainen, 483–520. Amsterdam: John Benjamins. ),

• and they also trigger participation and rapport-enhancement (Kádár 2013Kádár, Dániel Z. 2013 Relational Rituals and Communication: Ritual Interaction in Groups. London: Palgrave Macmillan. Kádár, Dániel Z. 2013 Relational Rituals and Communication: Ritual Interaction in Groups. London: Palgrave Macmillan. ).

The question unavoidably emerges: can face-threatening ritual behaviour be studied on a par with other rituals? Interaction ritual theory provides a clear answer to this question: many social groups use ‘anti-structural’ rituals (Turner 1969Turner, Victor 1969 The Ritual Process. London: Penguin.Turner, Victor 1969 The Ritual Process. London: Penguin.) in which insistent and aggressive behaviour is both accepted and expected. Accordingly, we interpret Chinese New Year advising as a rite of antistructure.

In studying the ritual characteristics of Chinese New Year advising, we rely on the following minimalist and finite typology of speech acts (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, 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. Edmondson, Willis, Juliane House, and Dániel Z. Kádár 2023 Expressions, Speech Acts and Discourse: A Pedagogic Interactional Grammar of English. Cambridge: Cambridge University Press. ), illustrated by Figure 1 below.

Our typology of speech acts is different from classic typologies such as Searle’s, which do not, in our view, offer detailed corpus-based categorial systems where speech acts represent basic human needs across linguacultures. Our system came into existence through the early corpus-based investigation of Edmondson and House (1981)Edmondson, Willis, and Juliane House 1981 Let’s Talk and Talk About It: A Pedagogic Interactional Grammar of English. Munich: Urban & Schwarzenberg.Edmondson, Willis, and Juliane House 1981 Let’s Talk and Talk About It: A Pedagogic Interactional Grammar of English. Munich: Urban & Schwarzenberg., and in Edmondson et al. (2023)Edmondson, Willis, Juliane House, and Dániel Z. Kádár 2023 Expressions, Speech Acts and Discourse: A Pedagogic Interactional Grammar of English. Cambridge: Cambridge University Press. Edmondson, Willis, Juliane House, and Dániel Z. Kádár 2023 Expressions, Speech Acts and Discourse: A Pedagogic Interactional Grammar of English. Cambridge: Cambridge University Press. we tested the replicability of this typology by using large corpora drawn from a variety of typologically distant languages, such as English, German, Chinese, Japanese and Hungarian. Finiteness in our system precludes ‘discovering’ new and cultural- or data-specific speech acts. According to our typology, advising is realised through the speech act Suggest by default, defined as follows:

The Suggest as an illocution is analysed as the case in which a speaker communicates that he is in favour of H’s performing a future action as in H’s own interests(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. Edmondson, Willis, Juliane House, and Dániel Z. Kádár 2023 Expressions, Speech Acts and Discourse: A Pedagogic Interactional Grammar of English. Cambridge: Cambridge University Press. , 126)

Yet, given our bottom-up approach, we do not assume that advising as an interactional behaviour is exclusively realised through Suggests. In conducting such a bottom-up analysis of speech acts, we applied team-coding to validate our individual interpretations of speech acts. In this procedure, our interrater agreement score was 75%. In using our typology, we interpret each utterance as one speech act, differently from conversation analysis. We will define relevant speech act categories whose meaning is not immediately transparent from their labels in Section 4.

In order to capture interactional patterns of advising, we also examine interactional moves through which utterances relate to one another, relying on Edmondson’s (1981)Edmondson, Willis 1981 Spoken Discourse: A Model for Analysis. London: Longman.Edmondson, Willis 1981 Spoken Discourse: A Model for Analysis. London: Longman. classic system. In this system, speech act realisations occur as individual moves, and we examine how moves relate to one another in exchanges, as Initiating, Satisfying, Countering and Contra-ing moves. Initiating refers to speech acts through which an exchange is started, Satisfying includes illocutions through which an Initiating speech act is positively responded to, Countering points to speech acts through which an Initiation is countered but not entirely rejected, whereas if it is definitely turned down it is Contra-ed in our terminology. For example, an Initiating speech act Suggest (“You should find a partner”) may either be Satisfied (“Okay”), Countered (“I’m trying, but…”) or Contra-ed (“I don’t want to have a partner”). Satisfying advising (e.g. with an Okay) does not necessarily imply that the advice is accepted.

Our procedure of combining speech acts with interactional moves allows us to differentiate illocutionary and interactional categories (Edmondson 1981Edmondson, Willis 1981 Spoken Discourse: A Model for Analysis. London: Longman.Edmondson, Willis 1981 Spoken Discourse: A Model for Analysis. London: Longman.). For example, for us the utterance “aren’t your cousins doing well now, unlike you?” is not an ‘act of criticising’ but rather a Request (for information) through which the interactional phenomenon of criticising is talked into being. We break down the interactional phenomenon of advising into speech acts and interactional moves, and we refrain from venturing into terminological niceties about issues such as how advice-giving differs from suggesting, persuading, etc.

The ancillary part of our research consists of an assessment task with follow-up interviews conducted with Chinese evaluators to obtain their views on the behaviour of the participants in our TikTok corpus. These evaluators were presented with two excerpts from our TikTok corpus which illustrated typical cases where older family members provide unsolicited advice for younger family members, and where the younger participants refute this advice either by realising a dismissive Satisfying move or a Counter move. We thus ensured that our evaluators are provided with cases which represent the essence of the phenomenon under investigation. We asked our evaluators to assess the acceptability of the behaviour of both older and younger interactants on a five-point scale. We recruited younger and older participants according to the following age categories (see House and Kádár 2020House, Juliane, and Dániel Z. Kádár 2020 “T/V Pronouns in Global Communication Practices: The Case of IKEA Catalogues Across Linguacultures.” Journal of Pragmatics 161: 1–15. House, Juliane, and Dániel Z. Kádár 2020 “T/V Pronouns in Global Communication Practices: The Case of IKEA Catalogues Across Linguacultures.” Journal of Pragmatics 1611: 1–15. ):

1. Younger evaluators: 18–35 years

2. Older evaluators: over 55 years

We relied on these age criteria to make sure that a clear gap of twenty years exists between the two groups.

3.2Data

Our corpus of face-to-face interactions consists of forty randomly selected TikTok videos featuring advising, self-recorded by younger family members in 2024. We chose TikTok due to its extreme popularity in China. Young people who recorded such interactions often appear not to have told older family members that they video-recorded them, with the consequence that older participants appear to behave naturally. We translated our corpus into English and annotated it by using our speech act typology and system of interactional moves. Our transcripts are arranged in tables, displaying the Chinese original, its English translation, and also speech acts and interactional moves in the same line. Our speech act annotation is based on the Chinese original. Our annotated corpus consists of 11,043 Chinese characters. The interactions represent a public domain and the participants are anonymised, and also we do not reveal the exact source of each conversation for ethical consideration.

Our study primarily focuses on what participants say rather than how viewers react, or technical features such as embedded videos. It is clear that the medium of TikTok shapes the frame of the ritual under investigation through both the functions of editability (the makers of a video can expose the most relevant parts to their viewers) and audience visibility. The TikTok medium therefore may trigger a performative bias in the mediatised scenarios. Yet, in our view this bias does not alter the core tendency of the participants’ behaviour, as evaluations in our assessment task have shown.

For our assessment tasks, we recruited two times thirty evaluators who were also our interview participants, and who represented the above-outlined two age-groups. We asked our participants to give their consent and we anonymised them.

4.1First step: Main analysis

We first considered the basic pragmatic dynamics of our TikTok corpus, examining tendencies of speech act realisation, summarised in Table 2:

As Table 2 shows, older participants realised a larger number and broader variety of speech acts than younger participants, i.e. older participants participated more actively in the interactions than younger speakers. Further, older participants generally realised Substantive speech acts (see Figure 1), including Tell,22.Tell is an illocution through which information is communicated in an ‘objective’ way. Request, Opine,33.An Opine is an illocution through which information is communicated in a ‘subjective’ way. If the content of an utterance is not a fact but an opinion and it is nevertheless formulated in a fact-like way, we interpret it as a Tell. Complain, Suggest and Willing (Offer). Younger participants more often realised speech acts which are defined as Ritual in our typology as follows:

ritual speech acts tend to occur in specific places in an interaction; are, therefore, highly predictable and conventionalised; and have, we may say, a social meaning, such that the literal meaning of the utterance — if any — is almost incidental to the significance of the utterance for both speaker and hearer.(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. Edmondson, Willis, Juliane House, and Dániel Z. Kádár 2023 Expressions, Speech Acts and Discourse: A Pedagogic Interactional Grammar of English. Cambridge: Cambridge University Press. , 45)

Typical Ritual speech acts realised by younger participants include phatic Remarks, Okays indicating agreement, Extractors indicating a strategic anticipation of a Leave-Take, and Wish-wells.44.The label ‘Ritual’ refers to the interactional phases of Opening and Closing which typically trigger ritual language use. However, all speech acts in our system can be realised in a ritual way (see e.g. ritual Apologise).

This pragmatic tendency points to the asymmetric nature of the role relationships between older and young family members in our corpus, as Extract (1) shows:

In turn 1, the aunt overwhelms her niece with advice, urging her to find a boyfriend. In turn 2, the niece responds with a brief Okay (zhīdào-le ‘I know’) as a Satisfying move, without however explicitly accepting the advice (see more below).

Power difference is omnipresent in our corpus, showing that New Year advising is a ritual with roots in Chinese traditions, although the topics and the flow of such interactions may vary. In the following, we consider the key features of New Year advising through the lens of interaction ritual, initially devoting attention to the behaviour of older participants who seem to dominate these interactions, and then turning to the behaviour of the younger participants. All these features were elaborated through our bottom-up procedure, i.e. we identified them through an empirical language-anchored analysis instead of having arrived at these categories at the very outset.

4.2Second step: Ancillary analysis

In our assessment task and follow-up interviews, we asked our evaluators to assess the behaviour of both older and younger participants in two samples drawn from our TikTok corpus. In both cases, the younger participants did not accept the advice: in the first case the young participant realised a dismissive Satisfying move by using zhīdào-le (‘I know’) and in the second case the young participant realised a Counter move.

We first requested our evaluator to assess the behaviour of older participants. Table 5 summarises the outcomes of this task:

Table 5 shows the ritual nature of advising in our corpus: none of our evaluators assessed unsolicited advice by older participants as ‘Unacceptable’, and the majority of evaluators in both groups assessed such behaviour as either ‘Acceptable’ or ‘Clearly acceptable’. In the follow-up interviews, various younger evaluators interpreted unsolicited advice by using evaluations such as bùshūfu dànshì kěyǐ lǐjiě 不舒服但是可以理解 (‘uncomfortable but understandable’), bùkěnéng qù dǐngzhuàng zhǎngbèi 不可能去顶撞长辈 (‘it’s impossible to talk back to older [family members]’) and tāmen shì hǎoyì 他们是好意 (‘they mean well’). Extract (9) illustrates a typical response drawn from an interview with a younger evaluator:

As Extract (9) shows, young evaluators unanimously argued that New Year advising is expected to happen, i.e. it is part of a ritual convention. Also, many evaluators, such as the one cited above, said that unsolicited advising behaviour only occurs during the Chinese New Year, i.e. they were aware of the ritually constrained nature of such advice. At the same time, all evaluators argued that the behaviour of older participants tends to be face-threatening, which accords with the fact that the ritual on hand is a rite of anti-structure.

The assessments of older evaluators accorded with that of younger evaluators, as Extract (10) illustrates:

As Extract (10) shows, various older evaluators referred to New Year advising as a rite of passage, arguing that they themselves were once also recipients of such advice and so young family members should tolerate it for their own benefit. In summary, the first part of our assessment task and follow-up interviews show that our Chinese respondents felt that the behaviour of older-generation participants in the samples is conventional and thus acceptable.

In the second part of our assessment task and interviews, we asked both younger and older evaluators to assess the responses of the younger participants. Table 6 summarises the outcomes of our assessment task:

Young evaluators provided different assessments of the behaviour of young participants in the samples: a majority of them evaluated it as ‘Acceptable’ and ‘Clearly acceptable’, while a minority evaluated it as ‘Unacceptable’ and ‘Clearly unacceptable’. Further, during the interviews even those who provided negative evaluations often defended the ‘Unacceptable’ behaviour of younger participants in the samples, e.g. by arguing that such behaviour is shuōchū le wǒde xīnshēng 说出了我的心声 (‘said exactly what I’ve been thinking’), suīrán wǒ bùgǎn dànshì tài lǐjiě le 虽然我不敢但是太理解了 (‘although I wouldn’t dare to say it myself I completely understand’), and fēicháng rèntóng 非常认同 (‘I totally agree [with such rude behaviour]’). Extract (11) illustrates a typical interview response of a young evaluator:

As this extract shows, our young evaluators often did not themselves find fault with the unsolicited advising of older family members in the samples, unlike those who posted the TikTok videos. However, they did sympathise with the behaviour of younger participants in the excerpts, even though they often found such behaviour unusual. The assessment by older evaluators, on the other hand, was by far most frequently ‘Clearly unacceptable’. Further, older evaluators frequently provided strongly negative assessments of the behaviour of young participants in the samples, such as bù dǒngshì 不懂事 (‘lacking maturity’), wéifǎn chuántǒng 违反传统 (‘violating traditions’) and bù zūnzhòng zhǎngbèi 不尊重长辈 (‘disrespecting elders’). Extract (12) illustrates a typical interview response of an older evaluator:

In summary, the ancillary part of our research helped us interpret the pragmatic dynamics we observed in the first step: while our TikTok corpus represents a conventional Chinese interaction ritual, only the behaviour of older participants can be described as clearly ritual. The behaviour of young participants in our corpus may well be influenced by the mediatised nature of the interactions, encouraging young participants who record these interactions to behave in ways which trigger interest in their behaviour. This might explain why many of our young evaluators found the rejection of advice by their peers unacceptable while at the same time endorsing it.