Indirectness in the Age of Globalization: A Social Network Analysis

Размещено на http://www.allbest.ru/

Indirectness in the Age of Globalization: A Social Network Analysis

Marina Terkourafi

Leiden University Centre for Linguistics

Abstract

Indirectness has traditionally been viewed as commensurate with politeness and attributed to the speaker's wish to avoid imposition and/or otherwise strategically manipulate the addressee. Despite these theoretical predictions, a number of studies have documented the solidarity-building and identity-constituting functions of indirectness. Bringing these studies together, Terkourafi 2014 proposed an expanded view of the functions of indirect speech, which crucially emphasizes the role of the addressee and the importance of network ties. This article focuses on what happens when such network ties become loosened, as a result of processes of urbanization and globalization. Drawing on examples from African American English and Chinese, it is argued that these processes produce a need for increased explicitness, which drives speakers (and listeners) away from indirectness. This claim is further supported diachronically, by changes in British English politeness that coincide with the rise of the individual Self. These empirical findings have implications for im/politeness theorizing and theory-building more generally, calling attention to how the socio-historical context of our research necessarily influences the theories we end up building.

Keywords: default indirectness; weak network ties; African American women; compliment responses; Chinese; English historical (im)politeness

Косвенность в эпоху глобализации: анализ сетевых связей

Марина Теркурафи

Центр лингвистики Лейденского университета

Аннотация

Косвенность традиционно рассматривается в связи с вежливостью и объясняется желанием говорящего избегать воздействия на адресата. Несмотря на эти теоретические предпосылки, в ряде исследований отмечаются функции косвенности, направленные на выражение солидарности и формирование идентичности. На основе анализа данных исследований был предложен расширенный взгляд на функции косвенности (Terkourafi 2014), в котором подчеркивается роль адресата и важность сетевых связей между собеседниками. В данной статье рассматривается вопрос о том, что происходит, когда в результате процессов урбанизации и глобализации сетевые связи между коммуникантами ослабевают. Опираясь на примеры из китайского языка и афроамериканского английского, автор утверждает, что эти процессы приводят к необходимости большей ясности в ущерб косвенности. Данное утверждение подтверждается также диахроническим анализом изменений в британско-английской вежливости, которые совпадают с усилением индивидуального Я. Полученные результаты имеют значение для теории вежливости в целом и подтверждают необходимость учета социально-исторического контекста при разработке теоретических положений.

Ключевые слова: косвенность по умолчанию, слабые сетевые связи. афроамериканские женщины, ответы на комплимент, китайский язык, английская (не)вежливость

Для цитирования:

Terkourafi, Marina (2019). Indirectness in the age of globalization: A social network analysis. Russian Journal of Linguistics, 23 (4), 930--949.doi: 10.22363/2687-0088-2019-23-4-930-949.

Introduction

By several accounts, ours is the age of globalization. Google's Ngram Viewer, an online search engine that charts the frequencies of words in printed sources from 1500 to 2008 using several text corpora, provides us with a first indication of this. As Figure 1 shows, starting in the 1980's, use of the term sky-rocketed, reaching an all-time high in the early years of the 21^st century.

Figure 1.Frequency of use of terms for "globalization" in Google's database of English books

(s/z spellings shown separately; retrieved from https://books.google.com/ngrams/ on 23 September 2019)

Despite its recent popularity, the term “globalization” is actually almost a century old, having been introduced in the 1930's in the field of finance to describe “the development of an increasingly integrated global economy marked especially by free trade, free flow of capital, and the tapping of cheaper foreign labor markets” (OED, s.v. globalization). This meaning is still found in economics textbooks, where globalization is defined as “[t]he phenomenon of growing economic linkages among countries” (Krugman & Wells 2013: 128). The recent expansion of the term beyond economics is not unrelated to these macroeconomic processes. In combination with technological advances such as satellite communications, the advent of the internet, and the availability of multiple means of cheap travel, which have in turn enhanced transnational mobility, and enabled mass-scale travel and tourism, these processes are jointly responsible for the unprecedented interconnectedness of our lives in the early 21^st century. This expanded understanding of the term is reflected in definitions of globalization as “an ever-increasing abundance of global connections and our understanding of them” (Barker 2012: 156) and as “the extension, intensification, and acceleration of consequential worldwide interconnections” (Sparke 2013: 3).

This increased interconnectedness has distinct cultural consequences. These consequences have been variably described as homogenization or even Americanization -- when globalization is seen as a one-way influence from US American models to the rest of the world -- but can also involve the creation of new cultural resources, artistic trends, and language varieties emerging from the dialogue of the global with the local (Ter- kourafi 2010a).In whatever way one thinks of globalization, one thing is certain: the average person in most parts of the world today connects with more people and with more diverse people than ever before -- whether one is geographically mobile oneself or not, social media and the transnational movement of people and goods can be counted on for this.

This exponential increase in the number and diversity of people with whom we interact can also affect the quality of our interactions with them. More contacts between people can also mean shallower contacts, especially if we subscribe to the idea known as “Dunbar's number”, after the British anthropologist who proposed it. According to Dunbar (1992), there is a cognitive limit to the number of people with whom we can maintain stable social relationships, and this number is around 150. Most of us have more than this number of “friends” in our Facebook accounts! This raises the distinct possibility that the new connections enabled by globalization are qualitatively different from the dense and multiplex ties of the social networks of the past. The change in the structure of our social networks brought about by globalization plays a central role in the argument developed in this article. default indirectness african american woman

Briefly, the argument goes as follows. While we may as a result of globalization connect with more people, these people are scattered across communities and we tend to know each person in one capacity only. This means that our social network ties with them are distinctly weak. This weakening “Weakening” is not meant to suggest that the same network ties were strong and became weak but rather refers to the overall picture that emerges from the three case studies analyzed below. As all three studies refer to both an earlier state in which strong ties were in place and a later state in which social network ties are weak, what emerges can be characterized as a weakening over time although that does not concern the same networks or the same individual's networks. can result in a shrinking of the common ground necessary to support indirect modes of communication, resulting in a shift toward more explicitness in conversation.

The article is structured as follows. Section 2 introduces the idea that, further to its strategic uses, indirectness can function recognitionally among members of a group. This notion of indirectness “by accident” is central to the argument developed in this article. Section 3 then goes on to outline the principles of social network theory from the field of sociolinguistics, highlighting how globalization can be one way of bringing about a weakening of social network ties. Section 4 lays out in more detail the hypothesis to be explored, while sections 5, 6 and 7 explore this hypothesis with reference to three case-studies from the existing literature on indirectness. These case-studies are drawn from different cultural and historical contexts: African American women's use of indirectness, Chinese responses to compliments, and the rise of conventionally indirect requests in British English in the 19^th c. These three case studies were chosen because they each involve a weakening, in different ways, of social network ties and allow us to observe the impact of this weakening on conversational styles. Compared with using studies from a single lingua-culture, the fact that these case-studies come from different cultural, linguistic and historical contexts yet still exhibit similar trends toward increased directness motivates us to seek a unified explanation for the observed changes, and is thus an asset of this meta-analysis. Finally, section 8 draws some implications of this analysis for the study of variation and change in pragmatics more generally.

Indirectness

Indirectness in pragmatics has generally been defined as meaning something more or something different from what we say. This typically involves generating an implicature by violating or flouting one of Grice's conversational maxims (Grice 1975). Indirectness thus understood is thought to be effortful for the addressee and used strategically by the speaker to achieve certain ends, which can include politeness (Brown & Levinson 1987), deniability (Pinker et al. 2008), expressing formality (Lakoff 1973), or stylistic prowess (Leech 1983). Note that this definition of indirectness is a formal rather than psychological one: it includes all those instances where what the speaker means is different from what her words mean (the literal meaning of an expression or `what is said') The notion of what is said is an especially thorny one and it is unfortunate that the classic notion of implicature is defined in relation to it. For a summary of some dominant positions about what is said, see Terkourafi 2010b., irrespective of whether what she means is immediately transparent to her listener(s) (e.g., because of convention or habit).

The above definition of indirectness has been claimed to capture only a small part of the full range of functions of indirectness in discourse (Terkourafi 2014). An expanded view of indirectness would also include “accidental” as well as “enabling” uses, with the latter capturing instances where indirectness might be the only option for the speaker, such as need-statements which are found among children's earliest directives yet are formally indirect, or metaphors used to describe physical or emotional pain, for which there is often no literal counterpart (Figure 2).

Figure 2.Proposed nomenclature for indirect speech (from Terkourafi 2014: 66)

Of particular interest to the hypothesis explored in this article is the notion of indirectness “by accident” (see Figure 2), which refers to an utterance that becomes indirect `in the ears of the listener,' so to speak, irrespective of what the speaker intended by it or was trying to achieve. This kind of indirectness has more to do with community and situational norms than with the speaker's intention and occurs when community members are attuned to reading more into each other's words than what is literally there. Hall (1976) defines such cultures as “high context cultures”. In such cases, indirectness becomes an interpretative reflex, or the default, expected state in conversation.

While rationally based indirectness, the type of indirectness calculated according to the Gricean maxims as explained at the start of this section, is an individual-level phenomenon, indirectness as default is a community-level phenomenon. It is not (always) intentional or agentive but rather an element of the habitus, which Bourdieu (1990) describes as follows:

The genesis of a system of works or practices generated by the same habitus (or homologous habitus, such as those that underlie the unity of the life-style of a group or a class) [...] arises from the necessary yet unpredictable confrontation between the habitus and an event that can exercise a pertinent incitement on the habitus only if the latter snatchesit from the contingency of the accidental and constitutes it as a problem by applying to it the very principles of its solution. (Bourdieu 1990: 55; emphasis added)

In the three case studies analyzed below, the event that “exercise[s] a pertinent incitement on the habitus” is indirectness, which, to the extent that it is automatically read into speakers' utterances by listeners, can produce amplified interpretations that are taken for granted by speakers as well as listeners and can, in their mundanity, serve as a `secret handshake' ensuring mutual recognition between them as cultural insiders.

It is this type of indirectness “by accident”, which presupposes socialization and functions as cultural credentials of community belonging, that is likely to be affected -- or rather, has no opportunity to develop -- when social network ties become weakened, as can be the case under globalization (see the next section). Although this type of indirectness as communicative style has often been talked about in national-ethnic terms (e.g., Tannen 1981 for Greek, Hall & Hall 1990 for Germans, the French and Americans), the relevant “communities” in today's world may be best understood as Communities of Practice (Lave & Wenger 1991) bound together by practices rather than pre-conceived attributes. As Terkourafi (2019) writes,

As national cultures are [... being...] redefined on new grounds (Pew Research Center 2017), variables other than country of origin, such as generational cohort, political affiliation, or professional expertise are gaining momentum in generating shared understandings and like-mindedness among people. The importance of these other factors explains at least in part why it is difficult to extrapolate from group norms to individual behaviours since “any results found [by aggregating data at the nation-level] cannot be applied to individuals living within these nations” (Fischer 2011: 5). An enactment view of culture provides an answer to this problem. In the practice-based understanding of culture advocated here, group-belonging is not presumed based on external attributes (e.g., nationality) but rather built from the bottom up, through specific behaviours and their having been interpreted in particular ways. (Terkourafi 2019: 1203--1204, original emphasis)

Indirectness “by accident” as defined above and exemplified in the case studies that follow can be one among many practices that bind a community of practice together, making a tightly knit social network part and parcel of this particular type of indirectness.

Social network theory

In social network theory, as applied to sociolinguistics by Lesley Milroy (1987) and further developed by, among others, Fagyal et al. (2010), our patterns of interaction with people we know form our social network, whose strength can be measured along two dimensions. Social network ties are dense in case the people we know also know each other, and they are multiplex if we interact with the same person in several different capacities (e.g., the same person is our neighbour, our co-worker, someone we socialize with, and a member of our family). Figure 3 provides diagrammatic representations of these two different measures of network strength.

Figure 3. Diagrammatic representation of an individual X's social network ties

Black lines indicate ties between individuals. Blue arrows on the left are used to show that the individuals that X knows also know each other, so X's network ties on the left are dense. The red dotted lines on the right indicate that X's tie with Z is multiplex (X and Z know each other in many capacities)

As these definitions make clear, the sheer number of people we know and the frequency with which we interact with each of them are irrelevant to the density and multiplicity of our ties. Spending a lot of time with the same person in the same capacity does not make our tie with that person a multiplex one, nor does knowing a lot of people who do not know each other make our social network ties denser. And although if a tie is multiplex this can result in increased frequency of contact, the opposite is not necessarily true: increased frequency of contact between two people, if it is always in the same capacity, does not increase the multiplicity of their tie.

Social Network theory maps speakers' patterns of interaction at a certain moment in time from an external observer's perspective rather than the participants' own. Thus, the fact that traditional network ties such as those of family or locality may now be conducted online and facilitated by the internet and social media does not mean that these ties are strengthened by becoming denser or more multiplex. They have simply been transferred to a new medium but this does not necessarily affect their strength, as this is measured in social network theory through the constructs of density and multiplicity, because these require different things: they require that the people we know also know each other and that we know each one in several different capacities. On the other hand, the possibility to meet new and more diverse people through both transnational movement and the internet does have specific consequences for the strength of these new ties. To the extent that these new acquaintances do not also know each other and that interaction with each of them remains limited to one rather than multiple capacities, the ties forged are distinctly weak.

Social network theory aims, centrally, to explain language change. It does this by using patterns of interaction between people to explain differences in the distribution of linguistic variants among groups which are otherwise hard to explain because the groups are indistinguishable in terms of the abstract macro-social categories to which they are assigned by the researcher (e.g. differences between groups of working-class people, or between groups of women). Clearly, not all working-class people speak in the same way, nor do all women. Early sociolinguistic work, however, could not do justice to this within-category variability. Social network theory claims, in this regard, that the stronger one's social network, the more vernacular norms they will display in their speech. Conversely, individuals whose social network ties are weak will not display a high proportion of vernacular norms in their speech but will tend to gravitate outward, toward more standard speech norms. The hypothesis forwarded in this article is inspired by this claim and applies it to a pragmatic phenomenon, the use of indirectness “by accident”.

The interface of social network ties and indirectness

This article explores the impact of globalization on conversational resources and styles of interaction and specifically the extent to which conversational styles may be shifting toward increased directness in response to the weakening of social network ties caused by globalization. The particular hypothesis I wish to explore is that, when social network ties become less multiplex and less dense, the common ground between interlocutors correspondingly shrinks and this can have an adverse effect on the use of indirect modes of communication. A shift can then be expected to occur toward more explicit modes of communication because the common ground necessary to support indirectness is now missing.

The relevant notion of common ground is taken from Clark (1986), who uses this term to refer to all the beliefs we can reasonably expect to share with others; in other words, what both parties in conversation, rightly or wrongly, take it that the other also knows. Common ground is a necessary prerequisite for communication and includes both personal common ground, that is, information we share with a particular person because of our direct personal experience with that person (as between friends), but also communal common ground, that is, information we share with someone (whom we may have never met before) because we take them to be members of a cultural community, where cultural communities are identifiable by their shared expertise. Because shared expertise is graded, ranging from central (assumed to be shared among all insiders) to peripheral (assumed to be shared by only some insiders), the contents of communal common ground also range from information about human nature (which we all have) through communal lexicons and information about cultural facts, norms, procedures (which can still be explained to an outsideras knowledge `that') to ineffable background information, which is knowledge `how' that cannot (easily) be explained to an outsider but must be experienced directly many times. This is characterized by Clark (1996: 110) as the “ultimate inside information” and it is the `know-how' that only community insiders can be expected to have. The type of indirectness “by accident” introduced in section 2 falls under this type of knowledge.

The idea that a weakening in social network ties can lead to a decrease in this type of indirectness is theoretically interesting because it is diametrically opposed to what received views about indirectness would have us expect. According to such views, as encapsulated for instance in the first-wave of politeness studies (Lakoff 1973, Leech 1983, Brown & Levinson 1987), under circumstances of weakened social network ties (e.g., increased Distance between interlocutors), indirectness, as a face-saving strategy affording interlocutors a convenient `out,' should be prized and expected to increase. Contrary to this, the hypothesis now put forward is that indirectness is disabled and less available as an option among interlocutors who do not know each other well. Nevertheless, these two predictions are not necessarily at odds. Clearly, what are concerned are two different types of indirectness -- individual-level or “strategic” vs. community-level or “by accident” -- and once the differences between them, as explained in section 2, are understood, both may capture different aspects of the same complex phenomenon. Thus, while this hypothesis is in contrast with the face-saving understanding of indirectness in first-wave politeness studies, which perceived indirectness as exclusively strategic, it is in line with more recent analyses that highlight the variability in interpretations of indirectness across cultures and emphasize that understanding it as the `safest' strategy is valid only for a narrow socio-cultural bandwidth of primarily Anglo speech styles (Grainger and Mills 2016).

A similar argument about the impact of social network structure has been made regarding not language use but language structure. In recent work, Raviv et al. (2019) investigated whether user group size affects the systematicity of a language, with linguistic structure being explicitly defined as encoding a meaning consistently via the same form rather than variably via different forms. It emerged that smaller groups could afford more complex, less transparent (one-to-many) mappings of meanings to forms because their members got to know each other better. As the number of people with whom one uses a language grows, that complexity is lost and the mapping from form to meaning in that language becomes more transparent (or systematic). In other words, the language gains in transparency of form-to-meaning mappings when it is spoken between more people, who correspondingly interact less frequently with each other. Considering that indirect speech reduces the transparency of form-to-meaning mapping, these findings are in line with the hypothesis presented here.

Before proceeding to present the empirical evidence for this hypothesis, two caveats are in order. First, it should be made clear that the claimed weakening in social network ties does not concern all of our network ties (as noted in section 3, the strength of some ties may remain unaffected), nor is globalization the only way in which such a weakening can come about. Globalization is just one way in which a weakening of social network ties can come about; urbanization is another, and there may be others. As such, what we are concerned with in this article is what happens to conversational styles when network ties are weakened in general and not only as a result of globalization. Indeed, of the three case studies discussed below, only one (Chinese compliment responses) can be seen as related to globalization, while the other two relate rather to the weakening of social network ties in the face of urbanization in the recent and more distant past. My goal, then, is not to provide definitive evidence for the specific hypothesis that globalization can result in increased explicitness in conversation but merely to establish this as a plausible hypothesis by showing how a weakening of network ties in general has led to a similar loss in indirectness in other contexts. This is not the final but rather the first step in that process, while the specific hypothesis concerning globalization clearly remains to be further investigated by means of empirical studies designed specifically for this purpose.

Furthermore, while it will be argued that, as a result of a weakening of social network ties, increased explicitness can be expected, it is not thereby also argued that this is incontrovertibly the result of either parallel but internally-motivated cultural responses to increased interconnectedness of people and ideas or of the spread of mainstream American models in a dichotomous fashion. In view of the fact that increased explicitness is also a feature of the more `inductive' -- as opposed to `deductive' -- conversational styles associated with the US (or a generalized `West'; Scollon & Scollon 1995), it is possible that the shifts toward increased explicitness documented below are, at least in some cases, a cultural borrowing from mainstream American-dominated discourses.As globalization has been used to refer to both processes -- the spread of American ways of doing things and the creation of new cultural resources in response to the needs of life in an interconnected world -- this possibility does not weaken the hypothesized relationship between globalization and increased explicitness in conversational styles. On the contrary, to the extent that the two processes are facets of globalization that operate in tandem, being aware that they can lead to similar effects only adds to our reasons to expect that these effects will occur. In other words, the present article is open to a multiple causality account of the relevant phenomena.

Younger and older African American women's indirectness in 1970's Chicago

For close to a decade, between 1974 and 1982, Morgan (1991) studied the conversational styles of three generations of African American women aged 18--72 living in Chicago, focusing on two types of indirectness, both potentially confrontational, identified in previous studies of African American discourse. Baited indirectness occurred when a speaker said something general which was taken by the audience to be specific or addressed to a specific person because of contextual clues. Pointed indirectness, on the other hand, occurred when a speaker said something to someone which appeared opaque or irrelevant in the current context and involved a `sham receiver' different from the intended target, who could in turn be identified based on contextual clues.

An example of baited indirectness occurred between two friends, when one invited the other over for a meal saying:

(1) “if you're not going to be doing anything, come by. I'm going to cook some chit'lins. (rather jokingly) or are you one of those Negroes who don't eat chitl'ins A contracted form of “chitterlings,” a type of food made from the cheaper parts of an animal, such as the small intestines of a pig.? (cited from Morgan 1991: 427--428)

to which the other responded rather indignantly that she had had enough of this “soul food” in her life and was only going to eat better quality food from now on. As the speaker's subsequent discourse after the recipient had left the room made clear, this indignant response was taken by her as evidence that the recipient had read the underlined part of her utterance as an indirect criticism, an interpretation she was prepared to own up to despite not having intended her utterance in this way.

An example of pointed indirectness, on the other hand, is given in (2):

(2) A woman chose to wear an overly bright shade of lipstick to a party. She overheard a[nother] woman say, “Oh, I thought your mouth was burst.” to a man whose lips were in perfect order. (cited from Morgan 1991: 429--430)

In this case, the woman's utterance was clearly not referencing her male addressee and it is this incongruency between her utterance and the physical reality of the addressee that led the overhearing woman to interpret the utterance as a criticism of herself. This is an example of what Morgan calls “if the shoe fits”: what matters is not so much whether the speaker intended her utterance as a criticism as the fact that it was so interpreted, suggesting that it touched a chord with the overhearing woman who took it that way.

In both examples (1) and (2) interpretation prioritizes community norms: what the audience understands counts as the utterance's meaning that stays on the conversational record, irrespective of what the speaker may have meant by her utterance. Such uses of indirectness, according to Morgan (1991; see also Morgan 2010), constitute a distinct African American identity; they are part of the oral tradition of signifying through which speakers construct -- both by performing and by recognizing it -- their belonging in this community. As such, the goal that these uses serve is not primarily informational but rather an identity-related one.

Even more interesting for our purposes is the fact that these intuitions were not uniformly distributed across the three generations of African American women that Morgan studied. As part of her project, she showed adaptations of actual stories like the exchanges above containing baited and pointed indirectness to African American women of different ages and to white women and asked them questions about what the speaker meant and the target of her remarks. African American women in their 40's or older recognized both types of indirectness and were open to ambiguity and communally attributed interpretations of the speaker's utterance, similar to the interpretations of the participants in the examples above. By contrast, African American women younger than 25 rejected these communally attributed meanings, despite recognizing them. As one of these younger women put it, she didn't like to be responsible for “...anybody who comes along accusing me of saying something I didn't actually say” (Morgan 1991: 440). In this respect, these younger African American women behaved more like the white women in Morgan's sample, who likewise prioritized what the speaker herself meant. However, as Morgan is quick to point out, a crucial difference exists between the younger African American women and the white women from the same community: the white women focused exclusively on what the speaker meant and were unable to locate any further implicated meanings; the younger African American women, on the other hand, perceived the indirectness but resisted it.

According to Morgan, the difference between African American and white women's systems of communication is that the latter give priority to what the speaker means and find it permissible to reconstruct speakers' intentions. For African American women, however, speaker intentions are much less prominent and responsibility encompasses not just those meanings which a speaker wishes to be credited with but also those which her audience attributes to her words -- whether she intended those meanings or not. Speakers must choose their words carefully because they carry responsibility for everything their words can be understood to mean -- speakers can't turn around and deny their words later. Younger African American women fall somewhere in between. Notably, their interpretations, like those of white women, prioritize the speaker's intention and speaker intentionality as the ultimate arbitrator of what an utterance means.

This case study of women's speech in the 1970's in an urban setting shows community-specific modes of indirectness being rejected by the younger generation and a rise in individualism going hand in hand with a preference for speaker-based interpretations. That these were community-specific modes of indirectness is shown by the fact that white women living in the same area did not recognize them. And that their demise was a result of a weakening of social network ties in the sense of a loss in density and multiplicity is suggested by the answer of the young woman, who, despite recognizing the indirectness, refused to be held accountable for it. Her emphatic defense of her individuality is typical of those less connected to the core of their communities, who have correspondingly less to lose by distancing themselves from community norms One may recall how, in Milroy's (1987) study of three neighbourhoods in Belfast, it was those with the strongest social networks (denser and more multiplex ties in their community), whether they were men or women, that were vernacular norm enforcers..

Morgan allows for this possibility when she writes that “another explanation for the young women's responses to the survey is that [they.. .are] opting for the non-African American system. They may perceive their role and relationship to society as one where “hidden” forms of discourse are unsuitable” (1991: 441) The other explanation Morgan considers is incomplete socialization (1991: 440).. Morgan leaves it open whether contact with white women, which would imply a weakening in the younger women's social network ties as they would now be interacting with members of different communities who did not mutually know each other (loss in density), and potentially only in some capacities (loss in multiplicity), was a factor in this development but her analysis certainly suggests so.

Chinese Compliment Responses

A loss of more indirect ways of communicating can be incurred not only by urbanization but also by globalization. This is shown by our second case study, which involves Chinese responses to compliments. Compliment responses pose a well-known problem for politeness theorists, which can be formulated as follows within Leech's maxim-based framework (Leech 2014). In this framework, politeness generally involves giving value to Other while withholding value from Self. To Agree, then, with the compliment is to elevate Other's assessment (and hence be polite) but also to elevate Self (and hence be impolite); while to Reject the compliment is to lower Self (which is polite) but also to lower Other's assessment (which is impolite). In other words, responding to a compliment creates a clash between the two components of politeness, raising Other and lowering Self, which cannot both be satisfied at the same time. Different cultures resolve this clash differently: Anglo-cultures tend to Agree, Chinese cultures tend to Reject.

Of these two strategies, the Reject strategy is indirect, since the speaker who uses this strategy is trying to communicate not disagreement per se but rather, through this disagreement, modesty. In other words, the recipient of the compliment doesn't reject it out of disagreement with the speaker (which would amount to lowering Other and hence be impolite) but in order to appear modest (which amounts to lowering Self and is polite). Only if the complimenter recognizes the disagreement as `fake' and motivated by modesty can the Reject strategy be perceived as polite. The Agree strategy, on the other hand, simply accepts the compliment and does not communicate this additional layer of meaning. As a compliment response strategy, then, Agree is not indirect.

Compliment responses are a widely studied phenomenon in Chinese that has attracted scholars' attention for some time (see, e.g., Chen 1993, Wang & Tsai 2003, Chen & Yang 2010, He 2012, among others). These works have identified three response strategies to compliments in Chinese: (i) Acceptance, (ii) Rejection, and (iii) Amended acceptance, which amounts to reformulating the compliment before accepting it. Generally, up until 2000 studies of Chinese compliment responses find an equal split between the Accept and Reject strategies or alternatively a preference for the Reject strategy (for instance, this was 95% in Chen 1993). This preference, however, shifts from 2010 onwards, when the Accept strategy begins to overtake the other two. Figure 4 charts the results of three different studies (Yuan 2002, Chen & Yang 2010, and He 2012), showing, in the case of He using an apparent-time methodology that compares younger to older generations, a shift from the Reject strategy to the Agree strategy Pre- and post-OPCG are the terms used by He (2012) to refer to the older and younger genera-tions in her data, using the introduction of the one-child policy, which required parents (with some exceptions) to have only one child, by the Chinese government in 1979 as a conventional cut-off point.. Using a range of methodologies including Discourse Completion Tests (DCTs), corpora, spontaneous observation, and exit interviews, different researchers have argued for a shift from the Reject strategy in older generations (before 1980) to the Accept strategy in younger ones.

Figure 4. Comparison of Compliment Response Strategies in three studies

(from He 2012: 46; pre-OPCG and post-OPCG stand for pre-One-Child-Policy-Generation and post-One-Child-Policy-Generation respectively)

The Reject strategy in Chinese compliment responses, which was analyzed above as indirect, has been attributed to the Chinese norm of modesty, reflecting Confucian ethical norms. On the other hand, Accept has been attributed to a rise in individualism by some researchers (He 2012), to which others add foreign (American) influence (Chen & Yang 2010). Lin, Woodfield & Ren (2012) found a similar increase in compliments overall (both implicit and explicit), as well as a predominance of compliments about appearance, and went on to associate both trends with Western influence. Still others urge caution in interpreting these findings pointing out that the presence of third parties seems to make a difference (the Agree strategy is hardly ever used in the presence of third parties) and that methodology also has an effect: specifically, DCTs may boost acceptance (Xia, Yin & Lan 2017). Clearly, a closer look at a greater range of data from both bipartite and multi-party exchanges and using different methodologies is needed to settle this.

Meanwhile, several researchers have suggested that a rise in individualism and self-confidence in generations from 1980 onwards goes hand in hand with the aban-donment of an indirect strategy (Reject), which was dictated by and interpretable within community norms, in favor of a more transparent one (Agree), possibly under Western influence and as a result of exposure to foreign norms. As He (2012) writes,

The finding that the post-OCPG [one-child-policy-generation] are much more likely to accept a compliment is an indication that their norms of politeness have deviated from those observed by the older generation. And it seems to indicate that, perhaps due to the dual influence of Western cultural values (cf.Chen and Yang 2010) and China's emphasis on individual aspirations and attainments under the reform policies, the younger generation have become more concerned about presenting a new self-image and identity by displaying self-confidence and individualism through accepting compliments. (He 2012: 47)

Taking these comments about an increase in self-confidence and individualism to reflect a weakening of social network ties, as it is hard to see how these traits could have occurred in the context of Chinese society without such a weakening, we can link the observed loss of indirectness in the retreat of the Reject strategy to a change, specifically a weakening, in social network structure. While more should be known about the networks of the subjects in He's (2012) and Chen & Yang's (2010) studies to confirm this, this interpretation gains some traction from a comment about the subjects in Chen's earlier study, namely that since the subjects in his 1993 study lived in a place that was relatively closed to the outside world, they “probably represented the traditional social values such as modesty” (Chen & Yang 2010: 1959). This comment links the indirect Reject strategy with the stronger social network ties that tend to exist in a place “relatively closed to the outside world”, leaving open the possibility that once a community opens itself up to the outside world, a loss in communal norms of indirectness can be expected to occur.

From discernment directives to conventionally indirect requests in English

The opening sentence of Leslie Hartley's 1953 novel `The Go-Between' reads: “The past is a foreign country: they do things differently there” Credit goes to Larry Horn, who brought this sentence to my attention.. This sentence is an apt reminder that similar shifts in conversational styles to those noted above, which are attributable to contemporary processes of urbanization and globalization, have also occurred in previous historical periods that underwent a loosening of network ties.

The transition from the medieval to the modern period in England offers an example of this. Medieval England was characterized by a fixed social hierarchy, geographical immobility and strong community belonging. As Jucker (2012: 177) notes, “belonging to the network of the society and having good relations in this network were central values.” By contrast, “the idea of an individual's psychological wants [was] not relevant” in this socio-historical context (Culpeper & Demmen 2011: 52). This was then a period of strong social network ties. According to Jucker (2012: 424), three types of politeness systems can be distinguished universally, of which a language may instantiate only some at a particular place and time. The first type is discernment politeness, which is not concerned with face-threat mitigation but relates to socially appropriate behavior guided by socio-cultural conventions. The second type is deference politeness, which relates to the use of titles and honorifics and thus overlaps with the discernment type; while the third type is non-imposition politeness, which employs strategies that give the addressee a choice and explicitly express non-intrusion and non-coerciveness on the addressee's wishes. Based on honorific use during this period, Jucker (2012: 184) argues that Medieval England is characterized by `discernment' politeness. Researchers have identified four different forms of Old English directives that were used during this period: directive performatives such as “I ask you to...”; constructions with a second person pronoun plus scealt/sculon; constructions with uton (= let's) plus infinitive; and impersonal constructions with (neod)pearf (= it is necessary for x) (Jucker 2012: 179). Of these, the first two were used by those in a position of authority over their addressees, while the last two were common among those not in a position of authority, for instance, in religious contexts where humility and obedience discouraged the display of authority. By indicating one's place in the social hierarchy, discernment politeness (as also proposed by Hill et al. 1986 and Ide 1989) thus reinforces the social hierarchy that it reflects.

Things began to change from the 16^th century onwards. Initially, the rise of the bourgeoisie (and concomitant notions of courtly behavior or curteisie) and a loosening of the grip of religion and fatalistic acceptance of one's place in the world led to social mobility (Terkourafi 2011). Then, as technological progress took hold, industrialization and urbanization accelerated geographical mobility. Both types of mobility led to a breakdown of established social networks. During this period, the older speaker-based forms that foregrounded the speaker's authority or sincerity as grounds for granting the request began to be displaced by addressee-based forms (if you please) that foreground the listener's right to non-imposition (Jucker 2012: 188). Supporting this claim, Culpeper and Archer (2008) found that if you please requests were among the most frequent in a corpus of English texts from 1640--1760.

By the 19^th century, under the influence of Romanticism, political, and economic liberalism, individualism became a positive value and was actively pursued in Victorian England. According to Culpeper and Demmen (2011: 61), “as social ties became weakened, the notion of privacy became stronger, and acquired positive value in the Victorian period. Th[is] notion is ... of course related to negative face”. It is during this period that the conventionally indirect request forms Canyou/Couldyou ...? prevalent in present-day English started gaining ground. In written records, these occurred first in trial contexts in the 19^th century and spread from the 1900's onwards (Culpeper & Demmen 2011). These requests are arguably less indirect than the Old English forms which foregrounded the speaker's authority or sincerity as grounds for granting the request. This is especially true of requests by those not in a position of power, who in previous times would have used either inclusive (`let's') or impersonal (`it is necessary') forms for their requests leaving implicit the identity of the person(s) responsible for bringing about the content of the request.Canyou/Couldyou...? forms, on the other hand, specifically reference the addressee and make explicit that this responsibility lies with them. They are therefore more transparent in this regard.

The rise of the Can you/Could you .? request forms signals a shift from deference politeness to non-imposition politeness in (British) English. Like requests using please during the previous period, these ability requests orient to people's individual abilities and rights to non-imposition, reflecting the period's emphasis on the individual, which was a distinct outcome of the socio-historical conditions in 19^th century England. This third case study, then, provides a further example where an attested weakening in social network ties (always in the sense of a loss in density and multiplicity) is accompanied by a shift toward less indirect, more transparent ways of speaking, this time from a historical perspective.

Summary and theoretical implications

The three case studies reviewed in the previous sections -- African American women's indirectness, Chinese responses to compliments, and the rise of can you/could you...? requests in Victorian England -- show community-specific conversational norms reflecting discernment being displaced by more transparent forms reflecting (and respecting) interlocutors' individualism. Researchers have explained these shifts as the outcome of forces of urbanization and globalization affecting the relevant communities at a specific time. In social network theory, urbanization and globalization can be accounted for in a unified way as a loosening of network ties, which, I hypothesized, promotes increased explicitness in conversational styles. The three case studies discussed in this article support this hypothesis.

This social network explanation for the shifts in pragmatic strategies noted above aligns pragmatic change (change in conversational styles) with other types of language change, that have been shown to be affected by social network structure. Specifically, researchers have posited that peripheral members in a network (`loners') are needed to introduce new forms and central members (`leaders') to diffuse them and help stabilize the use of some forms over others (Fagyal et al. 2010). That in the case of the shifts in pragmatic strategies discussed above it is specifically indirect forms that are replaced by more explicit and transparent ones is an interesting addition to this framework from the perspective of pragmatic change. This suggestion seems reasonable if we consider that peripheral network members, that is, those with weak ties to their networks, are also those who have the least opportunity to develop common ground with other members of the network. As common ground is necessary to support default or “by accident” (rather than calculated, strategic) indirectness, our ability to read through others' lines can become curtailed if our social network ties are weak, as can happen under the impetus of globalization. Although we are becoming members of an increasing number of networks, we remain peripheral members of those networks and thus do not get the chance to develop the common ground with other members of the network that would allow us to engage in default indirectness with them.