TWO KINDS OF TYPIFICATION

Lao pétanque players use each of the two EMTs ‘gambling for beer’ and ‘gambling for money’ for both generic and specific reference.Footnote ⁴ Generics such as ‘police dress in blue’, ‘ibises eat trash’, and ‘sourdough is delicious’ are statements that ‘express general claims about kinds and categories, rather than claims about particular individuals’ (Leslie & Lerner Reference Leslie, Lerner and Zalta2016). In generic uses of ‘gambling for beer’ and ‘gambling for money’, players refer to these types of gambling as types. They describe their general properties, compare, evaluate, and essentialize them, and so forth. For example, in transcript (1) below, a player uses the phrase tii3 ‘hit’ kin3 ‘eat’ ngen2 ‘money’ to describe the properties of gambling for money.

1. (1) Player:

   (lip smack)  suan1 laaj3   khòòj5   kaø-  mak1   tii3   kin3 ngen2

   mostly    1.sg   topic.link   like   hit eat  money

   ‘(lip smack) Mostly, I like gambling for money’

   phòq-   vaa1 nùng1 maa2 lèèw4 tii3 kin3 ngen2 lèèw4 man2 caø- bòò1 maw2

   because comp   one    come already hit eat   money already 3.sg irrealis neg drunk

   ‘because, first of all, in gambling for money [you] don't get drunk.’Footnote ⁵

In the case of specifics, players use the same EMTs to label, characterize, or typify actual or imagined games. Take, for example, how one player lectured an audience member who screamed during his turn (see Zuckerman Reference Zuckerman2016).

1. (2) Player:

   taa3  bòø-  mii2   qaq1   lin5  kin3  ngen2   nii4

   eyes  neg  have  q.polar   play  eat    money    deictic.prox

   naq1  hêt1  hoo  kaø- vaw4  dii3 dii3kaø- vaa1 san4  naq1

   fac  do  hoo  topic.link speak well-well topic.link  comp  like.that fac

   ‘[What], you don't have eyes? [We] are gambling for money here and you yell ‘hoo!’ Is that ‘speaking nicely’, [is that what you] are saying here?’

Here, the player used the form lin5 ‘play’ kin3 ‘eat’ ngen2 ‘money’ to put the particular game he was playing under a description.Footnote ⁶

In these two examples, the players did not produce exactly the same EMTs to refer generically and specifically. Rather, they used two options from a broader ethno-metapragmatic array (see Figure 1). Players use the forms in this array in pragmatically and semantically similar enough ways that I here discuss them under the umbrella of the English ‘gambling for beer’ and ‘gambling for money’.

Figure 1.

How to refer to the two kinds of gambling.Footnote ⁷

The forms in this array are the focal EMTs of this study. They contrast with the innumerable other—and for the purposes of this article, non-focal—ways players describe games of pétanque, as, for instance, ‘pétanque’ (pêêtang3 or pêêtòòng3), ‘sport’ (kaan3 kiilaa2), ‘gambling’ (kaan3 phanan2),Footnote ⁸ ‘wasting time’, ‘losing money’, ‘men working on their tans’, and so on. Any combination of the forms in the array can be used for either generic or specific reference without morphosyntactic modification. This flexibility is partly due to the fact that Lao, similar to Mandarin (Tardif, Gelman, Fu, & Zhu Reference Tardif, Gelman, Fu and Zhu2012:131) and Vietnamese (Sidnell, this issue), does not mark plurality, obligatorily signal aspect, or have nominal determiners. The sentence below, for example, has at least the following interpretations, including the patently generic (3a) and the patently specific (3c).

1. (3) maa3  kin3  kaj1

   dog  eat  chicken

   1. a. ‘Dogs eat chicken’

   2. b. ‘The dog eats chicken’

   3. c. ‘The dogs are eating the chicken’

   4. d. ‘A dog ate the chicken’ (said, for instance, upon discovering the carcass)

That utterances are polysemous like this in artificial isolation calls attention to what disambiguates them in practice, namely the semiotic cotexts and contexts in which they are spoken. In the small but growing work on generics cross-linguistically, scholars have found that languages tend to leave generics unmarked and to mark specifics (see Mannheim, Gelman, Escalante, Huayhua, & Puma Reference Mannheim, Gelman, Escalante, Huayhua and Puma2010; Meyer & Gelman Reference Meyer and Gelman2016; Gelman, this issue; see also Agha Reference Agha2007:43 and Sidnell, this issue, on generics as non-selective deixis). Lao speakers use a range of resources, including non-linguistic ones, to do this latter work of marking lexemes or utterances as specific. Among other things, they point to objects (see Meyer & Gelman Reference Meyer and Gelman2016), delimit with numbers, look at their referents, say statements that resonate with previous comments about other specific referents, and offer utterances pragmatically anomalous enough to seem unlikely generics, such as, ‘The lion is green’ (Gelman Reference Gelman2003:224–26). They also frequently mark their utterances with various deictics. In transcript (2) above, for instance, the man uses a proximal deictic nii4. Footnote ⁹ The variety and heterogeneity of these resources makes clear that construals of reference as specific or generic are inseparable from issues some might call ‘pragmatic’ or ‘contextual’ (Declerck Reference Declerck1991; Ni Reference Ni1996).

Although different languages, and semiotic repertoires more broadly, have different resources for ensuring EMTs are construed generically or specifically, all languages and their users have the capacity to use EMTs in both generic and specific ways. These two ways of referring and modes of typification, in turn, afford distinct kinds of social action. It is these affordances to which I now turn.

GENERICS ABOUT GAMBLING

Pétanque was once banned in Laos because the socialist state deemed it a ‘problem’ that stimulated gambling. In contemporary late-socialist Laos, however, it has become something of a national sport (Zuckerman Reference Zuckerman2018). While most men and women in the city play or have played the game, it is most associated with (and commonly played by) men in their late-twenties to mid-fifties. These men, who tend to play in teams of two or three, are proud of their skill in these games, which often draw audiences. But they are also mindful of lingering criticisms of pétanque as unethical and unproductive. To discuss and diffuse these criticisms, they often make use of the distinction between ‘gambling for beer’ and ‘gambling for money’. This distinction—alongside the moral properties players tend to attribute to it—echoes a more widely circulating distinction central to Laos's 1975 socialist revolution: between ‘mutual aid’ and ‘solidarity’, on the one hand, and ‘greed’ and ‘noise’ on the other (Evans Reference Evans1990; High Reference High2014). Whereas gambling for beer is frequently treated as if it produced the kind of camaraderie the socialist government has yearned for since its inception, gambling for money is often derided as divisive.

When speaking generically, players talk about these two EMTs easily and consistently (see Agha Reference Agha2007:119–24; Hampton Reference Hampton and Pelletier2009:81). This ease and consistency evidences their social-semiotic status as distinct kinds (Krifka, Pelletier, Carlson, ter Meulen, Chierchia, & Link Reference Krifka, Pelletier, Carlson, Meulen, Chierchia, Link, Carlson and Pelletier1995:11). While, with enough intervention from a researcher, people can always be led to compare and contrast different categories (Pelletier Reference Pelletier and Pelletier2009:5–6), pétanque players need no nudging to talk about gambling for beer and gambling for money. During interviews, nearby people frequently even interrupt to collaborate on these generics, as if they were joining in to help sing the chorus of a popular song that everyone knew. To one man's comment that, “When you lose you drink together, and when you win you drink together”, another added, “[It's like] constructing [solidarity] together”.

These generics are never exclusively about the literal stakes of the two types of gambling, beer and money. Rather, players use these stakes and the EMTs that refer to them as shorthands for distinct socialities and ethical ends. They genericize about ‘gambling for beer’, in which losers buy beer that is then drunk together, as simply for ‘solidarity’, ‘love’, and good sociality. They paint ‘gambling for money’ as an activity for money won at the expense of relationships. In responses to a written survey I passed out, for example, most respondents stressed the types’ distinct purposes (see Zuckerman Reference Zuckerman2018). Some used the language of ‘goals’ and ‘targets’ (paw4 maaj3). Others sketched the difference in ends using the form phùa1, meaning ‘for’: “Gambling for beer”, one man wrote, “means to play for (phùa1) friendship, for (phùa1) drinking beer, it does not mean you are playing to lose or win. Gambling for money [in contrast] means to play for (phùa1) losing or winning, because if you lose, you lose money and if you win, you win money”. These answers focus on the ethical question of ‘What is a game of pétanque for?’. ‘What is its rationality?’ in Weber's language. ‘What are its ends?’ in the register of moral philosophy. They frame the choice between the two gambling types as something like the choice between gift and commodity exchange (Carrier Reference Carrier1995): as a choice between forming social relations with beer or destroying them for money. And they tie this choice to different actual courts in the city (where people tend to gamble for one stake or another), and to distinct social and gendered figures: prototypical beer gamblers are morally upright co-ed civil servants, tied to the state; while the paragon of a money gambler is a rougher, masculine tuk-tuk or mini-van driver in the tourism industry.

Generics as social action

Because generic utterances are abstract in reference and often consistent across speakers, it is tempting to overlook the moments in which people utter them. And yet, I found that as people in Luang Prabang refer to the timeless qualities of gambling, they also orient to potential judgment or praise in the enchronic (Enfield Reference Enfield2013) flow of interaction. When men filled out my written survey, for instance, they wrote generically about ‘gambling’, but they also worried and joked aloud about how their answers would be received, and what those answers would say about them. This was a pattern. When I interviewed a civil servant named Buun at a money gambling pétanque court, for example, he told me that, “Gambling for beer means whichever person wins… whoever wins or loses, you drink together. But when you gamble for money, the winner puts [the money] in his pocket and gets out of there”. As I followed up, “So you think that playing for money is…” he hastily interjected that it was “not good”.

Talking about these kinds is ‘sticky’ with ethics (Mathias Reference Mathias2019), and their mention often leads to ‘moralization’ (Lempert Reference Lempert2013) and ‘defensive detailing’ (Drew Reference Drew1998). This is especially evident in interviews with civil servants where I was recording or filming, but it is not limited to these more formal or on-the-record interactions with me. In conversations over coffee or late-night drinking sessions with one another, players also talk about gambling with a marked attention to how what they say implicates their desires for money or friendship, and their opinions of their partners and opponents in games. Unlike more ethically neutral subjects, about, say, where someone likes to fish or a new shoe shop in the market, the topic of gambling almost forces one to take a stance; saying nothing is liable to be construed as a tacit sanction of money gambling and thus a hint as to what sort of person the speaker is.

Such construals are possible only because generic utterances are never as timeless as their content. As Goffman (Reference Goffman1974:247) put it, ‘Fanciful words can speak about make-believe places, but these words can only be spoken in the real world’ (compare Gal & Irvine Reference Gal and Irvine2019:181). In that real world, generics are not floating amorphously, but rather indexically anchored to the people who produce or are held responsible for them.Footnote ¹⁰ At times, people explicitly tie their generic references to themselves or to others, for example, with first person pronouns and stance-taking verbs, in phrases like ‘I don't like money gambling’, or by embedding similar utterances in reported speech. At other times the mere fact of producing a generic utterance invites listeners to guess at what it conveys about the speaker or her motivations for saying it, or to read the alignment or non-alignment of that generic with the generics others have spoken as a diagram of her social relations with those people (Du Bois Reference Du Bois and Englebretson2007). This point is often forgotten in research that studies the semantics of generics, but starkly obvious in experience: when a politician gives an impassioned denunciation of ‘corruption’ in the middle of an investigation into his finances, we know he is not merely articulating a general opinion. Rather, he is refuting the allegations against him.

And yet, while those who speak generics are liable to be held responsible for them through the indexicality of speakership, the connection between generics about kinds and the real events and people that might fit those kinds is fundamentally less direct.Footnote ¹¹ Instead, generics share a pragmatic affinity with other varieties of deictically non-selective forms of ‘indirect addressivity’ (Lempert Reference Lempert, Paulston, Kiesling and Rangel2012a), such as indefinites (see Brenneis Reference Brenneis1984; Whitehead & Lerner Reference Whitehead and Lerner2020).

For players in Luang Prabang, this makes generics good objects for stance-taking (see Pino Reference Pino2020). Their unanchored reference allows players to evaluate the types of gambling more strongly than they might otherwise. This is what generics afford. With them, a person can condemn and commend without specifying his target's connection to spatiotemporally individuated events and people; he can sing the praises of beer gambling while still claiming that he does not go to the courts much; he can distance himself from money gambling as a type of activity without directly censuring others or holding them accountable.

This is not to say, of course, that generics lack the capacity to be read as specific indictments of discrete people or events. They too can make addressees or overhearers ‘wonder, Is he talking about me?’ (Lempert Reference Lempert2012b:114). Such construals might follow a logic akin to the inferred syllogism in Table 1.

Table 1.

An inferred syllogism.

At times at the court, in fact, people grew uncomfortable making strongly worded generic evaluations about money gambling, presumably because of processes like these. Sometimes in interviews it felt as if the relevance of the abstractions was too clear, as if I were asking them to loudly discuss why healthy eating was important as we sat in a booth at McDonald's.

But, although possible, syllogistic movements from major premise to minor premise to conclusion are not always so natural, and generics often leave space for their utterers to plausibly deny that their generics implicate specific individuals or events. This is clear, for instance, when one looks closely at the way players describe the ethical limits of how they gamble for money. Almost all players distinguish some category of person with whom they would not wager money because doing so would be ‘bad’. Some men, like Buun, told me matter-of-factly that gambling for money was always bad, no matter against whom one was gambling. Others generically restricted their money gambling to strangers or acquaintances and refused to play with ‘friends’ (muu1 or siaw1).Footnote ¹² These two positions differ starkly in entailment: one entails no gambling for money at all, the other allows it in some circumstances with some classes of people. But this is where practical syllogism meets discursive practice: actors can, and frequently do, deny that actions or people fit the types referenced in their generic utterances. Later in my interview with Buun, for instance, he bet money on the game in front of us. When he lost, he reframed the bet he had made: if he had won, he said, he would have bought us some beers. And so, a bet for money became something like a bet for beer.

Social scientists often analyze circumstances like these as contradictions between talk and action (e.g. Evans Reference Evans1990:143–44), but both generics describing events and people and specifics which ascribe types and statuses to events and people are linguistic; they can both use the same EMTs. The disjuncture is thus not between talk and action but between two kinds of talk. It is an effect of how people make use of the differential affordances of generics and specifics. Simply put, generics make for good stance objects because they allow people to paint the world with sharp lines, and thus communicate something about the world and themselves, without so explicitly entering the fray and declaring what counts as what.

SPECIFICS

In contrast, sorting out what's what, using EMTs to label tokens as tokens of a type (see Rumsey Reference Rumsey, Enfield, Kockelman and Sidnell2014:410; Gal & Irvine Reference Gal and Irvine2019:95), is exactly what specifics do best: put swaths of experience under description. Players use ‘gambling for beer’ and ‘gambling for money’ in this way more often than one might expect. These utterances tend to be clearly purposeful. Instead of neurotically organizing their games into money and beer gambling categories, players use these categories to get things done, to evaluate how others are acting, and to convince them to act differently (Enfield & Sidnell Reference Enfield and Sidnell2017; on ‘labeling’ and its effects see Becker Reference Becker2008; Thompson Reference Thompson2014).Footnote ¹³ Specifics afford such social force because they have the capacity to directly implicate people in the world.Footnote ¹⁴

Sometimes players refer specifically with ‘gambling for beer’ and ‘gambling for money’ for relatively straightforward purposes, for example, to satisfy some late arriving audience member's curiosity as to what kind of game is happening. Although hopeful gamblers do not need to characterize the type of gambling that is going to happen to make a bet—some bets are sealed with sequences as simple as, ‘50,000?!’… ‘[I'll] take [that]!’—gamblers do at times refer to their games as money or beer games in the process of negotiating stakes, balancing teams, splitting-off outside bets for audience members, discussing whether a neutral party should hold the wagered money, and more generally preparing for gambling. One man, for instance, enlisted people milling about to start a new game by shouting, “Yeah, OK, [let's] gamble for money here!”. In putting the games he wanted to play under a description, he moved the crowd.

In more of my examples from games of pétanque, however, players use these EMTs with evaluative effect. In one game for beer, for example, a player was lifting his left leg up slightly each time he threw the ball, and someone complained that it violated the rules (see Figure 2). In response, another man defended the player who had raised his leg by pointing out that they were gambling for beer.

1. (4) Defender of the player who raised his leg

   qoo3   ñok1  khaa3  qiø-   pên3   ñang3   nòq1    tii3   kin3      bia3   nòq1

   oh   raise  leg   irr   cop   what   q.polar   hit  eat   beer     q.polar

   ‘Ohh, what's the problem with raising [your] leg? [We are] gambling for beer, right?’

The player himself later repeated the same idea, given in (5).

1. (5) Player who had raised his leg

   bòò1  ñòk2  kaø-   bòø-  pên3  ñang3  tii3  muan1  sùùn1

   neg  raise  topic.linker   neg  cop  what  hit  enjoy  happy

   ‘No, lifting [your leg] is not a problem, [we] are playing for fun.’

   tii3  kin3   law5  kin3   bia3

   hit  eat   alcohol  eat   beer

   ‘[We] are gambling for alcohol and beer.’

Here, the description of this game as ‘gambling for beer’ is used, in conjunction with other descriptions and claims, to hold the leg-lifter's critic accountable, to remind him not to take the game too seriously.

Figure 2.

Players talking about leg-lifting.

Players frequently use ‘gambling for beer’ like this: to mollify arguments or disagreements on the court or to correct others’ claims about the goals of the game, and thus, the goals of the interaction. For instance, when people sometimes voiced frustration at their own missteps or poor play, beer gambling opponents and partners often reminded them that we were playing for beer. Such descriptions assuaged guilt. In pointing out that we were gambling for drinks to be shared, they highlighted the social ends of the game rather than the material consequences of winning or losing.

Uses of ‘gambling for money’, in contrast, offer players a tool for emphasizing these consequences, for stressing what may be won or lost. With the EMT, players can underline the seriousness of games, enforce decorum, and get in the heads and ears of their opponents by forcing them to think about the money they are risking.Footnote ¹⁵ Take the example in transcript (2) above, for instance, wherein my partner yelled at the man who distracted him: “[What], you don't have eyes? [We] are gambling for money here…!” (see Figure 3).

Figure 3.

My partner chastises his distractor (image has been modified to preserve anonymity).

Notice how the force of this utterance centers around the type-identity of the game qua token: this is a money gambling game; such hijinks should not be allowed. Compare this with a scene from The Big Lebowski in which a bowler's foot goes over the line and his opponent, Walter, draws a gun in outrage. In response, the lead character, ‘the Dude’ tells Walter to calm down––it's “just a game, man” ––and Walter justifies his righteous officiousness with his own specific typification: “This is a league game”. These descriptions license action and hold others accountable (Enfield & Sidnell Reference Enfield and Sidnell2017).

Across specific uses, ‘gambling for beer’ and ‘gambling for money’ become a part of social action in slightly different ways. Take the below three snippets from two different games (Figure 4), all of which are for money and between four players. In (i), a player in the game challenges his opponent who is about to take his turn, asking if, given that real money is on the line, he is brave enough to ‘shoot’ his pétanque ball, a riskier shot where one tries to knock the opponent's ball out of the way rather than merely approach the smaller target ball. Here the definition of the situation as ‘gambling for money’ is used to unsteady the opponent and bring his fortitude into question. In (ii), a player yells at his opponent for taking a practice shot in between turns, threatening to count the shot and reminding the man that they are gambling for 100,000 kip each ($12.50 USD per person at the time). The player's point is that, given the stakes, his opponent is being inappropriately casual. In (iii), finally, an audience member who is betting on the game takes issue with one of the players in the game's incessant heckling of his opponent, who has been mostly quiet. When the player retorts that his other opponent (the quiet player's partner) has been heckling too, the audience member refutes this by describing the stakes. Because they are gambling for money, the audience member's logic goes, the player really should be focused only on the game.

(i) kaa4 tii3 bòò1 liin5 haa3 ngen2 nii4 qaw3  ‘Are [you] brave enough to shoot it? [We are] gambling for money here, OK!’

(ii) qaw3 leej2 daj4 laj1 qaw3 leej2 daj4 jaa1 tii3 phuu5 laø- sèèn3 lêêw4 vaal  ‘[We'll] count that, [we'll] count that, don't [take practice shots]!... [We're] gambling for 100,000 [LAK] per person!’

(iii) khaa3 khuu1 mèèn1 ñang3 kin3 ngen2 kan3 juul phii4 maa2 vaw4 khaa3 khuu1 nèè2  ‘So what about [your opponent's] partner? [We] are gambling for money here and [you're] talking about [his] partner?’

Figure 4.

An array of specifics (images have been modified to preserve anonymity).

Of course, my brief explanations of the effects of these specifics are, in contrast to the specifics themselves, clumsy and stilted. In fact, specifics often work through concision. With the sharp knife of description, they cut efficiently, and explaining how they work can resemble the joyless task of explaining a joke.

The following slightly expanded example (see Transcript 6 below), taken from a game where Bot was gambling for money with his siaw1 or ‘close friend’, provides more detail as to what such specifics often feel like on the ground (on siaw1, see Zuckerman Reference Zuckerman, Djenar and Sidnell2022). When Bot and his friend began to bicker, an audience member laughed and called out the mismatch: the siaw1 are “gambling for money (kin3 ngen2 kan3)”. Later in the same game, the men began to bicker again. This time a different man in the audience, Phuumii Hii, described their gambling, and added that the pair was violating a pledge they had made weeks before, sealed with a handshake, that they would no longer ‘gamble for money’ with each other.

1. (6) Phuumii Hii: qaaw4 kuu3     hên3 suu3   cap2 mùù2  vaa1  suu3

   hey    1.sg.bare  see  2.pl.bare grab  hands   comp    2.pl.bare

   bòø-  kin3  ngen2 kan3

   neg    eat   money reciprocal

   san4 suu3   kaø-  ñang2 kin3  ngen2 kan3

   thus 2.pl.bare topic.linker  still   eat money  reciprocal

   ‘Hey, I saw you two shake hands and say that you wouldn't gamble for money anymore, but you're still gambling for money.’ pointing at Bot and another man.

   Bot: bak2  qan3   nan4  kin3  ngen2  mùng2

   bare.prefix  thing  deictic.distal  eat    money  2.sg.bare

   san4       naq1

   thus    factive.particle

   ‘That guy's gambling against you!’ pointing at Phuumii Hii's siaw1       whom he is gambling against.

   Phuumii Hii:    man2 siø-  cap2 mùù2  kuu3

   3.sg.bare irrealis grab  hand  1.sg.bare

   ‘Did he shake my hand?’

   (several people laugh)

   caw4    tòòng4  cap2  mùù2  vaa1  bòø- qaw3  ngen2      kan3

   2.sg.polite  need     grab    hand  comp   neg  take   money  reciprocal

   thèè4   thèè4  qan3   nii4

   true   true    thing   deictic.proximal

   ‘You need to shake hands if you really aren't going to gamble for money.’

This passage shows the stakes of these types for these players. ‘Gambling for money’ is an important enough kind of activity that players take the time to pledge not to do it. They remember their pledges, and as they organize games of teams, they often reject otherwise balanced and appealing matchups because those matchups pit players against one another who should not be gambling for money: “we don't eat each other's money”, people say. The example also evinces the capacity of specifics to highlight and thus circulate the association of types with certain qualities and entailments. Phuumii Hii, like the audience member before him, labels Bot's ongoing bet as ‘gambling for money’ in response to, and as a sort of explanation of, Bot's bickering with his siaw1. The utterance is thus liable to be construed as underlining and concretizing this capacity of gambling for money, so often mentioned in generics about it: money bets make people, even friends, argue.

THE EXPLICIT AND THE TACIT

Recognizing that generics and specifics offer people distinct ways of using the same EMTs reveals why EMTs can seem so messy: the ‘explicit discourse’ that uses them is not only socially embedded, but it is differently embedded in ways that shape that discourse's form. This fact, in turn, brings into relief the difference between generic and specific modes of explicit typification and other, less direct, modes of typification.

In the last few decades, it is these latter modes that have received the most attention in studies of language in society. This is especially true in linguistic anthropology, my own field, where at least since Michael Silverstein's (Reference Silverstein, Basso and Selby1976) groundbreaking article on shifters, and in part as a corrective to the emphasis of non-specialists on ‘words’ (Ochs Reference Ochs2012:147–48), scholars have spent most of their energy excavating an array of more indirect, implicit, tacit, or underhanded typifications (see Silverstein Reference Silverstein, Enfield, Kockelman and Sidnell2014). These typifications function metapragmatically but are not necessarily identifiable as metapragmatic discourse (Silverstein Reference Silverstein and Lucy1993): think of the pitch contour of a ‘voice’ (Hill Reference Hill, Tedlock and Mannheim1995; Keane Reference Keane2011), the manner of a burial (Parmentier Reference Parmentier, Mertz and Parmentier1985:140), or the metrical structure of a stretch of discourse (Silverstein Reference Silverstein2005 on ‘type-sourced interdiscursivity’). While ‘indirect’ or ‘tacit’ typifications are a dizzyingly heterogenous bunch, two properties span many definitions of them: their lack of focal EMTs (Silverstein Reference Silverstein2010:345), on the one hand, and, their recognized capacity to, specifically and felicitously, be construed, glossed, or substituted with these same focal EMTs, on the other.Footnote ¹⁶ ‘Indirect typifications’ are thus ‘indirect’ insofar as they are ‘metapragmatically opaque’ (Silverstein Reference Silverstein1981:14), that is, insofar as they are formally distinct from the EMTs that are thought to gloss them most accurately. In this view, ‘indirectness’ is a situational, interdiscursive quality, identifiable only relative to a construal with a given focal EMT. The use of the word ‘illegal alien’ (Dick Reference Dick2011:E41), for example, may be construed as indirect in respect to ‘race’, but direct in respect to ‘immigration status’.Footnote ¹⁷

This understanding of indirectness means that even when focal EMTs, in all their messiness, are not the subject of study, they continue to be definitional of much of what counts as indirect for the analyst. Those who shift their focus ‘beyond labeling’ (Gelman et al. Reference Gelman, Coley, Rosengren;, Hartman, Pappas and Keil1998), but nevertheless want to investigate a subject that is locally salient, important, and named by an EMT, thus often continue to orient research questions and methods to that EMT. The EMT becomes a shorthand for a type—of genre, characterological figure, speech event, object, action, or material quality (see Wortham & Reyes Reference Wortham and Reyes2015).Footnote ¹⁸ In this role, the EMT is not the subject of analysis, but rather, a rough tool for delimiting the scope of research.

EMTs are good tools for this delimiting work—obviously enough—because they can be used in explicit discourse. In this role, they can offer unity and focus to the notion of types and connect tokens of them. They can serve as stand-ins for types, lexical links between the two modes of typification I have described here, through which actors (i) reflect on the properties of ‘ideal’ categories, generically, and (ii) label specific experience. And they can bridge and unite these scenes of typification, and consequently, anchor the idea that social ‘types’ exist outside of interactional time. When a person reflects on the essence of ‘lying’ over breakfast and, later that afternoon, calls a particular stretch of discourse ‘a lie’, there is a sense that the two utterances are connected by an underlying conceptual type, which exists by way of the shared lexeme ‘lie’. This lexeme becomes a semiotic hinge for deduction and induction (Agha Reference Agha2007:74–75; Gal & Irvine Reference Gal and Irvine2019:276). Its lexical identity makes the swing from generic discussions of ‘a lie's’ qualities to individual, identifiable ‘lies’ appear effortless.Footnote ¹⁹

But examining uses of EMTs in specifics and generics makes patently obvious that, in the flow of interaction, this sense of interdiscursive effortlessness is a mirage. An EMT may be a useful stand-in for a type, but one use of it is not inherently linked to another. Such links must always be construed. The role of construal is especially clear when speakers produce an EMT only specifically or only generically.Footnote ²⁰ These utterances are syllogism-fractions. If a person produces only a generic (e.g. “gambling for money is bad”), with ‘no directions as such for extension’ (Silverstein, this issue), then any minor premise concerning what that generic applies to is left to inference, as in Table 1 above. If a person produces only a specific reference, then one is left to infer what that choice of EMT means. In Table 2, I have represented these possible inferences as if they were co-occurring generics, but that such generics are never uttered, but only (possibly) inferred, is the point: EMTs do not enter into syllogistic logic automatically.Footnote ²¹

Table 2.

An inferred syllogism, with a spoken specific.

All of this means that even EMTs used in ‘explicit discourse’ can be plagued by dimensions of tacitness.Footnote ²² But the inferential slope that ‘indirect’ typifications must climb to enter into such syllogistic logic is far steeper. If we should be cautious, then, about claiming that one use of an EMT relates to another, we should be even more circumspect about making claims about how EMTs relate to ‘indirect typifications’, on the whole. As many have pointed out (e.g. Dick & Wirtz Reference Dick and Wirtz2011:E2; Agha Reference Agha2015:315–16), the metapragmatic opacity of ‘indirect typifications’ is part of what makes them attractive to speakers, as it allows them to be less ‘on record’ about what it is they have done (see Hill Reference Hill2008). But this quality also raises fundamental methodological problems for analysts. When an analyst treats a tacit typification (see Schegloff Reference Schegloff2009 on ‘possible actions’) and its object—that is, a swath of experience unlabeled in situ—as a token of a type, and then glosses it with an EMT, she is affecting the scene she is studying. In one reading, she is translating the indirect into the direct; in another reading, she is conjuring the indirect, pointing it out where it may have never been. Either way, she is ‘appropriat[ing] the metapragmatic function’ to herself (Philips Reference Philips, Schieffelin, Woolard and Kroskrity1998:223), lending a hand to the typifying work that is underway (see Stokoe Reference Stokoe2006:471).Footnote ²³

Focusing on specific and generic uses of EMTs by no means removes the analyst's interpretive role from the equation, but it does help us move beyond the coarse binaries of the implicit and the explicit, and the direct and the indirect (Gal & Irvine Reference Gal and Irvine2019:182). Taking the distinction between generics and specifics seriously makes clear that, at base, to understand how ‘types’ are brought to bear in discursive practice by EMTs requires us to explicate these different modes of typification and to begin by treating them as if they were non-substitutable. This involves, minimally, holding in suspension whether any ‘indirect’ or ‘tacit typification’ relates to some EMT.Footnote ²⁴ Doing this has an obvious drawback. Being circumspect about what is or is not a ‘tacit typification’ lessens the immediate social power of the analyst (Philips Reference Philips, Schieffelin, Woolard and Kroskrity1998:223), who often gains ethical and political efficacy insofar as she can deploy the same EMTs that circulate in public discourse, and thereby make the tacit explicit. With analytic confidence as to what is what, we offer more convincing answers as to what is good and what is wrong. But while such work is often illuminating of—and frankly essential for meaningfully affecting—problems in the contemporary world, recognizing that EMTs offer this political efficacy only further evinces the depth of the conceptual problem: focal EMTs are useful as tools for analysts for the same kinds of moral, political, and interactional reasons they are useful for pétanque players. They are tools not just for tagging or reflecting upon experience, but for also, among other things, holding specific people accountable and taking stances on the world at a distance. This insight invites a more fine-grained understanding of the varieties of typification and their relations with politically and morally charged ‘social types’, so that we can identify power without eclipsing or flattening its mechanisms. The distinction between generics and specifics is one more step toward such a fine-grained understanding.