OVERVIEW
Contribution to the field
This article develops the field of interdisciplinary research into the linguistic construction of identity by highlighting specific links between pragmatic particles (Holmes 1997) and social constructionist theories. I explore the ways in which such particles serve particular functions for the groups and individuals who use them, and how these in turn can be related to the group dynamics outlined by various social theorists. The essay highlights how the young people in the two London peer groups I have studied – one mostly working class (hereafter WCG) and the other mostly middle class (hereafter MCG) – reproduce, reinforce and occasionally challenge dominant discourse. It explores the link between Foucault's (1977, 1980) notion of power/knowledge and the verb “to know” in spoken interaction. Although the link can be traced through both sets of data, the ways it is manifested in the two groups show both similarities and differences. I link the similarities to the importance of the peer group in adolescence (Seltzer 1989) and to the potential effects of its characteristics on processes of identity construction. The differences I link to the young people exploring their differing positions within the wider society.
Why you know and I know?
Patterns
The findings reported in this article form part of a larger project that set out to look for themes and patterns in the conversational interaction of two groups of young people and to analyze what these could tell us about how they were constructing their identities. What first alerted me to the possibility that these particular pragmatic particles might serve particular functions for the respective groups was the preponderance of the verb “to know” in both sets of data. I therefore decided to consider the possibility that the similarity between the two pragmatic particles might be as important as their differences – that although the collocation of “know” with “you” in WCG and with “I” in MCG suggests differing speaker/hearer dynamics, the two might have similar functions. I observed that both pragmatic particles were functioning as norm setters or norm reinforcers. This suggested a strong link between norms and knowledge, and I have therefore used Foucault's (1977, 1980) concept of power/knowledge to frame my analysis theoretically. I discuss this below and thereafter discuss my own theoretical position in the light of other work on language and identity construction generally, and you know and I know specifically.
Explicit marking of “knowledge.”
By using the verb “to know,” the young people label certain information explicitly as “knowledge.” It could be argued that one of the functions of utterances in general is to relay speaker knowledge (albeit that this is often not the primary function). We therefore need to ask why the young people state this explicitly (or “perform” it) in some utterances and not in others. It seems likely that those utterances whose “knowledge” status is explicitly claimed are taken out of “neutral” territory and made more “visible” for some reason. It seems therefore to be important for the young people explicitly to state their knowledge about various things. It is notable that the possibility that they are stating precise knowledge of the various propositions can be refuted in the vast majority of utterances containing either you know or I know. It is impossible to claim someone else's knowledge of something (as in you know) because knowledge is personal. In the case of I know, it is often the case that the speaker could not in fact “know” what the previous speaker has said, for example because the previous utterance presented the previous speaker's emotions, which are also personal. Thus, in a discussion about the relative merits of giving up chocolate for Lent, Anna says I wish I could do that, to which Libby responds I know I would like …; later in the same conversation, Anna says yeah I know to Cassie's do not eat chocolate near me cause it's like I really want to eat it. We must therefore consider the possibility that the knowledge the young people are stating is less precise, that it is some kind of general social knowledge. It seemed pertinent to consider whether Foucault's theory of power/knowledge could help to illuminate the functions of these utterances (or utterance parts) for their speakers. Having analyzed the data in light of this theory, I claim that the use of you know and I know is one of the procedures by which these speakers come to accept utterances as truth or dominant discourse (Foucault 1977).
THEORETICAL POSITION
Foucault's notion of power/knowledge
Definition of power/knowledge
Foucault claims that there is a very close relationship between power and knowledge, so close that central to his analysis of society is the single concept “power/knowledge”: “There is no power relation without the correlative constitution of a field of knowledge, nor any knowledge that does not presuppose and constitute at the same time power relations” (1977:27). Power/knowledge is not an essentialist concept. It is not a case of knowing how the world works and thus making optimal choices that will give one power over one's existence. Knowledge is not absolute truth, because any material fact that is interpreted linguistically will necessarily be altered in some way in the process. However, Foucault claims that power/knowledge comes to be seen as truth because language use reproduces and reinforces concepts. These concepts come to be accepted as knowledge, and those who voice them reproduce and reinforce a worldview that is interpreted as “reality.” He is therefore interested in documenting not what truth is but rather the procedures by which something comes to be considered knowledge and, by extension, “truth.” Foucault calls this language use “discourse.” Thus, to have power/knowledge is to have access to the discourses that are dominant in a society. From this position individuals can reproduce and thus reinforce those discourses.
It is clear, then, that “knowledge” is reproduced relationally rather than essentially. Something can be evaluated positively only if something else is evaluated negatively. The reproduction of dominant discourses necessitates identifying as deviant other ways of viewing the world. Therefore, speakers who do not have access to dominant discourses or who actively choose instead to voice nondominant discourses will be marginalized. They will be considered not to have knowledge and therefore will be powerless. Where their behavior is seen as a threat to those in power, they may even be physically marginalized, for example in a prison (Foucault 1977). They then come under constant surveillance (Foucault studies the function of the panopticon) and can be “normalized” into behaving in ways desired by those in power. Because of this and other material manifestations of surveillance (a recent example would be closed-circuit television), the notion of being watched and evaluated enters the psyche of individuals in a society to the extent that the material forms are no longer required: Individuals begin to self-monitor from the point of view of the powerful. Furthermore, meaning within dominant discourses is (re)constructed by marking out deviant categories against which other categories can be viewed as the norm and, by extension, dominant. To claim dominant knowledge, then, is also to claim, implicitly or explicitly, a position of power for the speaker. These, then, are the two things that will be examined in this article: the (re)construction of the dominant knowledge accepted by the peer group, and the self-positioning of the speaker in a position of power.
Local reproduction of power/knowledge
Although Foucault illustrates the concepts of dominance, deviance, and power/knowledge with reference to large institutions, as I have mentioned, he constantly reiterates the idea that the power relations in any society are dynamic; they are not initiated “at the top.” Relations of power between individuals reproduce the norms that exist in society – indeed, it is due to these individual interactions that the dominant discourses and norms come to exist in the first place: “These (micro-level) mechanisms of power have been ‘invested, colonised, utilised, involuted, transformed, displaced, extended’ by more general forms, and these are what we tend to see as forms of social power” (1980:99). The fact that the norms which exist in different societies or in the numerous subsections of any one society are not identical is evidence for Foucault's theory of the dynamic nature of relations of power. It follows from this theory that, with each encounter, the individuals involved have the capacity not just to reinforce the existing discourse(s) but also to challenge them and perhaps even to develop or change them. This article, then, aims to show the procedures by which norms are reproduced by young people in their peer groups, to what extent they reinforce dominant discourses, and to what extent, if at all, they challenge them.
Why adolescent peer groups explicitly reproduce power/knowledge
Peership versus friendship
Having observed that members of both the groups I have studied explicitly voice their knowledge of dominant discourses, we could tentatively claim that adolescent peer groups in general do this. The question then arises as to why this should be the case. With reference to Seltzer 1989, I claim that this happens because the size of the peer group and the relations of difference therein necessitate more explicit norm setting. There has been much illuminating work on the construction of identity in friendship groups as opposed to peer groups (e.g., Coates 1999). Such groups offer much to such studies – for example, a potentially higher level of self-disclosure. However, I surmised that the looser relationships within peer groups would necessitate more explicit performance of identity and so would be an important context in which to view identity construction.
It has been shown that even the youngest children reproduce society in their play (Cook-Gumperz et al. 1986). However, Seltzer claims that whereas in earlier childhood it tends to be friendship groups that are the main site for identity construction, in adolescence the focus changes to the peer group (Seltzer 1989:21). Peer groups are less likely than friendship groups to be gender- and/or class-specific. Even if they are, however, they are still bigger than friendship groups and so make difference a more salient issue than it would be in friendship groups. This is not to deny that young people in friendship groups talk about people who are different from them. But young people in peer groups are much more likely to have to interact with people who are significantly different from them. Increased size and difference mean that it will take relatively more work to negotiate power relations and set norms. We could therefore expect to see this being done more explicitly.
The peer group as a nascent society in microcosm
Young people are of course influenced by the norms of, for example, the media, class relations, and gender relations that are reproduced by the society in which they live. However, in their peer groups they reproduce group-specific versions of these. For example, Eckert 1995 illustrates young people's using certain phonetic variables to construct an identity that orients either toward or away from the urban center. Membership in the “burnout” and “jock” groups bears some relation to the social class structure of the community in general, working-class and middle-class respectively. However, an individual young person's membership in one or the other group cannot be seen to be determined by his or her family's social class identity. So, the norms reproduced by the peer group can vary considerably from those that are reproduced by the wider society. Because of this, we must, in my opinion, view the peer group not simply as a subsection of society but as a society in microcosm. Moreover, the setting of norms and relations of power will be particularly salient to such groups because they are relatively “new”: They could be described as societies in their nascent stages. The onset of adolescence tends to coincide with a change of school and increased independence from the family. It is therefore likely that young people will form new relationships and spend the majority of their time with their peers. It is true that the majority of children experience different forms of peer group interaction from a young age. However, in many cases this interaction consists of structured activities overseen by adults, whether in school or in various clubs or organizations. Even during times of play, there is sometimes an adult within earshot who is in a position to monitor “problematic” activity. Thus, there is a continuum of observed/unobserved peer-group activity along which children move. It could be claimed that the period of adolescence is the first time that the balance is weighted more toward unobserved peer group activity for a majority. It is the first time that most young people spend most of their time with other people of relatively equal status to themselves. Relations of power are not seen to be predestined, as they are in the school or in the family. Rather, they are in a constant state of flux and must continually be negotiated. It is therefore not surprising that we see the young people in my data explicitly reproducing dominant discourses and relations of power.
METHODOLOGY
In an attempt to highlight links between young people's conversational interaction and their construction of identity, I decided to focus on spontaneous peer group interaction in a non-school setting. By contacting the group leaders of various drama groups, I gained access to two groups. I chose drama groups for the purely practical reason of voice recognition. Voice recognition was a potential problem for two related reasons, numbers of young people and lack of contact. I was looking not for friendship group interaction but for overall peer group interaction, and this meant that I could have many (if not all) the young people from any one group on tape. I would therefore need some kind of prolonged contact with the young people in order to get to know their voices. However, other considerations precluded this. I wanted my data to be as spontaneous as possible and therefore wanted my presence to alter the group dynamic as little as possible (cf. Labov's [1972] notion of the “observer's paradox”). I therefore decided to approach drama groups. In this context it would be possible for me both to record the young people interacting casually prior to the session and in break times, and to observe them interacting verbally in structured sessions. The former would provide me with relatively more spontaneous data, while the latter would enable me to familiarize myself with individual voices, something that would be essential when I came to transcribe.
One group (WCG) was based in a working-class area of North London, and the other (MCG) in a middle-class area of South West London. Group membership was not completely made up of individuals of one or other class grouping, but it was predominantly so. Initially I met with the young people, explained that I was interested in seeing how they interacted with one another, and discussed with them the possibility of taping some of their conversations. Once I had answered various questions about what my work would be used for and who would see it, the young people all agreed to be taped. There were on average twelve participants in each group each time I recorded. I attended each group six times over a three-month period and on average one hour of data was recorded during each visit, so the total dataset for each group was approximately six hours.
I explained to the young people that they could rerecord or destroy the tapes at any time if they were not happy with what was on them. The actual taping was done in unstructured pre-session or break times, and on each occasion I gave a Dictaphone to one young person. I asked her or him simply to press the record button and carry the Dictaphone around discreetly while interacting as normal. This meant that, while all the young people were aware that they would be recorded at some point, they were not aware of specific recordings until afterward. This, I believe, created the optimal balance between meeting ethical guidelines and gathering spontaneous data. In order to maintain the participants' anonymity, all names were changed during the transcription process.
Seltzer observes that adolescents' selves are not distinct, that boundary lines are blurred (1989:21). Considering this possibility, I decided that a “one at a time” representation of adolescent talk was likely to be unsuitable. For this reason, I have used the stave format of transcription suggested by Edelsky 1981 and developed by Coates 1986, 1988, and 1996. This aims to show conversation as a collaborative rather than a “one at a time” venture.1
The transcription conventions I have used are as follows:
RELATED WORK
Identity construction
Identity as co-constructed
Recent work in various fields such as sociology, psychology, and cultural studies has begun to see identity not as a fixed state but as something that is co-constructed by speakers as they interact with one another. Recent linguistic work in this area has become increasingly interdisciplinary, with various studies interpreting linguistic data in terms of social theory (early examples include Eckert 1988, 1993; Kiesling 1998).
Use of Foucauldian theory in sociolinguistic analysis
Particularly relevant to this article are those studies that have used Foucauldian theory in their analysis. The earliest of these was undertaken by Fraser & Cameron 1989, who highlight apparent contradictions in the input of individual adolescent speakers into conversation but challenge definition of these as contradictions. They claim that the idea of a fixed social world (and therefore of truth) is a myth, and that we should study instead the ways in which we construct our world through language (1989:28–29). They therefore analyze the “contradictions” in terms of the young people's attempts to make sense of society via the interplay of various discourses (1989:32). In later work, Cameron 1997 highlights the importance of considering both form and content of conversation. In her data, young men use an apparently cooperative style of talk but the content consists of “the same old gendered script” (1997:64). Coates 1999 observes that the adolescent girls in her data use various discourses to position themselves as women: liberal, patriarchal, pseudo-scientific, and so on. At the age of about 14, a change in linguistic features suggests the advent of a new discourse that she calls the “discourse of consciousness raising.” However, this discourse is in competition with the fact that the content of what they say often comes from a patriarchal, repressive discourse.
You know and I know
You know
In the previous literature on you know, there is disagreement as to whether or not use of you know has anything to do with knowledge. Macaulay 2002 claims that it does not, at least for his Glasgow and Ayr groups. He puts forward the argument that you know has been grammaticalized and, as such, any prior meaning has been “bleached out” (2002:755). He gives two main reasons for this. The first, more tentative, is that you know can occur in close proximity to the verb “know” in its basic function. The second, more relevant to my research, is that “‘you know’ is frequently used before or after statements that clearly do not represent shared knowledge” (2002:755). What exactly we mean by “shared knowledge” is important here – in brief, whether we mean knowledge about events or knowledge about norms. I return to this point below in light of my analysis.
Fox Tree & Schrock 2002, in their review of the literature on you know, cite differences of opinion on what random sprinkling of you know tells us about the speaker: Lalljee & Cook 1975 and Ragan 1983 (cited in Fox Tree & Schrock 2002:729) claim it shows uncertainty or anxiousness, while Erman 1987 (cited in Fox Tree & Schrock 2002: 729) claims it shows certainty. Lakoff 1975 and Fishman 1980 also claim the former. As work in this field has developed, researchers have come to see that, since language is always used in social contexts, we cannot take an essentialist approach to function but should view particles as polypragmatic. For example, Cameron et al. 1988 find that tag questions serve different functions for different speakers in different contexts. Holmes 1985, 1986 and Bonnano 1995 (cited in Holmes 1997) find that you know can function either as a hedge or as a booster, the former lessening and the latter increasing the force of the proposition. Holmes 1997 defines such particles as “interactive.” She claims that for this reason they cannot be considered to have essential functions but instead will serve particular social functions for particular speakers. Cameron et al. 1988 also find that tag questions in general serve different functions for different speakers in different contexts. While we must be aware of the fact that every utterance of you know or I know could serve a different function, it is also the case that repeated use within a particular group can result in relatively fixed group-specific social meanings, as Hewitt 1986 found in the “lexical inventiveness” and meaning alteration of older words (1986:114) illustrated by the young people in his study.
Fox Tree & Schrock also outline the work of researchers who have found that, for their data, you know does indicate some degree of shared knowledge. The two proposals for the basic meaning of you know are related but differ in that one invites the addressee to “fill out unspoken intention” (Jucker & Smith 1998, cited in Fox Tree & Schrock 2002:737), while the other checks on or demonstrates shared views. For example, Schiffrin found, in her interviews, that you know sometimes “marks the general consensual truths which speakers assume their hearers share through their co-membership in the same culture, society or group” (1996:274). Ostman's (1981, cited in Fox Tree & Schrock 2002:736–37) conclusion is a slightly more proactive variation of this: that “the speaker strives towards getting the addressee to cooperate and/or accept the prepositional content of his utterance as mutual background knowledge” (Fox Tree & Schrock 2002:736–37). I have found Schiffrin's and Ostman's work to be particularly relevant to my own research. I give reasons for this and position my work in relation to theirs below.
I know
There has not been any work done specifically on I know, as far as I can ascertain. However, we could perhaps view it as a relatively focused minimal response. It has been documented by various researchers (Zimmerman & West 1975, Fishman 1980) that women use minimal responses more than men do in mixed-sex conversation. Coates claims that this does not entail women's powerlessness: Minimal responses are used regularly during the conversations of women who are friends and equals (Coates 1988:106). From the previous literature, we could therefore view I know as a supportive device which encourages the continuation of the conversation.
DATA ANALYSIS: THE USE OF ‘YOU KNOW’ IN THE ACTIVE (RE)PRODUCTION OF POWER/KNOWLEDGE
The tonal nature of you know in these data
The tonal nature of the version of you know dealt with in this article is the same as that documented by Hewitt 1986: a low tone on you and low rising tone on know. I have illustrated this simply below.
This similarity to Hewitt's findings could suggest that the young people using you know in these data are constructing an identity that draws on recognizable markers of ethnicity. In the peer group Hewitt studied, markers of Jamaican identity were drawn upon not only by Jamaican young people, but also by black young people who were not from a Jamaican background, and, more notably perhaps, by white young people. This served to bring more solidarity to the group and to allow the young people to perform their resistance to authority more explicitly.
This version of you know almost always appears in clause-final position, for example you're bad you know (Maria to Jay). On occasion it appears in clause-initial position but can be differentiated from another version of you know that the speakers use as a topic initiator, both tonally and owing to the fact that know is more heavily stressed in the former version. Topic-initiating you know has a mid tone on you and a high tone on know. This is illustrated below.
You know with the latter tonal pattern appears in the following utterance from Kevin: you know you know s- you know sometimes you know sometimes you know when you're having (.h) a- an argument and you're really angry what do you say about them, where it can be clearly seen that he is initiating a topic. The following utterance from Jay includes an example of each type of you know and shows clearly their different functions: hey you know that other Jay he's gay you know.
Constructing self as dominant and other as deviant
Sexuality
In excerpt (1) below, Kevin uses you know to mark Jason's allegedly being gay as salient. He then goes on to reproduce a link between feminine gender characteristics and male homosexuality, suggesting that a boy's being gay is to be viewed as not just salient, but deviant. That said, it can be argued that the main difference Kevin reacts to in “othering” Jason is not sexual orientation but rather, although he does not voice it explicitly, social class. Once sexual orientation is on the agenda, however, Jay engages with it. He identifies partially with the deviant position but does so ambiguously so as to maintain his dominant position. Overall, an interesting interplay between class, gender and sexuality is illustrated in the boys' discourse.
In stave 1, Kevin talks about that other Jay. There is only one boy called “Jay” in the group, the one involved in this conversation. As I sat in on some of the group's drama activities, I was able to recognize from the utterance Kevin recounts that he is talking about Jason. Other parts of my analysis suggest that Kevin is making a salient point rather than an error here: Incorrect naming is a strategy for symbolically excluding members of the group or making them more peripheral.
Jason could be considered an outsider on a class basis. The majority of the young people in this group come from working-class areas in the boroughs of Hackney and Haringey. A minority, including Jason, come from middle-class areas in the borough of Islington. Even had I not known this, it is clear from my data that the young people divide themselves roughly into these majority and minority groupings. This relates to Eckert's (1995) work mentioned above, in which young people reproduce group-specific identities that draw on but are not determined by wider social class identities. The class difference is generally left implicit in my data; the salient categories by which groups divide themselves are sexuality, ethnicity, gender, age, and style. These categories are interlinked in various illuminating ways.
Kevin gives the way he talks (stave 2) as his reason for thinking that Jason is gay and then mimics him by speaking in a high-pitched, feminine way. In doing this, he reproduces a link between male homosexuality and feminine gender characteristics. However, Jason's voice is not particularly high-pitched compared with other boys in the group, so we could surmise that the way he talks actually refers to Jason's accent, which, while not Received Pronunciation, is significantly closer to it than Jay's, Kevin's, and Richard's. If this is the case, then Kevin is reproducing the equating of the working class with masculinity and the middle class with femininity.
In staves 5 and 6, Jay challenges Kevin's marking of dominant discourse. He introduces his utterance with well, suggesting what follows will challenge in some way what has gone before, but the nature of the challenge is ambiguous. He could mean either of two things: (a) that, because his voice is more like a girl's than Jason's, he is more likely to be gay than Jason is; or (b) taking for granted that he (Jay) is not gay, having a voice … like a girl's does not necessarily make one a candidate for homosexuality. In the first of these options, he can be seen to keep open the negotiation of which positions are to be considered dominant and which are to be considered deviant. In the second option, he questions the link between homosexuality and effeminacy, and by extension questions group boundaries and the conditions for group membership. He could be purposefully ambiguous here. He places himself in Jason's place in order to experience the reactions of the other two (option a), but has a “get-out clause” (option b). He thus identifies with what would be seen by the other boys as a deviant position, but this identification is tentative and partial.
In staves 7 and 8, Richard “accuses” Jay of option (a). Jay denies option (a), but this is not entirely successful (Richard continues to make fun of him). Jay then chooses to change the topic, leaving “gay as deviant” as a salient but open topic.
Ethnicity and sexual behavior
Excerpt (2) leads directly on from (1) above. In it, Richard uses you know to mark I'm fucking her and don't use your fingers as salient, suggesting that he is linking the two utterances in some way. The link appears to be that it is acceptable to engage in full penetration with the girl in question, but that other aspects of intimate contact with her are taboo. The boys lead up to the use of you know by differentiating between “sex” and “going out with” people and by differentiating between girls of different ethnicities in regard to whom they would engage in which actions. Overall, the conversation and uses of you know reproduce problematic discourses that link ethnicity with (un)cleanliness, in turn reproducing oppressive power relations in a “micro” context. Although in the previous example differences of social class are not voiced at all, differences of ethnicity are voiced here, but the majority of the specific “othering” is once again done on the basis of gender and sexuality. Once again, there is evidence of a challenge to the construction of the deviant category. However, in this example it is unclear whether this is used to explore identity, or as a cynical strategy to encourage reproduction of the boundary between dominant and deviant.
The construction of dominant and deviant positions in the early part of the extract can be summarized as follows. Jay answers Richard's challenge that he (Jay) is (sexually) deviant by saying his girlfriend is Indian (stave 1). We can surmise from this that he is meeting like with like and thus suggesting that Richard is deviant by associating him with someone who is deviant. We can therefore surmise that Jay views Indian people as (ethnically) deviant (or that this is the view upheld by the peer group). Jay introduces Kurdish girls into this ethnically deviant category (stave 2). When Jay is asked whether he would go out with an Indian (stave 5), he begins to answer no and then says I've only went out with a Kurd (staves 5–6, my emphasis). This suggests that Kurdish girls are less deviant than Indian girls.
Perhaps to underline the partialness of his identification by passing over it very quickly, Jay, without a pause, immediately shifts back from the “real” world to the “possible” world of the modal auxiliary and says in stave 6 the sex would be good (implying “with an Indian girl”). He has previously denied (or begun to deny) the possibility of his going out with an Indian girl and now he asserts the opinion that having sexual intercourse with an Indian girl would be good. This suggests that having sexual intercourse with someone whom one considers deviant makes oneself less deviant than does going out with that person. It is in fact treated positively: the sex would be good (Jay) / it would be quite good (Richard). Perhaps, then, the “otherness” of the “deviant” girl is being interpreted positively (but still problematically) as mysteriousness (cf. Said's [1978] theory of Orientalism).
In staves 7 and 8, Kevin introduces a girl of unknown ethnicity into the conversation. By saying, in stave 8, that Kevin has claimed the girl in question is much nicer than Jewish, Richard marks Jewish people as deviant, possibly more deviant than any group discussed up until this point. This is not challenged; in fact it is not even commented upon, suggesting that such evaluation of Jewish people has been previously cemented into the boys' discourse. Kevin then states but I'd never go out with them, again contrasting sex and going out as acceptable and unacceptable, respectively.
In staves 10 and 11, sexual intercourse is shifted for the first time from the possible or fantasy world into the real world. (It is unclear whether the her Richard refers to is the girlfriend mentioned at the beginning of the extract or the girl Kevin has just referred to.) Richard's use of you know suggests that he is (re)producing the dominant discourse of the group by offering advice or a rule to the other boys. He says that having sexual intercourse with this girl is acceptable but using your fingers is not, and the sex/relationship idea is thus broken down further. He suggests that acceptable sex is restricted to conventional heterosexual intercourse (fucking). He implies that using your fingers would cross the interpersonal boundary at the wrong level and lead to pollution from the deviant other (cf. Douglas 1966). So, Richard's two uses of you know have accomplished two things. First, they have cemented the dominant discourse norms that the boys have been negotiating – that is, having sexual intercourse with someone who is considered deviant is a positive thing whereas engaging in any other kind of sexual activity with that person is a negative thing. Second, Richard has escaped the negative position Jay placed him in by saying his girlfriend is Indian, and repositions himself as dominant.
It is unclear whether Kevin's utterance you're evil (staves 11 and 12) is a reaction to Jay or Richard, but in either case the strength of the reaction suggests a recognition of the salience of any utterance tagged with you know. He could have become aware that Jay has been recording the conversation. This interpretation could be supported by the fact that the tape recorder is turned off immediately after Kevin's utterance. On the other hand, it could show his awareness that what the boys have been discussing would be frowned on in the wider society (i.e., the racist nature of suggesting that a person would be “polluted” in some way by having sexual relations with a person of a particular ethnicity). He may think that Richard's objectification of the girl has gone too far, or he may wish to mark the fact that wider society would be of this opinion. The difference between these possibilities may be subtle, but their functions may be radically different. Is he criticizing Richard? Or is he actually drawing more attention to what Richard has said, thus applauding him for having voiced this taboo discourse so fully?
It is interesting that in the first example above, Richard uses you know as part of a statement of self-reference (I'm fucking her), whereas in the second he uses it as part of a statement of other-reference: He warns against something that would make the other boys deviant. In the first example Richard must regain the dominant position that he has partly lost, whereas in the second example he is simply reinforcing it (or making the shift complete). It could be, therefore, that when dominant positioning is in question, the reference is to self in order to reclaim that positioning. However, when the dominant positioning is unproblematic, it is reinforced by differentiating oneself from the deviant other rather than by restating one's dominance.
Gender issues in the use of you know
For Richard and Kevin, positions are claimed via differentiation from a deviant other provided by a third party. In contrast, when Jay and Maria converse with each other, the speaker positions the hearer as the deviant other. There is a consistent pattern regarding where you know appears. Jay tends to make the evaluation and then back it up, whereas Maria tends to give evidence before making the evaluation. Furthermore, Jay's “reasons” for positioning Maria as deviant involve physical boundaries and pollution, whereas Maria's “reasons” for positioning Jay as deviant involve the morality of interactive behavior. The following excerpts are examples of Jay's use of you know:
In excerpts (3) and (4), Jay evaluates Maria as rude before giving as evidence the fact that she has intruded on his space (ex. 3) and has sat on soiled furniture (ex. 4).
Excerpts (5), (6), and (7) are examples of Maria's use of you know:
In ex. (5), Maria debates the issue of why Jay has been cussing (criticizing) Joanna (staves 4 and 5) before labeling him as bad (stave 6). In ex. (6), she asks him what he means by saying that the space they are in is his house before concluding that it is not his house and therefore labeling him as deviant for not telling the truth. In ex. (7), the evidence comes from Irene (an adult receptionist): you're a bit on the lairy side (stave 2). It would appear that Maria respects Irene's view and thus feels able to use it as evidence in labeling Jay as bad (staves 2–3).
The above examples are evidence that Jay takes a more active/assertive/aggressive stance than Maria in regard to self and other positioning. The following utterance supports this. Maria's position is reactive: She reacts to Jay's accusations and defends herself against them using you know as part of a statement of self-reference:
However, in the utterance below Maria appears to share the other participants' preference for using you know as part of a statement of other-reference. She switches as Richard did in the utterance considered earlier, possibly suggesting that she considers other-reference to be a more strategic choice than self-reference as regards subject positioning.
She begins to self-refer (I tried to explain that I) and then switches to evaluating the situation via reference to the other party, Jay (he's lying). Obviously it is important for Maria to strengthen the force of her utterance and highlight the fact that she is telling the truth (as opposed to Jay, who is lying), and you know helps to serve this purpose. However, it is more complex than this. She fulfills two other functions concurrently with strengthening her utterance.
First, you know once again appears in its less regular clause-initial position. This suggests that she is using it relatively more consciously than at other times. This, coupled with the fact that she includes a third party, Joanna, to give additional evidence, suggests that she feels the position from which she moves to be problematic.
Second, she seems to alter the relative positioning of herself and Jay. During the course of this utterance, Maria moves from a reactive stance to an active one. Her conversation with Jay has been competitive, and Jay has been continually labeling her as deviant to claim a higher status position than hers. In this utterance she begins reactively: She is reacting to Jay's accusations and is about to deny them. However, she alters her position to a more active one by moving out of Jay's discourse and into her own choice of discourse. She uses a moral discourse to label Jay as deviant (lying when he should be telling the truth). So she does not answer him on his own terms but rather actively changes the means of evaluation.
So, in conclusion, we can see that Maria attempts to subvert the power relations that Jay has set up. To begin with, Jay labels Maria as deviant by using you know with other-reference, and his style can be considered active because he makes the accusation and then supplies evidence to support it. Maria uses two strategies. One of these is to label Jay as deviant by using you know with other-reference. However, in contrast to Jay, she first submits the evidence and then concludes with the labeling. Her other strategy is to deny the label Jay has given her by mirroring and counteracting his utterance: She uses you know for self-reference along with a syntactically negative utterance, in this case. All of this, alongside the basic fact that Jay is the first of the two speakers to use you know, suggests that, in cross-sex adolescent talk, you know could be considered to be primarily a masculine feature. Jay uses it to make an explicit verbal attack on Maria; Maria then hijacks the feature and counteracts Jay's attack, fighting him on his own terms; finally, Maria uses the feature in an active way but changes the discourse, moving into a more active position. She is, of course, still using “Jay's” feature, but she has changed the discourse it is used in conjunction with, thus partially subverting the power relations that exist between them.
Analysis of other data from WCG supports the suggestion that you know is a male feature. For example, in a long conversation between Alysha and Katrina, you know (in this tonal form) is used only once, in the following utterance from Alysha, in which she uses you know to claim that her coming late to the group is acceptable because others do:
It is interesting that you know appears in the marked utterance-initial position, suggesting irregular use. (Alysha pauses after cos the people do come late, which suggests that her you know is linked to the following I should come late Irene innit.) Similarly to Maria, Alysha uses it after she has given evidence for her case (other people come late to the group, therefore Alysha herself should come late).
Thus, we can claim that you know is a feature mostly associated with male speakers. The all-boy conversational group in my data use it to position themselves in relation to a third party deviant other by claiming knowledge of the dominant and deviant subject positions within the dominant discourse. In boy–girl interaction, the boy uses you know to label the girl as deviant and thus set up an unequal status relationship between the two speakers. The girls in my data use you know only when conversing with a boy, apart from the special case analyzed above (ex. 10). In the boy–girl conversation in question (between Jay and Maria), the girl uses you know reactively to counteract the unequal status relationship. We also have one example of her being subversive and setting up an opposing unequal status relationship (the girl has higher status than the boy).
Reviewing previous literature on you know in the light of this analysis
As stated above, Macaulay 2002 claims that, for his data, you know has been grammaticalized and thus has nothing to do with knowledge. One of the reasons he gives, and the one that is most relevant to my research, is that it is not used with statements that represent “shared knowledge.” Macaulay appears to be referring to the fact that the utterances in question relay items of information to the hearer for the first time (see also Schourup 1985 cited in Fox Tree & Schrock 2002:735). This differs from my interpretation of “shared knowledge,” which, in the context of this article, has to do with shared knowledge of norms. The analysis above has shown speakers using you know to reproduce norms, and I would claim that you know in several of the examples Macaulay cites can be interpreted in a similar way. The following example from Macaulay's data clearly deals with norms of gendered behavior in a partnership or marriage:
and he said “you didn't tell me you were going out” and I said “Yes I did” “No you didn't” you know so the next morning he was going out to work and he said “Well will I see you for tea tonight?” you know “Will you be home” you know. (2002:756)
It is interesting that the examples of Macaulay's that I would analyze as dealing with shared norms all come from the category in which you know appears utterance-finally, as these correspond with the group-specific use of you know in my data.
Another similarity between Macaulay's findings and my own is that working-class young males in Macaulay's data used you know more frequently than working-class young females, although the difference he finds is not highly significant, and this is the only part of the sample in which males use the feature more than females. That said, Macaulay's you know is tonally different from mine in that it has a falling intonation, and we have seen that the significance of gender in the use of you know varies greatly from study to study; both of these points remind us to maintain focus on the use of pragmatic particles for individuals and specific groups rather than attempting any mass generalizations.
It is also interesting to revisit Schiffrin's finding that you know sometimes “marks the general consensual truths which speakers assume their hearers share through their co-membership in the same culture, society or group” (1996:274). She found this conclusion more likely where you know had a falling intonation. This differs from you know in my data, which has a low rising intonation. Obviously this could be partly to do with idiosyncratic usage in this particular group, but perhaps it is also to do with the fact that whereas Schiffrin presupposes agreement, the young people in my data are still negotiating what the dominant discourses of their group should be, albeit proactively. On this point, of the studies cited above, Ostman's (1981) conclusion is the one that fits most closely with my own findings. To recapitulate, he found that “the speaker strives towards getting the addressee to cooperate and/or accept the prepositional content of his utterance as mutual background knowledge” (Fox Tree & Schrock 2002:736–37). More general conclusions on you know in this data set, and its relationship to I know, will be made at the end of the article.
DATA ANALYSIS: I KNOW
The double function of I know: To mark and claim knowledge of the dominant discourse
Excerpt (11) provides a good introductory illustration of the function of I know for MCG. The girls are discussing a planned visit by someone who wishes to speak to the young people about “teenage sex.” While part of the reason Diane says I know in stave 8 seems to be to mark the fact that she does actually know about this planned visit but had simply forgotten about it, I would argue that she is also marking the importance of what Lana has said about the status of 16-year-olds in terms of social norms.
Lana's words in staves 6–8 superficially comment on the fact that five years (that is, the time between being 16 and being 21) is not a long enough time to merit having forgotten how it feels to be 16. It would seem to me, however, that the evaluation goes further: Lana is actually commenting negatively on the fact that such a visible difference is made between 16-year-olds and 21-year-olds, to the point that the two are seen as being at different life stages. Presumably she thinks that 16-year-olds should be viewed more as adults than they are. Diane both marks this as dominant knowledge and claims the knowledge for herself by saying I know.
In excerpt (12), Tim uses I know to equate long hair with “horribleness.” Thus, even though he has long hair, he still positions himself in the dominant discourse of style:
In excerpt (13), the young people talk about chocolate. After some discussion about giving up chocolate for Lent, Libby uses I know to mark as dominant the discourse about the positive nature of denying oneself chocolate (particularly if this can be permanent).
In excerpt (14), Cassie uses I know to mark as dominant discourse Celia's “rule” regarding the incompatibility of spending a lot of time with a person and saying negative things about her. It is notable that, having positioned herself as dominant by marking and claiming knowledge of the dominant discourse, she then proceeds to develop the discourse herself. This is evidence of the reactive nature of the use of I know in MCG.
I know as a tool in crossing peer group boundaries
During excerpt (15), the girls are looking at a magazine about skateboarding. At the time of data collection, skaters are a visible subgroup of adolescents as a whole and are considered particularly style-conscious. Anna is the only one of the speakers who is a skater, so it is likely that Alison does not in fact know that these ones are cool. She uses I know to align herself with skater discourse and by extension position herself as cool:
Excerpt (16) suggests that Anna considers her full membership in the skating peer group actually to dilute her membership in the general adolescent peer group. At a point when all speakers are concurring once again on the adolescent norm of denying oneself chocolate (all say yeah, and Alison adds it's hard), Anna is the only speaker who uses I know, perhaps because she is the only speaker who feels the need to mark explicitly her knowledge that this is dominant discourse.
As well as having her membership diluted by being a skater, if we consider chocolate denial to be more a case of “doing adolescent femininity” than simply “doing adolescence,” there may be an issue of gender identity here as well. The vast majority of skaters are boys, so Anna may feel relatively more obliged to “prove” her femininity than the non-skating girls do.
Revisiting previous literature on I know in the light of this analysis
Above I concluded that previous literature on I know suggests an analysis in which it functions as a supportive device that encourages the continuation of the conversation. I know appears to serve a more complex function than this for the speakers in MCG. Supportive devices are other-oriented, encouraging the current speaker to continue. The present analysis shows that the use of I know to create a link between previous speaker and current speaker can be seen to benefit both speakers. In fact, it could be argued that the examples of I know in this article in fact benefit the speaker more than the hearer. Bernstein's (1974) research into the discursive styles of middle-class and working-class adolescents may be useful here. Although he does not write specifically on I know, he observes that the first person pronoun is used more frequently by the middle-class young people whereas the second person and third person pronouns are used more frequently by the working-class young people. In addition, he points out that self-referring expressions such as I think are used more frequently by the middle-class young people, while sympathetic circularity sequences (SCs), such as you know, are used more frequently by the working-class young people. This may suggest that the differential usage of I know and you know by the young people discussed in this article may suggest differential orientation toward egocentric and sociocentric modes of interaction, which Bernstein claims are encouraged by socialization into the relevant class background. That said, Bernstein claims that SCs request affirmation and signal feelings of “uncertainty of the appropriateness of the message.” You know in my data does not signal either of these things. This difference could well be to do with the context: Bernstein's young people were discussing a rather formal topic in a relatively formal situation, whereas mine were experiencing informality on both counts.
THE USE OF I KNOW BY THE YOUNG PEOPLE IN WCG
The young people in WCG also use I know to signal knowledge of peer group norms, although they do not use it nearly as often as the young people in MCG, and their use of I know would appear to be more closely related to their use of you know than it is to MCG's use of I know. The marked usage occurs because there are specific reasons why it is necessary to break out of the regular pattern, often because the speaker has been challenged and therefore has to state something defensively rather than proactively. This can be seen in the following example. While MCG use I know to mark what someone else has said as dominant discourse, here Maria uses it to mark what she herself has said as dominant discourse. Maria says that if you say you hate someone you must not mean it. Donna mocks this, and Maria claims that it was in fact logical and relevant by using I know. The context is that Kevin has the tape recorder and appears to be encouraging speakers to disclose their dislike of other members of the group.
CONCLUSIONS
Processes of identity construction: Similar or different?
I have shown that the members of the two peer groups I have studied explicitly mark their knowledge of dominant discourses by their use of you know and I know, and that in so doing they reproduce these discourses. This is evidence that knowing the norms of the group, or not being a deviant, are important to both groups of young people who use these forms.
I would claim that the use of both you know and I know in my data could be seen as “comparative acts” (Seltzer 1989:60). Seltzer describes a “peer” as a “counterpart” (1989:21). She suggests that during the time that adolescents interact in peer groups, their selves are not distinct. She further claims that adolescents build their identities through engaging in “comparative acts” with their peers (1989:60). In both groups I have studied, speakers compare their knowledge with that of hearers. In the case of you know, the speaker does not simply present knowledge; she or he also claims that the hearer's knowledge is comparable to the speaker's own. In the case of I know, the current speaker claims to have the same knowledge as the previous speaker and as such compares herself or himself favorably with that speaker.
That said, the differing status of the hearer in each case is interesting. Speakers in WCG do not use you know to request acknowledgment from the hearer. Why, then, not say I know X as opposed to X you know? Speaker knowledge would still be expressed. It would appear that the speaker, as well as claiming her or his own “knowledge” of the point in question, also wishes to claim the hearer's knowledge of the same. I would claim that this presents the knowledge in the first instance as group knowledge rather than individual knowledge, and as such makes it more readily acceptable as dominant discourse. It may appear, therefore, that dominant discourses are reproduced in an active manner by this group: The discourse is voiced, and the hearers are party to it unless they challenge this position in the following utterance. MCG's process, in contrast, seems more reactive. First, the speaker who says I know has waited for another speaker to voice the potential dominant discourse and then claimed knowledge. Second, the speaker to whom the speaker of I know reacts does not personally claim that her or his utterance is dominant discourse. Thus, the status of any utterance as knowledge is provided by someone other than the speaker who produced that utterance. We can see this as a collaborative venture in which no participant need take full responsibility for claiming that something is dominant discourse.
To end the analysis here is to simplify matters, however. On the majority of occasions on which speakers in WCG use you know, they reproduce the dominant discourse indirectly. They are generally talking about other people's deviant behavior. They therefore claim non-membership in the deviant group. Membership in the dominant group is therefore only claimed by extension.
Wider claims: The relationship between economic conditions and symbolic identity construction
My main aim has been to analyze the group dynamics of these two groups in terms of what they suggest about individual processes of identity construction. However, I would put forward a tentative claim that the processes of identity construction evidenced in this data show the speakers making sense of social class and gender positions in the wider society. Foucault 1977, whose work I have drawn upon here, recognizes that dominant discourses and power relations develop into recognizable structures in society. Also useful here is Bourdieu & Passeron's (1970) notion of “reproduction”: While individuals and groups construct their identities in interaction, the resources they draw upon while doing so are very much tied up with material reality. So the likelihood remains that, while dominant knowledge can be challenged, society will be reproduced in a similar fashion.
I would claim that the speakers who use I know (the middle-class speakers) see themselves as potentially inhabiting a dominant position in society because of their families' (upper) middle-class status. That said, their conversational behavior illustrated in this article suggests that they remain markedly insecure in regard to this positioning, perhaps because, as young people, they have not yet been able to inhabit it in their own right. They do not actively position themselves as dominant: They reactively identify with a dominant position in order not to jeopardize their potential social positioning. The speakers who use you know (the working-class speakers), I would claim, are aware that the likelihood of their gaining a dominant position in the future is not as great as it may be for some others. They therefore actively construct dominant positions for themselves locally within the group. These are constructed relationally by defining other speakers or third parties as deviant.
There are two notable points to observe in relation to this. First, it is predominantly the working-class boys who use you know to construct dominant positions. The working-class girls do use it on occasion, but only when a boy has set the pattern previously. You know does not appear in all-girl talk in this group, so we can surmise that the girls use different strategies to position themselves unless they are reacting to boys' strategies. Second, although I have claimed that the boys are reacting to their position within the social class structure, they do not subvert this, or even voice any recognition of it. Instead, they base their dominance on gender and ethnicity. They often construct themselves as dominant in relation to a deviant female. This is particularly clear in Jay and Maria's dyads. In Richard's discussion of his sexual relations with an Indian girl, he sets “Indian” up as a deviant position and suggests that the Indian girl is even more deviant because she is a girl. When Jay places Jason in a deviant position, it is not on account of his being middle-class: instead, he projects onto that identity a homosexual identity and constructs the homosexual identity as deviant. This again relates to Eckert's (1995) finding that adolescents reproduce society broadly along recognizable social class lines, but they do not voice their difference from others along those lines, instead reproducing group-specific categories.
So, to conclude finally, I claim that the speakers in the groups I have studied use the pragmatic particles you know (WCG) and I know (MCG) both to show their knowledge of the dominant discourses and norms of the group and to reproduce these. Both groups of speakers can be seen to be engaging in comparative acts, and such acts have been documented as playing a significant part in adolescents' identity construction. The use of you know shows relatively active identity construction while the use of I know shows relatively reactive identity construction, but both groups construct their identities indirectly. In reproducing relationships and dominant discourses within the peer groups, the young people can be seen to draw upon aspects of wider society, and there is an interesting interplay of issues of social class, gender, sexuality, and ethnicity in their discourse.
