33
Sui Generis Micro Social Structures: The Heuristic Example of Poker Author(s): Garry Potter Source: The Canadian Journal of Sociology / Cahiers canadiens de sociologie, Vol. 28, No. 2 (Spring, 2003), pp. 171-202 Published by: Canadian Journal of Sociology Stable URL: http://www.jstor.org/stable/3341458 . Accessed: 15/06/2014 00:31 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. . Canadian Journal of Sociology is collaborating with JSTOR to digitize, preserve and extend access to The Canadian Journal of Sociology / Cahiers canadiens de sociologie. http://www.jstor.org This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AM All use subject to JSTOR Terms and Conditions

Sui Generis Micro Social Structures: The Heuristic Example of Poker

Embed Size (px)

Citation preview

Sui Generis Micro Social Structures: The Heuristic Example of PokerAuthor(s): Garry PotterSource: The Canadian Journal of Sociology / Cahiers canadiens de sociologie, Vol. 28, No. 2(Spring, 2003), pp. 171-202Published by: Canadian Journal of SociologyStable URL: http://www.jstor.org/stable/3341458 .

Accessed: 15/06/2014 00:31

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

Canadian Journal of Sociology is collaborating with JSTOR to digitize, preserve and extend access to TheCanadian Journal of Sociology / Cahiers canadiens de sociologie.

http://www.jstor.org

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

Sui Generis Micro Social Structures: The Heuristic Example of Poker'

Garry Potter

Abstract: This article uses aspects of poker (no prior knowledge of the game is required) as a heuristic example to facilitate arguments to the following conclusions. Micro social structures have a sui generis aspect to them. Social structures are simultaneously a set of relations between positions and interactional relations between people. Micro structures have some of the features usually only associated with macro structures but nonetheless there are qualitative differences in kind between macro and micro structures. The article thus retains the insights associated with both interactionist and structuralist positions with respect to sociology's long history of debate about structure and

agency. But it also shows both poles of debate to be significantly mistaken.

Resume: Cet article emploie des aspects de "poker" [un jeu des cartes] (aucune connaissance ante- rieure dujeu n'est exig6e) comme exemple heuristique pour faciliter des arguments aux conclusions suivantes. Les structures sociales micro ont un aspect sui generis a elles. Les structures sociales sont simultanement un ensemble de relations entre les positions et les relations interactionnelles entre les

personnes. Les structures micro ont certains des dispositifs habituellement seulement lies a de macro structures mais n6anmoins il y a des differences qualitatives en nature entre de macro et micro struc- tures. L'article maintient ainsi les perspicacites liees aux positions d'interactionist et de structuralist en ce qui concere l'histoire de la sociologie longtemps de la discussion au sujet de la structure et de l'agence. Mais il montre egalement les deux poteaux de la discussion a confondre sensiblement.

1. I would like to thank the anonymous reviewers of this article for their detailed critical en- gagement with it and their many helpful suggestions.

Canadian Journal of Sociology/Cahiers canadiens de sociologie 28(2) 2(X)3 171

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

172 Canadian Journal of Sociology

Introduction

The intention of this article is to provide a new lens through which to focus attention upon the long standing debates and theoretical issues concerning the relationship between human agency and sui generis social structure. We shall examine the emergence of a particular kind of micro social structure found in a particular example of social interactive phenomena: games of poker. Two

things must be made clear from the start. First, no knowledge of the game of

poker will be required to understand the sociological points made in this article. Good players will find the descriptions of the game's features, examples of strategy etc. relatively unremarkable and incontrovertible. Those who have never played should also find nothing difficult to grasp about such features of the game as are analysed here. On the other hand, those who do

play often yet are not very good at it (which is to say most players) may import some of their poker confusion and misconceptions into the interpretation of the article. This is perhaps unavoidable but if the points concerning structure and

agency are missed, such readers might at least find some suggestive ideas for

improving their play. The second thing which must be made clear is that though poker is a game

and I am using it to make some sociological points, I am not a "game theorist." I have serious reservations about such theory in all its formulations and

applications. Nor do I wish to make the claim that the game of poker is in

every sense analogous to the "game of life." Poker as I shall be using it here is not intended as either metaphor or model. The aspect of poker which I shall

analyse here as a type of sui generis social structure is very definitely not

analogous to social structure in general. It is not intended as a metaphor at all. Rather it is an example of a form of social structure. But it is a form, not the form. One of the crucial points which I wish to make is that there are social structures and social structures.

The example of poker is ideally suited to demonstrate features of social

reality focussed upon by interactionist perspectives. It involves rules, explicit and implicit interpretation (including the body language of unconscious clues), calculation and deception, "impression management," (variably) stable patterns of play, the accessibility and inaccessibility of knowledge, memory and many other subtle features of interaction. In particular, the causal efficacy of belief, including mistaken belief and beliefs about the beliefs of others, is shown to

possess an interactive reality of relation to action and actuality. However, there is also a structural reality to the game of poker, which may

or may not be recognised by individual players yet nonetheless exerts causal force upon the outcome of events. The description of this highlights sui generis structural features of the game which an interactionist account would omit. These features are broadly supportive of a Durkheimian inspired structuralist

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

The Heuristic Example of Poker 173

perspective. Sui generis structure emerges but it emerges in part from inter- action. That is to say that there are many aspects of the formation of social structure which the interactionists have understood correctly.

The presentation of these two perspectives cuts across the grain of polarised positions in the debate about structure and agency. The poker example shows the complexity of self-reflexivity and interaction. However, some of its micro-

sociological features can only be understood correctly through this notion of the emergence of sui generis social structure. Structures emerge not only through the repetition of patterns of interpretation and interaction between self- reflexive human agents but also through sets of relations with other larger scale structures and a form of structural causality. Aspects of both the traditional

poles of debate thus are correct ... and also wrong. The poker example makes it clear the way in which this is so.

Poker structures are micro structures and one of social reality's features I wish to clarify in this article is the manner in which micro and macro features are different. One of the arguments made is that differences in scale (that is to say quantitative aspects of phenomena) produce differences in kind (i.e. qualitative difference). The focus upon poker, it is hoped, will at once make this conclusion both clear and obvious and also open a new dimension of reflection upon and empirical research into social structure.

Methodology

The arguments presented in this paper rely upon hypothetical examples rather than an empirical study of actual poker games. Poker is being utilised here heuristically to make theoretical points. Poker also serves well here as an example of the necessary connection between the empirical and the theoretical and its dialectical nature. The hypothetical examples used certainly do derive from an empirical study, albeit not a scientifically controlled one. The formu- lation of hypothetical examples derives from the author's and others' experience of playing the game and self-consciously reflecting upon this experience. The examples may be hypothetical but they nonetheless possess an empirical grounding in reality. A reflection upon the nature of this ground- ing provides us with justification for their use.

First, the hypothetical examples chosen possess a heuristic superiority to the examples which might be culled from an empirical research project. In part this derives from exaggeration. The examples of "bad" strategy and "bad" players in the next section can be shown most evidently to be so in the extreme case. Players less consistent and less extreme in their styles of play (as might be found in an actual game) do not make the reasons as to why their strategies are so obviously bad (particularly to the reader not familiar with poker).

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

174 Canadian Journal of Sociology

More importantly, however, are the differences with respect to the potential knowledge obtainable through an empirical research project into an actual game (or games) and that which is affordable through an analysis of empiri- cally informed hypothetical example. There are very real practical barriers to even the potential of obtaining some kinds of knowledge through empirical research projects. Knowledge (situationally limited knowledge, unequally dis- tributed knowledge) is a very crucial component of the game of poker. Some of the "hypothetical knowledges," which are crucial to the heuristic purposes of the examples chosen, simply could not possibly be made available to the

empirical researcher observing (or participating in) an actual game. Concealment and deception are very much a legitimate part of poker. Let

us use here a hypothetical example to demonstrate the limitations to any empirical research study of the game. First, let us imagine the researcher as

participant observer. In such a role the limitations to researcher's knowledge are the same as those of any of the game's participants. A bet is made and the

participant observer decides not to call it. S/he may never know if this was a wise decision or not as the cards held by the winner may never be shown. However, it is possible to move back and forth in the hypothetical example from a "God's eye view" of the interactions to the limited knowledges and

perspectives of the game's participants. Let us imagine the researcher now as a non-participant observer attempting

to achieve this "God's eye view" of the game. S/he walks around the table

looking at all the players' cards. It is hard to imagine situations in which this would be permitted. The researcher would be doing more than making the

players nervous. The researcher, putting aside the concerns about cheating that would be in the minds of some, nonetheless could influence outcomes in the

game. Players would have not only the body language of their opponents to

study but also that of the researcher. However, even allowing for the possibility that a researcher might be allowed to view all the cards dealt as a game unfolded s/he would still be a long ways short of the God's eye view afforded

by hypothetical example. The researcher would not be able to see into the minds of the players. We shall later be showing just how important are beliefs about the beliefs of others ... and beliefs about the beliefs about the beliefs of others.

While the above points perhaps seem somewhat obscure because of their abstract nature the actual analysis of strategy and explanation of the game in the next section should make them clear. There are absolute limitations to

knowledge achievable by any poker player concerning what is going on at the time in a particular game. Players can infer whether their opponent really has a pair of jacks but the inference can be mistaken. Players can likewise make inferences as to the beliefs and states of mind of their opponents and, of course, also be mistaken. These limits to the players' potential knowledges are as well

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

The Heuristic Example of Poker 175

limits to the potential knowledge of an empirical researcher. However, as I shall demonstrate, such barriers, (some of them at least) are capable of being transcended by other means. In general, we can add here, without at all downplaying the value of empirical research and carefully controlled scientific study, that the limitations of such are not the limitations to potential knowl- edge. Such is the value of theorising and close reflection upon hypothetical example that it can extend knowledge beyond the barriers reality imposes upon empirical research.

Poker: General Features

The first point to be made about poker is that it is a game of skill. "Luck" is also a feature of the game but good players do not win because they are

"lucky." What cards are dealt to which individual player (in a fair game) is purely chance. Thus, a complete beginner could easily win a hand against the best poker player in the world. But winning a hand is not winning the game. Winning is measured in money. In the long term "luck" will tend to balance itself out. Why? Because the good players will tend to lose the minimum amount of money on their losing hands and maximise the money won on winning hands.

How does this happen? Let us use an example of a particularly bad strategy to illustrate how the process works: only bet heavily on "sure things." The player folds (does not call the bet) and so is out of the betting early on in

nearly every hand. S/he is waiting for a very good hand before risking any bets. Such a strategy will result in losing a small amount of money each hand as every player must "ante up" (pay in a small amount of money at the beginning of every hand). No matter such minuscule losses, says the wisdom of this stra- tegy, when I get a "good hand" I will win. The good hand is finally dealt and the player of course wins it. But how much did s/he win? Nearly nothing, only perhaps the ante of the other players. "Bad luck," says the bad player; good luck to get the good hand, but bad luck that nobody called the large bet made upon it. But, of course, luck had nothing to do with it. When the player got his winning hand everyone knew s/he had it and no one was going to call the bet. The other players did not arrive at this knowledge through cheating, they arrived at it through a reasonable (and easy!) inference based upon a sudden

large bet and previous super-cautious play. Perhaps the bad player helped their inferential process along by jumping up and down in his seat. Such "subtle" clues of body language and facial expression are also part of the game.

Then there is the wild player. S/he will attempt to "bluff' the other players into folding all too frequently. But they do not fold because they know that most of the time it is a bluff. This player will also call the large bets of other players under the mistaken assumption (or because of other irrational motivation) that they are usually bluffing. So most often s/he loses.

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

176 Canadian Journal of Sociology

A third example of bad strategy is less obvious than the previous two. There is a purely mathematical aspect to the game and some bad players guide themselves predominantly in relation to it. All players must make (tentative at least) predictions as to which cards will be dealt in the future and take

probabilities into account on some level. But while good players generally do this quite well, it is not their ultimate strategic guide. Why focusing upon the mathematical aspects of the game is not a good strategy, illustrates why poker is so interesting.

One cannot predict blind chance but in poker one is (usually) not wholly blind. Rather, there are probabilities and clues. For example, in many types of

poker played (stud poker such games are called) cards are dealt face up as well as down. One bets on each round of cards dealt. Thus, you are not betting on the ultimate hand you end up with but in some sense upon a calculated guess as to what hand you are likely to get if you stay in the game. Let us imagine in a game of five card stud you have been dealt three cards out of the five you will receive if you stay in until the end. You have been given two kings. What are the "odds" of you finishing with four kings? They are variable. If you can see that two other kings have been dealt to other players already, you know the odds absolutely i.e. that they are zero. If you can see one king has been dealt to another player you know that the odds appear to be significantly less than if you cannot. Players make such calculations in a relatively precise fashion or

crudely. The latter is expressed in such old poker adages as "never draw to an inside straight." Why not? Because the probabilities of getting it are low.

Dramatic mistakes in this calculation of probabilities may be made. The

player with two kings may stake a lot of money upon receiving a third, unfortunately having failed to notice that the two other kings have already been dealt. This too is a skill dimension of the game; the awareness of which, may tempt the mathematically inclined beginner to a bad strategy. That is, this side of the game is important, very important; but if one concentrates upon it to the exclusion of the other sides of the game one will lose. Such bad strategists will often tend to attribute the winnings of players apparently paying little attention to poker probabilities to extraordinary luck. Having discounted "luck" themselves in favour of the mathematical rules of probability they are

perplexed as to why they lose. They lose because probabilistic calculation is

only part of the game. Such knowledges have to be set beside other "clues" and

knowledges available. For example, if your opponent has a pair of kings showing and you know

that your hand beats this but does not beat three of them, then your decision to call the bet rests upon whether you believe s/he has (or has not) a concealed third king. You know that you have a king yourself and this in probabilistic terms reduces the probability that s/he has that third king. But this is far from all the relevant information that there is. You know your opponent's past

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

The Heuristic Example of Poker 177

betting behaviour. Are they anything like the first example of bad strategy? If so, then you probably should fold whatever the probabilities. And what of

physio-psychological knowledge? You watched your opponent's eyes as they looked at their cards. Their pupils dilated. This indicates that they were pleased with the third king they were looking at. Or perhaps you observed an association over time of a particular unconscious nervous mannerism that was manifest every time they bluffed. Such "knowledges" may be of far more relevance in many situations than the laws of probability.

The above examples illustrate some of the most important general features of the game and how skill (many different sorts of skills) interacts with chance. However, there is still more. Each player makes decisions (to raise the bet, to fold, to call etc.) in relation to the actions of the others. Each player takes into account the past performances of the others in attempting to "read them" and thus to predict their actions. Each player is aware that all the others are also

doing so. It is with respect to this that the common poker knowledge of non-

poker players is somewhat misleading. Everyone has heard the expression "poker face." The "poker face" is the player's attempt to prevent others from

"correctly" reading them. The word "correctly" in the preceding sentence indicates a disjuncture between the good player's understanding of "poker face" strategy and the non-poker playing public's understanding of it. What is at issue is impression management and intentional deception. The "bland poker face" displays no emotion. The good player does not shout excitedly when dealt a card s/he needed. The good player does not (usually!) wish the other

players to know whether or not s/he has a good hand. S/he works hard to maintain control over facial expression and unconscious body language. But sometimes the good player wants the opponent to believe that unconsciously something has been let slip. For example, s/he may have spent the last three hours stacking and unstacking their chips every time s/he had a good hand. But it was a ruse, part of a plan leading to a really big hand in the future where the

opponent will make a costly betting decision based upon a mistaken belief

concerning psycho-physiological pattern. This is all legitimately part of the game and everyone is aware of it ... in

general. But games of bluff are actually games of double and triple bluff e.g "the logic of probabilities says I should do this, which s/he can clearly see but s/he will expect me not to do it for precisely that reason, so I will do it; and this will change his impression of me from a wild player to a conservative one, which I am not, etc.." There is no end to the double triple guessing and

counter-guessing, pretence and pretence of pretence. There is no end to the

potential levels of impression management and prediction; except that one must choose and act in relation to the choices and actions of others.

The actuality of the cards that are dealt, and the mathematical laws of

probability, are for poker analogous to physical reality and the laws of nature

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

178 Canadian Journal of Sociology

with respect to social interaction in general. Physical reality, the laws of nature, exist independently of our alleged knowledges and/or relatively true and false beliefs, perceptions etc. They exert causal influence upon events. So too with the cards actually dealt in poker. However, knowledge and belief (including false belief) exert their own causal force upon events. Jo really has or does not have three Queens; this is so whatever you may believe is the case and your belief doesn't affect that reality. But Jo may believe that you believe she has only two Queens. However, the reality of your belief that she has three, is independent of her belief to the contrary. Both your belief and her's affect

your actions, the betting decisions you each make. This notion of independ- ently existing social realities is crucial and we shall come back to it later.

The conceptual intertwining of meaning and reality is integrally connected to the intertwining of structure and agency. Among the structures which emerge in poker are structures of meaning, structures of interpretation. The on-going strategy of the game is to attempt to create false impressions (or at least

unpredictability) with respect to one's own play but to try to correctly "read" the other players. This is a highly complex process with many levels of different sorts of skill involved. An example should illustrate this complexity. There are two people left in the hand. It is your turn to either call Jo's bet, fold or raise the bet. You have seen a number of the cards dealt: all of yours, some, but not all of Jo's, and some, but not all of those of other players. If the bet is raised by you and Jo does not call, you win. If you do not call the bet Jo wins. If the bet is called more cards will be dealt. Who wins or loses thus only partly depends upon who really has the best hand at the time of the betting decision, and also only partly upon who would end up with the best hand if all bets were called until the end. Probabilities are relevant but only partially. Actuality is relevant, but only partially at the moment, and only potentially with respect to the future. The things of greatest relevance to your decision to call raise or fold is your belief as to what cards Jo has, your belief as to what she believes you have and your assessment as to Jo's poker playing character. The reason why these are the most important factors is that one can win or lose in poker quite independently of who really has the best cards or even the best chance of end-

ing up with the best cards. Jo could have easily by far the best hand but if s/he is scared into folding you win. On the other hand, if Jo really has the best cards and your large bet is an unsuccessful scare tactic you have compounded your losses. Thus, beliefs and knowledge about Jo, including belief about her beliefs about you, are every bit as important as the reality of the cards. Let us make our

example more specific to further demonstrate the complexity of the interaction. You are a crafty player. You have been setting a trap for Jo and the others.

You have played over-conservatively for some time at some expense i.e. the other players have labelled you as conservative in their minds and tend not to call your big bets. You have further projected a false "unconscious" tell-tale

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

The Heuristic Example of Poker 179

sign: every time you have a very good hand you light a cigarette. You have done so often enough so that you "know" that "everyone" has noticed this. So, you light a cigarette and raise Jo by the maximum bet allowed by the rules ... and you lose a lot of money.

Why? Because your hand was very bad indeed and you were bluffing against Jo's real "ok" hand. But why didn't Jo fold? The first possible reason is that Jo didn't fold because she wasn't a very good player. There are times in both poker and life where insensitivity, foolishness and ignorance are triumphant. Jo simply hadn't been paying enough attention to your past conservative betting pattern. She didn't get the message that you only bet big on sure things; and she didn't make the connection between your lighting up a cigarette and having a good hand. Instead, she had concluded that you quite possibly had not much of a hand because she often bet big on mediocre hands. You lost a large amount of money because you had gotten too clever, because

you had "placed pearls before a swine." Jo's luck or your mistake? Your mistake ... because one can usually smell a swine. You should have known that your subtlety was going to be wasted on Jo and that her calling your large bet was entirely predictable.

A second possibility is also not flattering to you. Jo was a much better

player than you realised. Your subtle strategy of misleading clues was not so subtle. Jo saw right through you. She knew the cigarette lighting clue was

phoney. But unbeknownst to you, you projected some real body language indicators - an almost, but not quite, imperceptible hand shake at a crucial moment.

The two possibilities outlined above demonstrate the importance of belief and knowledge in the game. Knowledge is never absolute beyond the particular cases when all the cards are dealt and displayed but players have beliefs about the cards other players have. Players also have beliefs about the "character" and strategic orientations of the other players. These beliefs are sometimes correct and sometimes not. Indeed, how often one's beliefs are correct is one of the principal factors that distinguishes good players from bad ones. The (extreme) bad player examples heuristically used above are known to be conservative or wild to (nearly) everyone. This is because of their strict conformity to a pattern and the other players' memories of it. This would also be true of the "ideal type" of the mathematically governed player. They are so consistent with their probability calculation betting decisions that to "read" them one need only calculate the probabilities oneself. But these extremes seldom occur in actuality. Players exhibit inconsistency and the patterns are only tendencies. Players are also deliberately attempting impression manage- ment. Poker skill involves the "correct reading" of the other players in spite of their attempts to prevent this. One could create here two other extremes: good players possess knowledge while the bad players have false beliefs.

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

180 Canadian Journal of Sociology

Let us look at some more strategic nuances to further illustrate the com- plexity of the game. Many poker hands are not played out to the end and the last possible betting round. That is, one player makes a bet at an earlier stage in the hand, perhaps even in the first round of betting and everyone folds. Let us imagine a player has four spades showing. This means s/he potentially has a flush (five of the same suit - a very good hand). Whether or not s/he actually has this flush is not known to the other players, as it is dependent upon whether s/he has another spade among the cards dealt face down. S/he bets heavily and all the other players fold. S/he has won the hand and the money but still has another decision to make - to show or not to show the "hole" cards (those dealt face down) and answer the question in the other players' minds: did s/he or didn't s/he really have that flush? Usually the cards will not be shown and the question will remain unanswered. Usually. It is a matter of

strategic decision.

Impression management involves more than the conscious or unconscious, true or false, projection of body language and control of facial expression. It also involves a selective and partial display of what is allowed by the rules to remain hidden. Thus, in the above scenario the player with the four spades may wish to show that s/he did not really have a flush. Why might s/he wish to

display this information? It is certainly not intended to give the other players useful information. Quite the contrary. To show that you sometimes bluff is

necessary if you both really are a rather conservative player and also have the

reputation of being so. If the other players believe that in the "four spades showing situation" you would only bet heavily if you actually had the fifth one, then your amount of money won will be considerably lower when that is

actually the case. They will never call you in such situations. In general what

you want, of course, is for people to call you when you are not bluffing and to fold when you are. The selective display of information is actually an attempt at disinformation.

Impression management thus works upon more than one time scale. There is the time scale of the particular hand and the timescale of the longer term in which impressions of your poker playing character is established, maintained or transformed. For example, a large bet upon four spades showing is

signifying the message: "I have a flush." That is, it is signifying this upon one level, that of the "poker universal." It means the same thing for everyone mak-

ing a large bet in such situation. However, on another level, its meaning is variable and individually specific. It is dependent upon the other players' beliefs as to what sort of player you are. Do you often bluff or not? You wish to be mistaken with respect to this both in general and in the specific case. If

you show your "hole" cards all the time they will easily be able to make an accurate assessment. If you never show them most players may still be making an accurate assessment. Thus, you sometimes reveal your cards to offset long

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

The Heuristic Example of Poker 181

term accurate impressions of your betting character or to reinforce incorrect

readings of this. Thus far, sociologically speaking, one might be tempted to conclude that

poker is ideal subject matter for a symbolic interactionist investigation of its symbols, its predictions and deceptions. One might even think its theorisation and description of the relation between structure and agency is adequate to the task as far as poker is concerned: complex double bluff perhaps, but an interactional account of social structure could only be confirmed by it. How- ever, while some aspects of such theory are indeed confirmed, others are not.

Let us return to the previous example of the "crafty but apparently not so crafty" player's situation considered above. S/he had played a (relatively) long term strategic game of impression management and it failed. The bet had been called. The deceptive clues which had been placed had been missed and the

past history of play leading to this situation had been ignored or forgotten. The hand was actually lost because the opponent was a worse player than was believed. Or, so we concluded above, the hand was lost because "Jo" really was a much better player. The ruses had been seen through and other more "truthful" signs had been there to be read. But there is another aspect to the situation quite apart from what we have considered thus far. There were structural poker social forces at work! These forces transcended each and every player's individuality. Thus, let us move now to the particular aspects of

poker which clarify the emergence of micro social structures and their sui generis character.

The Betting Pace: An Emergent Social Structure

In every poker game a betting pace emerges from the social interaction of play as a changing and relatively malleable but nonetheless sui generis social structure. To explain this emergence we must first perform a "bracketing operation." That is, the betting pace, like every micro social structure, is set in relation to other structures, which also exist in a complex relation to interaction and agency. Thus, the individual character, the style of play and the behav- ioural patterns of every player, are in some sense determined (in exactly what sense will later be clarified) by broader social structures. That is, for example, the class backgrounds, ethnic cultural heritages and genders of the particular collection of individual players in any particular poker game are relevant. Players do not go into any poker game as blank slates with their behaviour and beliefs wholly interactively formed by the actions of the other players in that game. They are the products of their past quite outside the small scale space/time continuum of a particular poker game. Their judgements and actions in the game reflect this. Nonetheless, at this point we must bracket such features off ... though only, of course, to reintroduce them later in the article.

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

182 Canadian Journal of Sociology

It is necessary to do this because it is precisely at this point that the ongoing confusion with respect to the structure/agency debate in sociology is most acute. There not only is a failure to properly recognise that differences of scale matter with respect to the question of agency, but also the sui generis nature of micro-structure has been misunderstood with regard to its interactive character ... most crucially by interactionists. So we will first consider the micro structure of the betting pace abstractly, where the players involved are not real people, with life histories, characters etc., formed before and outside of the game. Rather we will consider them solely as self reflexive interacting entities responding to the conditions of the game as they are manifest in

particular situations. What is the betting pace? This is an interesting question in the first place

because it is not in fact asked by many players. Most players are entirely un- aware that there is such a thing as the betting pace. Even many quite good players are unaware of this. That is, they are unaware of it consciously. But

consciously aware or not, each and every player responds to it. This feature of the betting pace in poker makes it an example of sui generis social structure that brings out features of structure quite inaccessible to other more obvious

examples. Let us consider Durkheim's examples of language or currency for instance. While many (most!) people do not perceive language as a social structure in the same manner Durkheim did, they are all aware of its existence as some sort of phenomenon. His arguments concerning externality and constraint are readily appreciable by everyone on one level or another, even if one does not arrive at the same conclusions as he did. However, the betting pace in poker is not even perceived as existing by many whose actions are in fact significantly "determined" by it. I played poker for a number of years without in the least being aware of the existence of such a phenomenon.

So, again, what is the betting pace? It is a complex multi-faceted phenome- non. Let us look at its emergence. Individual players change their own ten- dencies towards either conservative or very loose play in response to the past play of others. Every poker game will go through periods of collective action that will tend to either the conservative or wild betting pattern poles. Each

player makes their own decisions. Each player makes their own interpretations of the cards dealt and of what is going on with the other players. This is so at each and every point in a game or a hand. Nonetheless, collective patterns of

play emerge. Hands and hands go by with everyone seemingly afraid to risk a bet on any but the surest of sure things. Hands and hands go by with everyone seemingly willing to risk everything on anything.

Individuals retain their poker playing individuality during this. The more conservative of players will tend to be more conservative than the majority even when everyone is playing relatively conservatively. Wild gamblers will tend to be even more wild than usual during the collective turns towards the

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

The Heuristic Example of Poker 183

bet happy pole. Or rather usually this is so. Or rather this is so for most con- servative and wild players most of the time. But what does the good player do? The good player (consciously planned as a strategy or unconsciously decided upon moment to moment as a tactic) attempts to go against the betting pace. The strategic wisdom of this is fairly obvious. If people are tending to call anything, then bluffing is a foolish thing to do. The wise player does not attempt it but waits for really good hands before s/he bets. If people are

displaying a tendency not to take chances, then bluffing and heavy betting can "buy pots" (win because all the other players have folded).

We wish to argue that what is described above as a collective pattern of

play is in fact an emergent sui generis social structure. But before proceeding with this let us ask and answer one more preliminary question: is there always a betting pace? Yes and no. The betting pace by definition is a phenomenon which persists over time but there is a point at which the time frame diminishes into nothing. The previous hand everyone was betting wildly but this hand

everyone tightens up. Does the betting pace change within the time frame of a single hand? Does the betting pace change within the time frame of a single round of betting (there are usually many within a single hand)?

Let us consider a physical analogy to answer these questions. There is a

point where descending precipitation crystallises into a snowflake, which perhaps, if this point occurs exactly upon the embryonic snowflake structure

touching down upon the warmer temperature of the ground, it dissolves

simultaneously. Or perhaps "simultaneity" is only an abstract conception in such cases and really it is a matter of micro seconds that the crystalline structure exists. No matter; two things are clear: 1) snow flakes sometimes exist and possess structure; and 2) the endurance of physical structures over time is temporally variable from near non-existence to the age of the universe. I would argue that this temporal variance is also true for social structures. Such inter-generational stability may emerge in human social life as to produce structures as would last as long as the human species. On the opposite end of the scale the existence of the structure may be so brief as to approach non- existence. The betting paces of poker are perfect examples of the latter, though they may also sometimes be far more relatively enduring.

A reflection upon the temporally relative endurance of the betting pace poker structure should help us to clarify its emergence. Let us look at a

hypothetical scenario. For the past six or seven hands the first cards dealt to all the players are relatively poor. The player with the best hand (in reality) makes a cautious bet and everyone else folds. This happens repeatedly. Then a player who does not have the best hand (in reality) makes a relatively cautious bet and again everybody (including the player who in reality has the best hand) folds. Such runs of betting occur frequently. Why? In the first place, there is the distribution of the cards dealt. This (in an honest game) is purely a matter of

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

184 Canadian Journal of Sociology

chance. The cards could be dealt so that everyone receives three of a kind of something - an extremely improbable event - or such that nobody receives much of anything - a rather more frequent occurrence. The cards actually dealt affects the players' betting. But this, of course, is not the only influence upon their play. Not only do different players make different decisions with respect to the hands they are dealt but different groups of players would play out the same distribution of cards quite differently as well. They would do so because they are all making informed guesses as to what cards the other players have been dealt and how they are likely to bet and adjusting their actions accordingly. Nonetheless, even if different players have left the game and been replaced by others during this seven hand run of cards the betting pace will likely still endure. One can observe this in casino play where players enter and leave games frequently and the betting pace endures even across changes of dealer.

The betting pace will consciously or unconsciously affect all the players' "informed guesses." That is, it is itself one of the pieces of the puzzle applied to arriving at predictions and decisions as to tactics and strategy. If, for whatever reasons, no one has been betting much, then this past is used (in part) in arriving at predictions as to how they all would potentially react to bets of various sizes i.e. there would be an increased probability that everyone would fold. This last prediction, of course, is tempered by the awareness that most of the other players are also aware of this and that there is a strong possibility that after a long time of cautious betting and quick folding someone will try to "buy the pot." The collective awareness of this possibility tends to moderate its actual occurrence and prolong conservative betting paces.

It is possible that at this point the reader may have arrived at a slightly incorrect understanding of what the betting pace is. It may seem from the

preceding description that it is something like a "collective mood." This is partially true but only partially. The betting pace is far more concrete. It is

interactively shaped by multiple factors, only some of which connect up with the conscious or unconscious aspects of "mood."

The reality of the run of cards dealt is concrete and is usually a partial causal factor. The players as well as interacting with each other are also interacting with the cards. It is possible to have a long string of cards dealt such that nobody receives a pair before everyone folds. Of course, the

qualification "before everyone has folded" precludes the real possibility that someone could have received a pair with the next card dealt ... if they had only called the bet. This, perhaps in part occurring because of "something like a mood," produced two real effects: a collective and individual (to varying degrees accurate) memory of the cards dealt and the actuality of the cards dealt (influencing the collective and individual memory). This in turn effected the "collective mood" of the next hand and so on with a reciprocally influential mutual determination -a kind of snowball effect.

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

The Heuristic Example of Poker 185

This mood/cards dealt interaction, of course, is not the whole story. As has

already been said the causes for the emergence of the betting pace are multiple. Intertwined with the card influenced mood is the (particularly the very recent) past of winning and losing, and the amounts of money won and lost, and the way the money was won or lost (i.e. through having had a moderate hand and

losing to a better one or through having attempted to bluff and been called). This is an individualised effect. However, there are periods when nearly all the

players have had a recent costly losing streak. There are sometimes periods when nearly all the players are close to having lost all their money. These factors interact with the others. In fact, each and every factor interacts with all the others. The factors range from the ephemeral vague and unconscious to the

very concrete and/or conscious. The betting pace does involve "something like" a "collective mood" and appears somewhat ephemeral and nebulous for that very reason. But it is also very material. If the players were totally unobservable to a disinterested observer who saw only the basic poker realities of call, raise or fold, that is, the players' bets, the betting pace would still be evident in this very concrete form.

At each decision making point for every player the "informed guesses," predictions etc., are individualised. That is, (most) players will evaluate (whether correctly or not is another question) a certain sized bet from John or Mary differently. The very worst of players, of course, are free to import an

egalitarian strategy to the game and respond identically to the cards and bets of any player. But this is unwise. This poker tactical error is analogous to the theoretical error of many structuralist and poststructuralist literary theorists

implied by the slogan the "death of the author." The insights taken from the

analysis of the synchronic structural production of meaning, I have argued elsewhere (Potter, 1999), become distorted if taken too far. "Authorship" can

profoundly matter in the production of meaning. It is the same with betting in

poker. We will come back to this point. However, as well as such "individual" factors there is also a collective response to the betting pace.

One could argue that the betting pace simply is a collective general pattern of play that persists over some limited time frame. However, while on one level this is true, on another it is somewhat misleading. Putting it crudely, an interactionist asks if the whole is equal to the sum of its parts and answers yes; while a structuralist asks the same question and answers no. They are both

right but upon different levels of analysis. Imagine a poker game with seven players in it. But at this point in time there are only three left in the hand. The three are interacting with one another, attempting to "read" each other, to pre- dict each other, to deceive each other etc. They are also interacting with the other four players not in the hand. This is part of their general long term strategy of impression management. However, the four players no longer in the hand are not interacting with them (at least not in the poker terms of bet, call or raise). Yet the past play of these four players has contributed to the emergent social

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

186 Canadian Journal of Sociology

structure of the betting pace. The betting pace emerges out of the past patterns that are "crystalized" in the present as something other than, and more than, the sum total of interactions. There is an interaction not only between players but with the cards actually dealt. There is an interpretive interaction with the other players in part informed by the reality of the past but also in part with a collectively reified interpretive memory of it. This "congeals" to form a discursive structure of the meaning of bets that transcends individual vagaries.

The betting pace demonstrates classic features of social structure. It is external to and independent of any individual player. It exerts constraint upon their actions. It also enables certain actions that would not otherwise be possible. As said before, this response to and awareness of the betting pace, may or may not be conscious. But conscious or not, not only are the decisions made (to raise or fold etc.) constrained by it, so too are the educated guesses and predictions of the others, which on one level motivate the decisions. Players are responding not only to other individuals but to an emergent phenomenon of which they themselves are apart, but which begins to manifest itself as transcendent of individual interaction. It is an emergent structural reality to which they contribute to both its reproduction and its change but yet do not wholly control.

It is possible that thus far we have over-emphasised the rational over the irrational in terms of players' judgements and behaviour. However, in a game of poker the two coincide and interact in a complex fashion. For example, let us imagine a particular player, we'll call him "lucky Jim," who has been

having a very good run of luck. This player not only received the best cards many times in a row but consistently bet heavily, was called and won big. Now he has placed a large bet upon the table. Among the factors "superstitious John" takes into account in making his "guess" as to whether "lucky Jim" is

bluffing or not is the immediate past run of good luck. "Lucky Jim" really did have this past run of luck; that is, to say he really was "lucky." One could in

theory be dealt four aces forty times in a row, even in an honest game (though the odds against this happening are astronomical). But there is also the cor-

rectly mathematically grounded old poker saying "the cards have no memory." However, not all players act on such ancient wisdom. "Superstitious John" is more inclined to believe that "Lucky Jim" really has another good hand because he had a lot of them in the recent past. Rationally he should not make his "educated guess" as to "Lucky Jim's" hand on the basis of such a factor.

Rather he should be attempting to predict "Lucky Jim's" hand according to factors which really are relevant. For example, he should try to decide if Lucky Jim was the type of player who upon receiving a bad hand would switch to conservative play and fold, or one who instead of simply counting his lucky streak winnings, would try to cap off this streak with a bluff. And "Lucky Jim," if he was not only "lucky" but also a good player, would certainly take into

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

The Heuristic Example of Poker 187

account the knowledge (based upon past observation) of other players as being "superstitious Johns" or not, and try to use his streak of good cards to substantiate a bluff after the luck turned ... and conversely to refrain from such a temptation if they were not.

Essentially, both rational and irrational beliefs are utilised in the prediction of other player's beliefs, including whether or not they are irrationally formed. All these factors interact with one another. Good players play differently with others that they recognise as good players than they do with ones who are not so good. Nonetheless, the correct or incorrect, rational or irrational, beliefs and decisions of both good players and bad, produce and reproduce a collective

product: the betting pace. My argument is that the betting pace is a relatively short lived and changing

phenomenon but nonetheless is a sui generis social structure. It is likely, however, that the critical reader, most particularly those of an interactionist

persuasion, would see the argument for the sui generis character of this social structure as yet to be proved. It could be argued, for example, that the

persistence over time of patterns to social interactions can be wholly explained in terms of the interactions of the individual participants. The symbolic interactionist would likely understand the bets, conscious concealment and

deceptions of body language as a form of conversation, part of the symbolic order present in every social interaction. And they would not be wrong in this. Where they would go wrong, however, is in believing that through an

exploration of the full complexity of this collection of individual interactions, readings, mis-readings, self indications etc. etc. (in theory perhaps infinite in its complexity) they would have exhausted all the levels of the phenomenon.

The question concerning the sui generis nature of social structure in one form or another is repeated through the entire history of sociological theory. Macro theorists tend to answer the question in the affirmative; micro theorists tend to answer it in the negative. The smaller the scale of the social phenome- non in question, the less stable and the more subject to individual agency a

pattern of social behaviour is, the more plausible it seems to explain it in terms of a simple multiplication of individual interactions. Macro theorists see this form of explanation breaking down when applied to larger scale and relatively more enduring social phenomena. The micro interactionist theorists would say that indeed they do have a concept of structure - it is the patterned outcome of many many individual interactions. But to the macro theorist this is no theory of structure at all. Where does it fall short? To answer this the macro theorist only need return to Durkheim as recounted above. The English language and our system of currency were here before we were born and presumably will outlast us. But the changing betting paces found from one moment to the next in poker games began with us and perish long before we do. We cannot through an effort of individual will change the grammatical

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

188 Canadian Journal of Sociology

structure of the English language but a few rounds of a few individuals' wild betting will likely alter a conservative betting pace. The whole dynamic of agency and constraint appears to be different. And it is precisely here that the micro and macro theorists, armed with quite distinct notions of both structure and agency, part company and follow their separate projects, to analyse a

particular youth subculture or global capitalism. There is yet a further difference between the perspectives of the micro and

macro theorists, between the interactionists and those who emphasise structural determinants. At the extreme ends of the spectrum is an apparent absolute

dichotomy of mutually contradictory points of view. Structure is, or is not, a set of relationships between people! The anti-humanist theoretical end of the

spectrum insists that it surely is not: social structure is a set of relations between positions! People merely "inhabit these positions," or "functionally activate them"; the language used to describe where people fit into the schema of social structure is variable but the essential component of the anti-humanist

position is that we can and should conceptually ignore them ( Woodiwiss, 1990, 2001). For example, we should look at the structural relation between

capital and labour; and if we consider particular examples of employer and

employee we should not be concerned with the persons who perform such "roles" or fill such "positions." Both employer and employee as people are far more than their roles or positions. They are wives and fathers, generous or

stingy, basketball fans or friends of famous people. None of this is relevant to the employer/employee structural relationship.

This, of course, strikes "humanist" theorists as either theoretically disturbing or completely incomprehensible. Phrases such as "theory thinks us" seem to them an anti-humanist self-refuting reductio absurdum. The argument of this paper is that properly understood such phrases are certainly not absurd and in fact capture some truths about the nature of structural determination. And yet, so too does the perspective of the interactionist "humanist." The

argument of this paper is in fact that the "absolute contradiction" between the base assumptions of these positions is no contradiction at all. Structure is

simultaneously a set of relations between positions and relationships between

people. Further consideration of the example of the poker betting pace should elucidate how such is possible.

The poker betting pace (at any point in time, in any particular poker game) lacks the easily identifiable labels for the "positions" that structural theorists would assert that social structure is a set of relations between. There seems

only to be a set of interactions between people out of which the patterned nature of the betting pace emerges. However, the patterns having once

emerged are structurally "crystallised" and the relations between positions can be identified. They are not labelled of course. The betting pace structures are both too briefly enduring and multi-variable in kind and degree to warrant such

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

The Heuristic Example of Poker 189

abstract labels as for example capital and labour with respect to the economic structure. The betting positions in poker exist but always exist un-labelled. Or do they? Actually the positions are sometimes both recognised and labelled, albeit somewhat crudely.

Another poker anecdote should clarify this. Most stud poker games deal a mix of cards face up and cards face down. You see your opponents' cards but not all of them. In this hypothetical seven card stud situation after five cards have been dealt you see three spades face up in an opponent's hand. S/he may or may not have another two spades face down, which would make a flush - a very good hand. S/he is betting heavily: a signification meaning that s/he really does have this flush. However, this signification contains two other pos- sible meanings: either s/he does not have the flush yet (perhaps has one spade "in the hole" but not two) and is betting upon the probability of receiving another spade in the future two cards that will be dealt, or that s/he is purely attempting a bluff. One's understanding of this player as being either a cau- tious or wild better will affect your interpretation of which of these possibilities are in fact the case. So too does the betting pace. In a conservative betting pace your interpretation, your decision, is made for you. You have to fold. You make the decision but which decision you make is in fact structurally "determined." Their "position" was the "better."Your "position" is simply that of the "folder."

One of poker's features is that multiple potential meanings can be conveyed simultaneously. Meaning is conveyed but it is structurally embedded in the rules of the game that meaning contain some measure of ambiguity. You have (for a variety of reasons including an assessment of poker "character') concluded that possibility one (above) is not actually the case: whether bluffing or hoping for another spade in the next cards dealt, s/he does not have the flush now. That belief, combined with the fact that you have a pretty good hand, was what motivated you to have bet so heavily up to this point. However, the next round of cards dealt renders completely irrelevant most of your past specula- tions and tactical betting decisions. S/he is dealt another spade.

Now, virtually your only hope of winning is that she was totally bluffing before i.e. that she had no other spades face down. If s/he didn't have a flush before s/he very probably has one now. You should cut your losses and fold in the face of their large bet. Except ... except that it is only another twenty-five dollars and you have already put in over two hundred into this pot and you do still have some chance of winning. What do you do? Except in cases of conservative betting paces you call the bet. You say out loud "damm, I have to call it and I'm going to lose!" The players observing nod affirmation in response to both assertions. Your "position" is recognised.

Something like the above scenario occurs to players relatively frequently. "Ah, the old hundreds in and facing a near certain flush situation," a player

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

190 Canadian Journal of Sociology

might comment, "looks like you're [cooked]!" The "position" is even referred to in old homilies and adages of poker and life wisdom: "don't throw good money after bad" or "in for a penny in for a pound." We can remove your individuality from this "position." In some cases our individual identities as players are interchangeable. The situation we are in is most significantly composed of a structural relation between two positions in the game. This will be recognised by the other players. They will look at your "position" with sympathy or amusement because they understand the strategic/tactical dilemma of decision it provokes. The structural relation between the two

positions is the same for each and every possible player. We could remove you and I from our hypothetical scenario and substitute

any other two individual players and the structural relation remains unaltered ... or at least one of the structural relations remains unaltered. I put in this last

qualification because there are other structural relations involved in any situation and some of these do possess components of individuality. In order to theoretically fashion the rest of the "structural story" we have to reintroduce some of the complexity that we had previously bracketed off. This will, to a

symbolic interactionist, have a degree of irony. We wish at this point to reintroduce the "personhood" of the individual players, which we have thus far

only considered in the abstract as typifications of kinds of players. We shall do so, however, to make a powerful point about structures, even interactional micro structures such as this one: structures are always in a structural relation to other structures!

The players do not enter the game as blank slates waiting to have their

playing persona written for them as a response to the actions of the other

players. Rather they are human beings whose enduring "pre-game - outside of game" characteristics inform and causally determine their play. One could

put this in Bourdieu's language: they have socially "inherited" dispositions to play in a certain fashion. They respond to the particularity of the game's situations and consciously strategise in relation to this. But this strategisation is also a manifestation of their "habitus."

Players interpret and act in relation to the past history of the game; they adjust their play in relation to their memory of the past play of different individuals. But the individuality of players is also taken into account outside of poker's narrow context. Ironically, this involves powerful structural features of the way poker is conjoined with other structures. Let us take gender for

example. It is both a matter of individual difference and a structural "determinant" of particular situations. As stated at the outset, poker is not

being used here as an analogy. It is, however, very tempting because the game contains virtually all of the superstitions and prejudices found in life. The good player, of course, wisely attempts to bring none of these to the table. Nonetheless, poker is a male game. There are proportionally few female

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

The Heuristic Example of Poker 191

players. Frequently there is a male "machismo" atmosphere, which few women are likely to find appealing. There are possibly elements in gendered socialisation with respect to risk that also make the game less attractive to women. Some of these same elements perhaps affect the strategy of women beginners. They play timidly, conservatively. It is, by and large, not a winning strategy. So they don't win and they don't take up the game as a major pursuit. That much said, there are, of course, also some excellent women players. However, we are concerned here with probabilities and beliefs, whether true or not.

The notion of women players being timid betters is a popular prejudice. However, like most prejudices it has some grounding in some sort of reality. It is imported into poker from the larger social reality in which the player's habitus is formed. A good female player can capitalise upon this. Poker can be educational. The sexist player maintains his prejudices in the face of reality. He believes his female opponent will be afraid to call his large bet. She calls the bet and wins. She does so repeatedly. The sexist player after losing enough money has perhaps learned a lesson. For the good woman player the strategy of playing to stereotypes can be effective. For the female player feigning timidity, feigning deference to male power, feigning stupidity, can achieve the same successful results in poker as in other areas of life. Poker involves impression management and many women have become very adept at

managing the "little boy" within the man. Now on one level we could (and should) consider the interaction between

a man and woman as an interaction between two individuals. They each brought their individual characters and characteristics into the game, of which their gender was but one. They each consciously strategised. But each of them also had their strategisation unconsciously affected by past socialisation. They each interpreted the actions of the other informed by a set of beliefs about the particular situation, the game, about life, about the other, about the other's beliefs about them and about themselves. These beliefs in each case were rela-

tively true or false, and more or less consciously and unconsciously formed. The stereotypes about gender with respect to play are, of course, relatively

false and a good player will not operate upon the basis of them. So if we have a good male and female player one on one against each other in a hand, does that mean that gender is an irrelevant consideration? No, because gender as a factor transcends the individuality of the players. Gender may not be taken into account by any of the players involved; but this is, at least potentially, a costly error. There are statistical probabilities to gendered patterns of play. More importantly there are widely held beliefs about these patterns of play. The

game of impression management and the deciphering of the complexity of such impression management cannot avoid a consideration of these factors. The discursive structures of poker interact with the larger scale discursive

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

192 Canadian Journal of Sociology

structures of society. The woman player must come to conclusions about the

possible sexism of her male (and sometimes female) opponents in order to

interpret their actions and "deduce" their true beliefs. Will they stereotype her? Will they pretend to do so only to mislead her? Her opponents must engage in the converse of this process. And because this is so, gender remains a constant structural "determinant" (this term's potentially misleading nature will be later clarified) of the game.

Let us bring in another couple of factors: money and attitudes towards

money. Unlike some other games - chess, checkers, football - poker can never be divorced from the serious game of life. That is, the realities of money cannot be taken away from it and it still remain the same game. One can (and of course many do) play poker for pennies or matches. But this is not really playing poker at all. It has a resemblance to the real game but all the subtlety and complexity disappears. It is frequently the goal of good players to treat the amounts bet as simply counters, as merely "points" to be won or lost. But while a worthwhile goal to aspire to, the reality of life necessarily intrudes. Poker is a game involving risk, real risk. If one is playing for stakes so low that one literally does not care at all about the actual money won or lost, then this too becomes a factor in the game. It is only with the games of penny bets that this will be true for everybody.

Socio-economic inequality is directly imported into the game. Neither ten dollars nor a thousand will mean the same for everyone. Poker is a game in which rich and poor can certainly play together but their income disparities are not checked at the door with their coats. Wagering one's monthly rent on a

single turn of the cards is an intense experience. The same amount of money may be worth virtually nothing to a richer opponent. The advantage, as in life, lies with the rich. Far easier for them to merely treat the money as "points." However, lack of valuation can also engender carelessness. The rich player does not receive the same intensity of experience. Betting wildly is a way of

turning up the intensity, of fighting off the boredom (poker is also often a game of patience). But, betting wildly is (usually) also a strategy designed for losing.

One's betting behaviour and attitudes towards money are also "determined"

by ethnicity and class. Cultural socialisation with respect to money makes itself felt here. One is representing oneself to others through one's betting behaviour. And one is also representing oneself to oneself. In terms of poker strategy, attempting to represent oneself to others "truthfully" is a matter of variable tactical decision. However, considerations imported from the "game of life" make their presence felt here both consciously and unconsciously. Portraying oneself as risk-taker, as being unafraid and unwilling to "back down" before an opponent's big bet (macho challenge?) may place the cultural demands of ego maintenance at odds with strategic logic. Individually, conscious choice is real. But the effects of class and ethnicity produce patterns

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

The Heuristic Example of Poker 193

and probabilities with respect to socialisation at the same time as the un- conscious internalisation of common culture nonetheless still produces individuals.

The interaction of habitus and field (in this case poker is the field) displays a logic of "unconscious strategisation."2 One has, at any given moment in the

game, tactical decisions to make and virtually infinite number of factors to consider in order to help one arrive at these decisions. However, the interpreta- tions, the processing of this information and the decision taking, is performed through the lens of cultural conditioning and the formation of a particular sort of character (with respect to risk taking, impression management and other

factors). This is in part conscious and part unconscious. This is done in part purely with respect to the poker game's particularity and in part as an unconscious portion of the ongoing life project of ego negotiation and

personality formation. One may, for example, intentionally choose to lose a hand. This may some-

times be a wise tactic in order to set up the other players for a future coup de

grace. However, it may also (and simultaneously with the preceding) be an unconscious desire to lose. A Freudian could offer up a number of plausible (and possibly true) explanations for why a particular player made this parti- cular bet at this particular time in terms of the unconscious. Would such, if true, wholly cancel out the conscious logical calculations of the player? No, they operate simultaneously. The complexity of conscious representation, interpretation, calculation etc. is enormous. So too are the effects of a variety of unconscious determinants. Nonetheless, in every situation one is left with

only three basic choices in response to an opponent's bet: call, raise or fold. All the complexity of determination and choice is distilled into an instant of decision.

It should be clear that I am using the verb and noun, "determine" and "determinant," quite differently from their usually understood meaning. Determination is generally conceived of in terms of absolutes but we must abandon any such notion. Causality is complex and multi-levelled. There are not only multiple causes for any social phenomenon but multi-levels of causation. These levels are not simply analogous to Chinese boxes of

generality and proximity, macro and micro forces. There is that aspect to

causality but it is not the sole factor in play. For example, poker contains a

2. The analysis here not only is obviously derived from the work of Bourdieu but it is hoped is

developed in the same manner. Bourdieu's analysis of the interaction of habitus and field in

many significant respects can be summed up as the logic of"unconscious stategisation." This can be viewed as either oxymoron or brilliant insight. The former argument is propounded by Alexander (1995) and the latter by myself (Potter, 2000). Hopefully the poker example of this article further demonstrates that it is no oxymoron.

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

194 Canadian Journal of Sociology

general set of rules to do with the utilisation of money in the game but further variable specifications must be made for particular games or sets of games (e.g. the number of bets allowed per round, maximum and minimum bets etc. is very flexible). The structure of the betting pace will be (partially) deter- mined by these rules. We could establish a hierarchy of causality and determination here - the "Chinese boxes" referred to above. It would be a matter of levels. However, it is not at all accurate to see determinations as

working in this fashion to the exclusion of other factors and other sorts of levels.

The "Chinese box" analogy is accurate in one sense because each level of causation is contained within another of greater generality. But it is inaccurate because at each of these levels there are differences with respect to the determining interaction with factors outside of that particular "Chinese box." The "money" determinants of poker strategy and tactics which produce the effects of inequality of financial resources would fall within one set of "boxes," while the "attitudes toward money" (the causes and effects that I applied Bourdieu's notion of habitus to) determinants would fall within quite a different set of boxes and within quite a different hierarchy of causation. Yet all of this is in operation simultaneously.

Let us summarise the position arrived at. The betting pace(s) achieved in

any poker game is malleable and relatively fleeting but nonetheless is still a sui

generis social structure. It is both an emergent pattern of relationships between

people produced by them, reproduced and transformed by them, and also, a set of structural relations between positions. The structure of the betting pace stands in relation to the structure of the game of poker in general. It also stands, in a very broad sense and in a variety of different ways, to the structure of capitalism, patriarchy and other macro structures. Individual bets made and

patterns of betting stand in relation to relatively enduring poker versions of discursive structures. The betting pace stands in relation to the "personality structures" and habitus of the individual players. It stands in relation to a discursive structure of interpretations resultant from the past interactions of those players and more enduring meaning structures both quite specific to

poker and also to wider social intercourse. The meaning of a particular bet has a lot of ramifications and possible

interpretations on one level. That is, this is the game of educated guess, calculation and miscalculation, unconscious give-aways, truth telling, deliberate deception and prediction. However, on another level the meaning of a raise, call or fold is fixed. This level of meaning temporally transcends individual vagary, the particularity of the hand or the game and even

capitalism. Poker pre-existed capitalism. Capitalism as an economic system, as a cultural determinant in personality formulation and collective values produces its own effects upon the meaning of particular betting situations.

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

The Heuristic Example of Poker 195

Individuals playing poker in today's capitalist political economy quite likely play somewhat differently than did individuals in the Roman Empire. But on another level of meaning, a raise contains the same meaning in either historical era and in any poker situation whatsoever. The relationship between the individuals involved in the game, which as said above, produces, reproduces and transforms the betting pace, is not adequate to fully explain the phenome- non's sui generis nature. The relations between the structure and the other

"determining" structures are not explanatorily reducible to the interactive relationships between the players. The interactions of the players considered atomistically, and simply multiplied or added up, are not causally sufficient to

produce the phenomenon in question. Structural causality is involved as well.

The Conflation of Debate about Social Structure and Agency

The argument of this paper is that there has been a double error of contradic-

tory conflation in the debate upon social structure by macro and micro theorists, interactionist and structural theorists. On the one hand, interactionists see relatively enduring patterns of behaviour emerging out of individual interaction. This is implicit (and sometimes explicit) in the utilisation of terms such as "role." These interactions are interactions betweenpeople, people who utilise pre-existing structures of meaning in their interpretations, self indications etc., but people nonetheless. This view stands upon a number of obvious and true aspects of the issue. No people, no social structure! Regardless of all the sociological factors present that causally affect the situation, individuals make choices.

Notwithstanding the foundational truths of this position, structural theorists would assert that structure is far more than patterns of interaction. To reduce it to this is no concept of structure at all. Not only does it vain-gloriously enthrone individual agency and creative power of interpretation and choice but it obliterates the real sui generis nature of social structure. This seems apparent on a macro scale (pre-existence, externality etc.). To such theorists, six or seven people interacting in a room, does not a social structure make. To them the fundamental error of the interactionist is to take this woefully lacking notion of structure and simply multiply its complexity to apply it to macro structures. To repeat, it is the argument of this paper that both such positions are mistaken.

The structural theory position is mistaken because six or seven people interacting in a room may well result in the emergence of social structure. The interactionists are mistaken because this micro structure does not result simply from an interaction of individuals. In reality, there is a simultaneity of individual interactional and structural determinations of structure, even micro structure. The sui generis character of social structure emerges from both these

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

196 Canadian Journal of Sociology

sources ... simultaneously. This newly emergent phenomenon itself exercises

determining causal power. The individual stands in relation to this social structure as an insider and outsider simultaneously. Individual interpretations, decisions and actions reproduce and transform the structure. However, the individual also stands in a relationship of externality to this social structure. It forms determinate relations of both enablement and constraint.

Without the particularity at any time of this structure, of this betting pace, for example, many interpretations, choices etc. would not have been possible otherwise (enablement). Because of the particularity of this structure, of this

betting pace, certain interpretations, choices etc. are not possible (constraint). It has been asserted by many theorists that because of the "structuring"

aspect of social structure most physical metaphors of structure are misleading. To conceive of structure as a wall, for example, they would say is not right. And in a sense, and some of the time, in reference to some aspects of social

phenomenon they are correct. However, with respect to others they are wrong. With respect to social structure's externality, it is in some respects, precisely like a wall. One can choose between various doors but one cannot walk

through the wall.

Differences between Micro and Macro Structures

There is a final set of points which must now be made. My poker example has, I hope, demonstrated the existence of, and proper way of theoretically conceiving, micro structure. This hopefully clears up some of the misconcep- tions on both sides of a long history of debate. However, the historical polarity between the macro and micro theorist was not without good cause. As stated a number of times throughout this article, my poker example was not intended as an analogy illustrative of all the features of social structure in general. There are, in fact, real differences between such micro structures and other larger sorts of structures! Through my example of poker I have given an account of the features of micro social structure that shows it to have features previously associated only with macro structures - most importantly a sui generis character. I stand by this argument.

Nonetheless, I do not feel the phenomena of macro social structures can be theorised merely through a conceptual multiplication of my account of micro structure. I do not wish to replace the individual atoms of individual interaction with a different but still atomistic account, merely providing a structural account of the social molecules which emerge and then reducing complex life forms to collections of molecules. Rather, we must take into account

qualitative changes in scale and complexity. Reality is a quantitative continuum. The lines and boundaries of reality are

determined relative to human interest. There is much "division" of dramatic

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

The Heuristic Example of Poker 197

scale in reality's continuum that is of no significance to us. There is also a

great deal of such "division" necessarily of significance to us. There are facts

concerning the human condition which determine much of this. For example, if we were a colour blind species it is likely we would have no categories of red and blue. Light's spectrum would still be there to be measured if we were so inclined but the qualitative distinctions of colour we break it into would not

necessarily be organised in the present fashion. The argument here is

essentially that there are reasons for qualitative categories relevant to facts about "our nature," about the human condition.

How can the above argument be applied to the issues surrounding social structure? My argument is that there are structures and structures, which are

qualitatively different in kind. This qualitative difference is also a quantitative phenomenon of size, complexity, malleability and longevity. What are the "facts" about the human condition which determine the necessity for a

qualitative distinction between micro amd macro social structure to be made? There are two such facts. One is the life span of human beings. The other is the set of causal powers possessed by humans and the limits of such powers.

Let us consider the temporal factor first. Social structures (and we are here in general now speaking of macro social structures) are reproduced inter-

generationally. The externality of macro social structures (a feature which I have argued is an aspect of micro social structures as well) impacts us as individuals much more obviously and dramatically on the macro level. It was not for nothing that Durkheim chose language and systems of currency as

examples of social structure to illustrate its sui generis nature. The observa- tions Durkheim made considering the pre-existence to any individual of such structures is powerful for its very obviousness. Pre-existence strongly reinforces an argument concerning a kind of independence of existence.

Earlier, I asserted that it is often the case that there are no ready to hand labels for a variety of micro-structural positions. The lack of such points us toward another qualitative difference between macro and micro structures.

Length of persistence of existence over time indicates a more solid and noticeable "crystallisation" of structural positions and the relations between them. This also usually makes them causally more efficacious. Social structures such as the betting pace in poker are causally efficacious whether the

players are aware of its existence or not. Nevertheless, it is but one causal force

among many and its power, relatively speaking, is not great. As a structure it is relatively "fragile" and easily subject to change.

The last point brings us to the issue of individual agency. Again, we can say it is not for nothing that the issue of human agency (most importantly individual human agency) has been historically imbricated with debates about the nature of social structure. It is only unfortunate that such debate has so

frequently been couched in terms of absolutes ("free will" versus "determin-

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

198 Canadian Journal of Sociology

ism" to cite the worst example). The second "fact" of the human condition relevant to the distinction between macro and micro structures is the limited nature of human causal powers. All human beings are not created equal; and

they are certainly not born into an equal society. They are imbued with vast differences in terms of degree of social power. However, regardless of differences, all are limited in terms of personal power. This simple fact

possesses enormous significance with respect to the differences "in kind" between macro and micro structures.

I argued that the micro structure of the betting pace causally affects the

players. However, it is also quite susceptible to causal influence, even conscious manipulation, by the individual. It is so small a structure, there are so few agents reproducing it, that it is extremely responsive to their action. All are not equal in this. A relatively bad, relatively poor (in financial terms) player, is perhaps severely circumscribed in his or her ability to affect a change in the betting pace. In the first place s/he is probably not aware of its existence. In the second place, while consistently bad interpretations and decisions will do as much over time to affect the betting pace as good ones, such actions will

likely result in the bad players retiring broke before much structural effect has

emerged. But let us look at a rich "bad player." Repeated heavy betting in the face of a conservative betting pace will surely change it. Conscious effort by good players is sometimes made to change the betting pace - to "loosen up the game" - but in general this is not their intention. Rather, it is generally to

capitalise upon the manner in which it determines the play of others. Both the

good players and the "rich" players possess much causal power to effect

change in the betting pace. To make our point about scale and human powers, let us juxtapose the

previous example with a rich person consciously attempting to change a macro social structure. Let us imagine a lunatic media tycoon (Ted Turner perhaps?). On a bizarre whim he decides to attempt to change the English language. Languages do change over time; their deep structural features change in relation to actual usage. Not all are equal in terms of their individual power over such change and further it takes place almost entirely on an unconscious and supra-individual level. With respect to this, we can say that virtually all individuals are for all intents and purposes nearly impotent. But this is not the case for the media tycoon. He possesses power (potentially even to enact a whim) over many, many others (his broadcast employees) speech, and thus over what millions will repeatedly hear. But if his language edicts for his broadcasters are too bizarre the viewers will simply change channels; and, in

theory at least, even a billionaire can go broke. We can cite here as a real

example the French government's futile efforts to control what they perceive as "language pollution." Thus, in terms of the temporal scale of human

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

The Heuristic Example of Poker 199

lifetimes and the causal powers of individual human agents, we can say that macro social structures are relatively stable and relatively impervious to

agency. This is not so for micro structures; thus, we can conclude they are importantly different in kind.

However, having reached this conclusion we must now ask a philosophical question. Do we understand this conclusion in realist or nominalist terms? In asking this question we actually find that it directs us to a specific empirical question.

Other Kinds of Structures? The Issue of Critical Mass

The preceding section presented an argument that micro and macro structures were importantly different in "kind." Earlier the argument was made through the example of poker that this difference was not that of the sui generis aspect of structure. Even micro structures may possess an emergent sui generis character. The argument thus profoundly diverges from the traditional fault lines of sociological debate. However, in another sense the paper has followed the traditional fault line; and in doing so has produced another important theoretical issue to be researched: the question of "critical mass."

This paper, like much of the theoretical debates in their reference to

examples of micro and macro structure respectively, has focussed upon the

very small and the very large. What of structures of an intermediate size? If one admits of qualitative difference (as understood in the terms of the

preceding section) between micro and macro structures an empirical question arises: are there (in some sense quantitatively with respect to size) two and

only two kinds of structures, to wit, micro and macro? Or are there more than two kinds? In this case there would be the very small micro structures such as poker's betting pace, the macro structures of world economy and so on, and a continuum of different sized structures admitting of malleability and suscepti- bility to individual agency to various degrees.

Let us re-phrase the above questions and briefly move from sociology to philosophy. Do micro and macro refer to "real" referents or are they merely convenient and useful categorical labels? With respect to such questions in general, however, there is an asymmetry between nominalism and realism. For the nominalist all concepts of this sort can be understood as labels. The converse is not true for the realist. Let us take for example a Marxist realist and the concept of class. For this sort of Marxist (there are others for whom this would not be true) class exists. It exists independently of its identification, labelling and conceptualising by the theorist. It is not a mere categorisation, which can be defined however one pleases, but is an existent social fact. Class struggle to this sort of Marxist is an ongoing material reality regardless of

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

200 Canadian Journal of Sociology

human awareness and understanding of what is going on in these terms or not. Obviously a nominalist would not accept this.

However, there are other concepts upon which the realist and nominalist would be agreed. Not every concept has the same ontological status to the realist. We like to "cut reality at its joints" but we do not do so every time. Sometimes our categorisations have other purposes. Sometimes the category is employed before the ontological status can be known. Such is the case for the concepts of micro and macro social structure. In the extreme cases of very small and very large scale social structures we see a difference of degree of

susceptibility to individual agency sufficient to warrant an ascription of

qualitative difference. The very small is very susceptible to such agency whereas the very large is near totally impervious to it. But what happens to the correlation between size and agency (other factors being temporarily bracketed off) in "middle sized' structures? Is there a sudden break whereby the degree of agency sharply diminishes towards the vanishing point? Can we, now allowing in some of the features previously bracketed off, ascertain how size correlates with these features in relation to agency? These are questions susceptible to definitive answer through empirical research.

Summary

I have argued that micro social structures possess some of the sui generis features usually only associated with macro structures but not others - size does matter. These similarities and differences between macro and micro have

generally resulted in opposite sorts of conceptual errors by micro and macro theorists, which have further served to occlude the real insights achieved by both camps. Social structures are not reducible to either relations between

positions or to relations between self reflexive interactive human agents. Both factors are in play simultaneously. Different levels of analysis are called for to

correspond to the ontological depth of reality. However, a synthesis of these levels is also required to grasp the phenomenon of social structure in its full

complexity. Finally, though micro social structures do in fact possess many of the features previously only associated with macro social structures there are still important differences between them. Two contrasting implications of this article apply to human agency in relation to social structure. Micro social structures are perhaps less malleable in this respect than is frequently believed

by interactionists. Macro structures, at least potentially, if human agents consciously collectively organise to attempt to affect change, are perhaps more malleable than is usually conceived by structuralists.

The game of poker is not equivalent to the "game of life" but nonetheless it can teach us many important things about it.

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

The Heuristic Example of Poker 201

References

Alexander, Jeffrey 1988 Action and its Environments, New York: Columbia University Press. 1995 Fin de Siecle Social Theory: Relativism, Reduction and the Problem of Reason, London

and New York: Verso

Althusser, Louis 1971 Lenin and Philosophy and Other Essays, London: New Left Books.

Archer, Margaret 1995 RealistSocial Theory: The MorphogeneticApproach, Cambridge: Cambridge University

Press.

Bourdieu, Pierre 1977 Outline of a Theory of Practice, Cambridge: Cambridge University Press. 1990 The Logic of Practice, Cambridge: Polity.

Bhaskar, Roy 1979 The Possibility of Naturalism, Brighton: Harvester.

Blumer, Herbert 1969 Symbolic Interactionism, Englewood Cliffs, New Jersey: Prentice Hall.

Bryant, Christopher, A.G. and Jary, David (eds.) 1991 Giddens' Theory of Structuration: A Critical Appreciation, London: Routledge

Callinicos, Alex 1999 "Anthony Giddens or Pierre Bourdieu," New Left Review, 236: 77-102.

Collins, Randall 1981 "On the Microfoundations of Macrosociology," American Sociological Review, 86:

925-942.

Douglas, Jack (ed.) 1971 Understanding Everyday Life, London: Routledge and Kegan Paul.

Durkheim, Emile 1982 The Rules of the Sociological Method, London: Macmillan 1895.

Foucault, Michele 1971 The Archaeology of Knowledge, New York: Pantheon.

Glucksman, Miriam 1974 StructuralistAnalysis in Contemporary Social Thought, London: Routledge and Kegan

Paul.

Goffman, Erving 1959 The Presentation of Self in Everyday Life, Harmondsworth: Penguin.

Layder, Derek 1985 "Power, Structure and Agency," Journal for the Theory of Social Behaviour 15:

131-149.

Lopez, Jose and Scott, John 2000 Social Structure, Buckingham: Open University Press.

Parsons, Talcot 1937 The Structure of Social Action, New York: McGraw-Hill. 1951 The Social System, New York: The Free Press.

Potter, Garry 1999 The Bet: Truth in Science, Literature and Everyday Knowledges, Aldershot: Ashgate.

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions

202 Canadian Journal of Sociology

2000 "For Bourdieu, Against Alexander: Reality and Reduction," Journal for the Theory of Social Behaviour, 30:2.

Schutz, Alfred and Luckman, Thomas 1973 Structures of the Life-world, Evanston: Northwestern University Press.

Woodiwiss, Anthony 1990 Social Theory After Postmodernism, London: Pluto Press. 2001 The Visual in Social Theory, London: Athlone Press.

This content downloaded from 62.122.77.83 on Sun, 15 Jun 2014 00:31:21 AMAll use subject to JSTOR Terms and Conditions