THE MEXICAN CANTINA AS A SETTING FOR INTERACTION' Brian Stross University of California, Berkeley INTRODUCTION Participant observation in a specialized interaction setting, a Mexican urban cantina,2 led to an inquiry into the question of deter- minacy and indeterminacy in interaction within a given social situation. Interaction implies communication, and communication implies the sending and receiving of information by means of some mutually intelligible system of codes. Communications engineers spend a lot of time coping with problems of misinterpretation, and social theorists have based the concept of culture on the assumption that meanings are shared by parti- cipants in the culture. To this extent the tendency has been to conceive of communication as a determinate system, or nearly so, in which people base their decisions and relationships upon the accurate determination of the intended meanings of communicative acts. Any ambiguities or misinterpretations are further assumed to be a fault of the code system or of the participants rather than a necessary feature of social inter- action. I am proposing to call the misinformation, misinterpretation, and withholdingof information that seems to be inherent in all communi- cation systems, the indeterminate aspect of interaction, and to view it as a necessary feature which functions to maintain the order that exists in a cultural system. When seen as a necessity rather than as a fault, the indeterminacy intrinsic to communication in interaction can be legitimately studied from the standpoint of specific functions and methods of maintenance. The opposing tendencies of determinacy and indeterminacy of meaning in communication can be found in varying degrees in different social situations. For example, a man riding the train may not care 58 how much private and personal information he reveals to a stranger so long as he knows that he will never see the stranger again. In the classroom, the same man in the position of teacher must carefully re- strict such information in order to avoid the conflict of roles. Crystallized in folklore we find cautions regarding information control such as "familiarity breeds contempt," but this appears to be contra- dicted by "to know someone is to love him." In any situation the diffi- culties of knowing someone, but not too well, are managed in part through careful manipulation of the determinate aspects of codes. This manipulation depends upon a knowledge of the norms for proper behavior that are specific to particular situations. The Mexican cantina is a situation that is well suited to the study of interaction norms and the manipulation of the determinate and indeterminate aspects of communication, because one of the primary reasons that a person enters the cantina is to interact with others in ways that would be socially unacceptable within other settings. Further- more, there are definable rules and a number of interaction problems to be dealt with by various classes of participants. Knowledge of the structure of the situation and its relation to Mexican culture are pre- requisites to an understanding of these rules and problems. Therefore the purposes of this paper are twofold: to present a coherent structural picture of the cantina as a social setting, and to bring out some facets of social interaction that have been hitherto ignored in- the literature. Goffman (1964) has suggested that the situation as a subject for study has been neglected in social science research. He maintains that situations warrant analysis in their own right, like that accorded other fundamental forms of social organization. I shall treat the cantina as a social situation and provide an ethnographic descriptive analysis of it which is as complete, yet as concise as possible. The problem is to find and employ an adequate framework for the dissection of the structure of the cantina situation while focusing on the process of social interaction. To this end I have separated the structure into 59 three aspects: 1) the concrete, 2) the abstract, and 3) the processual. Each of these aspects can be divided into two parts: 1) paraphernalia and participants, 2) roles and regions, and 3) behavior cycles and rules. The first two aspects are conventional enough that their inclusion in a structural description needs not be argued (cf. Frake 1964). The rationale for the use of the processual aspect is as follows. The study of structure is the study of ordered relationships among units in a given system. Because social interaction implies change, the structure of a social situation cannot be described in interactional terms without reference to the patterned changing of relations between social units. The social units involved are social identities, herein- after called roles. These roles are enacted relative to other roles through cycles of action and in accordance with rules or norms for action. Consequently the aspect of structure-process will be presented in these terms. An ethnographic description of the structure of a social occasion3 is incomplete unless it is fitted into the context of the culture in which it occurs. For this reason it will be necessary to relate the cantina situation to Mexican culture as a whole. There is difficulty, however, in deciding what of Mexican culture is relevant enough to necessitate inclusion without writing an ethnography of Mexico. I have met this problem by focusing on the ways that the cantina functions for the participants. The order of presentation will be the following: first, the formal structure of the cantina situation described in terms of the three above mentioned aspects; second, the relationship between the situation and the broader cultural context; third, a discussion of the results as they relate to determinacy and indeterminacy in communication. 4 60 STRUCTURE: CONCRETE Paraphernalia: Objects and Spacing Salient features to be found in every cantina are a bar, tables and chairs or benches, extra chairs, some barrier to sight at the entrance, a urinal, a door to the toilet, and a jukebox. Minor variations among cantinas in the spacing of physical objects force me to present one "typical" cantina in order to maintain simplicity of expositi6n. The spacing of objects in the cantina can best be appreciated from the point of view of the entering customer, who will notice people as well as objects. At the doorway to the noisy, smoke-filled cantina the prospective customer can part the curtains covering the entrance or peer over a five to six foot plaster wall between the door and dance floor in order to decide whether or not to enter. This is permitted, but few remain for long in this transitional zone, forthese partial barriers to perception symbolize the fact that the situation is a closed system. That is, it is private to the participants rather than a public performance. One generally feels more comfortable either in or out of a situation, not halfway between. Once inside the cantina, the customer usually stands watching the dance floor, often surveying the other patrons as well. From this position he can get a good view of the layout of the place. This is what he sees, to take a characteristic moment: Straight ahead lies the almost empty dance floor covering some two-hundred square feet. It is almost empty because the band has just taken a twenty-minute break and one of the women present is putting a "toston" (fifty centavo piece) into the jukebox while another is making a selection. The dance floor has only one couple on it and they are waiting for the music to start so they can dance together again. Almost surrounding the dance floor are chairs on which women are sitting facing the dance floor. Surrounding these chairs are about twenty tables, each with two to four chairs, most of which are occupied 61 by the male patrons. A few of these patrons are talking to women to whom they have offered a drink. To the left is a bar and a short flight of steps leading to a balcony on which the band members have left their instruments while getting drinks at the bar. To the right is a wall, half covered by the tables against it. The other half has an eight foot long urinal set into it and a door leading to a small room with a toilet. The jukebox is facing the dance floor from the wall that is straight ahead. The whole room, with its impersonal atmosphere of intimacy, is about nine hundred square feet, and has had its ceiling lowered by the use of cloth, accentuating the intimate feeling of the setting. Figure 1 portrays the setting graphically. Participants and their Attributes The participants recognize and comprise four essential classes based upon functional roles in the situation: patrons, waiters, musicians, and women. Most urban Mexicans have heard about cantinas and have at least a vague notion of what goes on in them. It is also common knowledge that "decent women do not enter the cantina," although nobody seems to recall where he first learned this fact. In addition it is well known that "well brought up young men should not patronize cantinas but always do anyway." These two value statements show clearly some of the forces that impede or encourage the entry of various people into the situation. The first statement and the social force behind it explains why women are never patrons of a cantina. The second statement appears to imply that the cantina experience should not be missed by any male if he is to keep up with his fellows. In fact, among many segments of the urban population, the visit to the cantina is one of the rites of passage denoting a venture into manhood. There are, then, forces acting to draw in or keep out various classes of participants. These forces also affect the accompanying paraphernalia brought into or left out of the situation (e.g., periodic police searches for weapons keep certain types of persons out and also 62 tend to ensure that people leave guns or knives at home). It is the effectiveness and selectiveness of these forces that limit the number of social roles in the situation and allow this attempt to describe categorically the participants involved. Patrons are always men, spanning approximate age limits of eighteen and fifty, the majority between twenty and thirty-five. The older the man, the more likely it is that he comes from the lpwer income class. There is *no real basis for grouping patrons by personality traits, but a small sample of informants distinguished three types: the peasant who is often socially inept and somewhat introverted; the macho (being a derivative of machismo, a complex Mexican phenomenon involving males' assertion of masculinity and dominance in numerous ways, and applied to anyone whose machismo is evident) who is overly aggressive; the ordinary person, a city dweller with no particular distinguishing characteristics. It will be shown later that there are a variety of factors accounting for the entry of the patron into the cantina. Suffice it as this point to identify the patron with one or more of three verbally expressed intentions; drinking, intimate interaction with cantina women, and male companionship. The first two are for sale in the situation. The third follows from the first. Most of the patrons drink beer, some drink hard liquor such as rum or tequila, and occasionally someone will order a soft drink such as Seven-Up or Coca-Cola. Everyone drinks something. About half of the patrons will sooner or later dance with a woman and perhaps ten percent will eventually treat a woman to beer at the table. There are between twenty and fifty patrons in the situation at any given time and the rate of flow at the door is about one per minute. Waiters are men between the ages of twenty-five and forty-five. There are from one to three waiters on duty during a single shift, and they are almost constantly moving about. After having visited a large number of cantinas, I was struck with the impression that the waiters were all alike in personality and mannerisms. Closer observation 63 revealed little to change this impression, so I tried to formulate an explanation for the fact that they all appeared to move fast, to be very jolly and possessed of a fine sense of humor, to laugh a lot, and to generally play the role of clown. The explanation to be given here is that the role of waiter in this situation has certain necessary attributes that are independent of the individual. Anyone who will fill the position of waiter successfully must embrace this role and display the characteristics that are so spe- cifically advantageous to maintaining a harmonious situation. Under normal circumstances the waiters have the most influence over and respon- sibility for the behavior patterns in the situation (cf. Barker 1961:464). Their job is to please their customers but they must also deal with potential disruptions to the situation, whether in the form of a customer threatening to start a fight with someone, or in the form of a policeman invading the cantina and making everyone nervous. The consistently jolly attitude allows the waiter to leave an engagement with a drunk who wants to talk, to deal with the macho with less chance of provoking a fight, or to jestingly tease the women. The playful, never serious, jesting waiter is not likely to be seen as a threat to the manhood or dignity of drunk or half-drunk patrons. He is a contributor to the informality of the situation, and his playful mock-fights with the women provide enter- tainment and satisfaction for the spectators. In short, the jolly waiter has a built-in mechanism for the maintenance of situation-specific social distance (cf. Murphy 1964:1273) for the purpose of maintaining order and satisfaction in the situation. He has built up a potent image that allows him to break rules of social order without giving offense.S The waiter's more explicit job is to serve alcohol and collect money. Associated with this is his job of unobtrusively distributing chits to the women at appropriate times (a chit being a token representing designated amounts of money that can be claimed by the woman when she goes home). 64 Cantina women are between the ages of sixteen and about fifty, with the largest representation between twenty and thirty-five (eighteen is the legal age minimum). The typical cantina harbors about fifteen women. They are not hired by the management but they do receive percentage kickbacks on drinks that they can persuade customers to buy for them and on exit fees paid by the patrons to take the women out of the cantina for a while. These cantina women are polar opposites of "decent" Mexican women. The ideal of the marriageable (decent) woman holds that she be chaste, faithful, careful of her reputation, ignorant about sex, euphe- mistic in language, religious, Cnd pure until marriage. She has verguenza (shame), the female complement to 1-ismo. The cantina woman by contrast usually has one or more illegitimate c. ldren to support, smokes, drinks, curses, sells her body, stares invitingly at a man without feeling shame, fights in jest or in earnest, and often playfully grabs at the private parts of her female co-workers. On the whole the cantina women submerge their individual iden- tities in order to convey a totally impersonal impression. This is not to say that they succeed perfectly in doing so. They each have their own idiosyncracies and when interacting together they exhibit a definite small group structure, pecking order and all. Nevertheless, as a rule the cantina situation is purely business for them and the very intimacies that their behavior exhibits are part of a role constancy that maintains social distance by protecting their individuality. Their behavior is intimate and inviting to the patrons, but they all display an inviting attitude, and to everyone that can pay. They therefore are not acting as individuals and do not consider the patrons as individuals. It can be said then, that, in protecting their individuality from the patrons, they are injecting social distance between themselves and the customers. At the same time, however, these "women without shame" use their shame- lessness and approachability to facilitate communication with the patrons for the purpose of making money. This leads to an apparent paradox in 65 the juxtaposition of determinacy and indeterminacy, for they communicate both unequivocally and ambiguously in their role behavior. This paradox will be discussed more fully below. Musicians are males between the ages of about twenty and fifty. In contrast to women, their age, looks and personality have little to do with their statuses as musicians, although these factors do become im- portant in the context of the small group of musicians, possessing among themselves a "back region" where behavior occurs which is concealed from others. 6 The musicians are paid by the management to play music for dancing. The typical cantina has five musicians. Their instruments are: trumpet (one or two), bass, electric guitar, marimba or saxaphone, and drums. Two to six minute selections (danzones, pasodobles, sambas, etc.) are played. When the musicians take breaks the jukebox provides the music. There are other people present in the situation at various times and for varying lengths of time. They are all undesirable within the situation, and contribute nothing important to its structure. These people are beggars, street vendors, policemen, and the homeless. STRUCTURE: ABSTRACT Roles and Regions In describing the participant aspect of the concrete structure of the situation, I have found it convenient to separate four functional classes of individuals (waiters, patrons, women, and musicians) and to describe some of their attributes. These attributes pertain, on the concrete level, to individuals rather than to statuses. Now these four classes of individuals can be placed on another analytical level where each class label will designate a status. It is the relationships be- tween statuses that most uniquely define the cantina as a "situation." These relationships comprise the "role structure." Figure 2 illustrates this structure. 66 The exchanges that are part of the role relationships involve money in all four cases. The waiter occupies the central position in the diagram, and is involved in the most exchanges. This fits well with the above mentioned hierarchy of situation influence and responsibility. From Figure 2 it is apparent that there are four positions relevant to the situation. Each position implies a different number of roles within the respective role sets, for a role implies a relationship with another role. The waiter has three roles, the patron and woman each have two roles, and the musician has one role. It is obvious, however, that this configuration does not account. for all of the interactions. One could observe that there are patron-patron relationships, waiter-waiter rela- tionships, and musician-musician relationships. These are, in fact, not structurally necessary on the level of the situation. They exist on an analytical level that could be called the sub-situation. This distinction between the situation and the sub-situation follows from the fact that the last mentioned relationships are not necessary to the definition of the cantina situation. If there were no interaction, no relationships between musicians, for example, this would make little difference to the situation. On the other hand there could be no cantina in which the waiter did not interact with the patron. Some relationships in the cantina are therefore necessary and some are not. The "role structure of the situation" includes only those role relationships that are vital to the functioning and maintenance of the situation. There are four "regions" associated with the four statuses. The musician's region is on the balcony and is clearly bounded. The woman's region is in and around the chairs that line the dance floor, and is less clearly bounded. The patron's region lies in the area of tables and on the seats at the bar. The waiter's region is in back of the bar and in the aisles separating the other regions. There are also two areas that could be called neutral regions: the dance floor, and the toilet-urinal area. The different participants have varying claims to these neutral regions. For example, all participants make use of the 67 toilet-urinal area, whereas patrons and women have the major claim to the dance floor. The aisles, which are primarily for the waiters, are used in part as a neutral region by all participants. There are definite regions in which the structurally significant role relationships occur. The waiter-patron transaction o-ccurs on the patron's territory. The waiter-woman transaction takes place in the patron's territory, the waiter's region, or the woman's region (rare). The waiter-musician transaction occurs in the waiter's region. The patron- woman relationship takes place on the dance floor, a neutral region, and in the patron's region. To summarize, the cantina comprises a setting of paraphernalia and classes of participants selected by the situation for their attributes. The classes of individuals can be called statuses which occupy regions and which have situationally functional role relationships with other statuses. Because the situation takes place in time there is another as- pect of its structure that must be considered, the structure of process. This will be discussed below under the headings of behavior cycles and rules. STRUCTURE: PROCESS Behavior Cycles The patron comes to the cantina to spend his money. The others are there to make money. Therefore the behavior of the patron is less restricted by the situation in the sense that he has a wider variety of choices for action open to him than do the others and that he can termi- nate his cycle in the cantina at almost any time. In the depiction of the action cycle of patron (Figure 3) the basic cycle will be that which most fully exploits the facilities of the situation, ending when the patron leaves the cantina with a woman, despite the fact that fewer than ten percent actually leave in this manner. 68 The patron, if he wishes, can complete a single cycle of be- havior, beginning with his entrance into the cantina and ending with his exit from it, without repeating any aspect of the cycle. It is this cycle that will be described, and any alternative choices that would lead to repetitions will be mentioned but not followed up. As can be seen from Figure 3, when the patron enters the cantina he immediately has three choices: leaving, standing and watching, or going to a table and sitting down. If he sits down he must order and re- ceive a drink. At this point he can either sit and drink or go to an unoccupied woman and ask her to dance with him. If he dances with a woman, there are two choices at the end of the dance: that of going back to his table alone, and that of bringing the dance partner to his table and standing her to a drink. When the woman's drink is finished, provided he has taken the latter alternative, he can decide to send her away or he can leave the cantina with her, thus completing the cycle. At any point during the cycle when he is not otherwise engaged, the patron may use the toilet, interact with another patron, or repeat some aspect of his cycle. The waiter is more limited in his choices for proper situation behavior. When he enters the cantina his first duty is to go to the table of a patron who wants service and to collect the money for the patron's order. He then returns to the bar and fills the order. Next he brings the drink(s) back to the table, at which time he gives any woman who has ordered a drink from that table a chit after having made change for the patron. This action cycle is repeated until the waiter goes off duty. At any convenient time between or during cycles the waiter may interact with patrons, women, or musicians, or collect an exit fee from a woman if she is leaving with a patron. At some indeterminate time, when the musicians go off duty, the waiter pays them. The woman enters the situation and sits down in her region until a patron asks her to dance. After dancing she either goes back to her region and sits down or goes with the patron to his table. If she does 69 the latter, when she finishes her drink, she either returns to an earlier portion of her cycle or leaves the cantina with the patron. At any con- venient time during the woman's cycle she may interact with someone in her region. The musician comes on duty and goes up to the balcony where he plays music for three to six minutes. He then rests for a minute or so and resumes playing. This goes on for up to an hour. After that the musician takes a break for less than an hour, during which time he per- forms no role at the situation level. At the end of the musician's work stretch he gets paid and leaves. It should be understood that the real life situation is not as neat as that just depicted. One important reason for this is that the cycles of behavior relevant to the situation structure are overlaid by much behavior that is not relevant at this level. In the description of the process rules some of this overlaid behavior will be taken into account. The description of these rules calls for another analytical distinction between rules created specifically for the situation, and general cultural rules modified by the situation. Rules created for the situation are those rules which are ex- plicit for the cantina setting and which a newcomer cannot be expected to know. If necessary, these rules will be explained to the patron. Rules modified by the situation are those rules or norms of conduct that come about as a result of the cantina's place in the culture as a whole. These rules are implicit in the cantina and will not be explained to the newcomer. Rules created for the situation must be followed. Rules mo- dified by the situation generally are followed if the patron is to fit in, but these involve ranges of appropriate behavior so that deviations from the norms involve sanctions that depend upon amount of deviation. All rules will be described here largely from the patron's standpoint in order to minimize redundancy. 70 Rules: Created 1. When a patron enters the situation and sits down he is obliged to buy a drink. He may either sit and wait until a waiter spots him, or he may bang the table once or twice with his hand or with some object on the table (such as an empty beer bottle). Attention can be attracted by merely hissing in the direction of the intended receiver, but this signal is more general, can apply to anyone, and is simply a sign of the desire to attract someone's attention. Banging the table is a specific signal denoting the intention of ordering a drink. 2. When the waiter arrives to take the order, the patron must pay for what he has ordered. Bottled beer is one and a half to three pesos depending upon the brand. Corona, Carta Blanca, XX, and Superior are the most frequently ordered. Canned beer (Tecata) is served with lime and salt and costs three pesos. Payment for beverages must be made in cash and the waiter will bring any change due when he de- livers the order. Money transactions between patron and waiter usually take place on a waiter's tray or on the table. 3. The patron must take the initiative in an encounter with a woman. He may ask any unoccupied woman if she will dance with him, or he may signal his intent through some other channel such as gesture. The woman may or may not re- fuse him. She usually accepts. 4. After each dance with a woman the patron must hand her the fifty centavo charge or more. The woman will make change unless expressly instructed by the patron to keep the change. She puts her fee in a small coin purse that is often held in the hand or secreted in her clothing. 5. If the patron wishes a woman to accompany him to his table he must offer her a drink. This is usually done during or after a dance. Most women will order a Tecate if the patron allows her to, for which the patron must pay the waiter eight pesos. For this the woman receives a chit worth two pesos. The woman will only remain at the table so long as there is beer left in her glass. 6. If the patron wishes a woman to leave the cantina with him he must pay an exit fee for the woman. This fee is about twenty pesos which the woman gives to the waiter on the way out, receiving in exchange a chit worth from five to ten pesos. 71 Rules: Modified The cantina is a situation in which many of the rules modified by the situation can be most clearly seen as relaxations of rules of proper conduct that apply among people outside of that situation. The consumption of alcohol in large quantities provides a partial excuse for much of this relaxation of conventional rules of order. The fact that cantina women are not considered ladies also helps account for the norms of behavior there. Language behavior for all patrons generally approximates the language of male intimates in the urban working class population. The participant peasant from the country is often made fun of in the cantina because of his linguistic ineptness and his sometimes excessive formality. On one occasion a peasant was the butt of many jokes made by the women because he failed to remove his straw hat. Speech in the cantina is loud, syntactically simple, and some- times difficult to hear because of constant background noise. There tends to be less use of diminutives which on the outside serve the functions of signalling status differences and acting an euphemistic re- placements for words of harsh sound or meaning. Euphemistic words and circumlocutions are less often employed in cantina discussion of topics in which they are usually used, such as matters relating to death, sex, and body functions. This less frequent use of euphemisms parallels a tendency to employ tabooed vulgar expressions. Muscular control is far less important in the cantina than elsewhere. A general looseness of the limbs is permitted and staggering is common. Aim at the urinal is careless, and knocking over a bottle or glass on the table is not cause for embarrassment as it might be in a restaurant. In much of Mexico there tends to be a high tolerance for phy- sical contact in interaction. This tendency is increased in the cantina situation. Male gestures of affection such as the embrace or the grasping of arms or legs are common, even among chance acquaintances. 72 The intimacy displayed by physical contact extends to the verbal aspect of interaction as well. Paz notes of Mexican men that relationships with other men are always tinged with suspicion. Every time a Mexican confides in a friend or acquaintance, every time he opens himself up, it is an abdication. He dreads that the person in whom he has confided will scorn him. There- fore confidences result in dishonor, and they are as dangerous for the person to whom they are made as they are for the person who makes them (1961:30). This does not apply with much force within the cantina, where it is much less dangerous than elsewhere to confide in others. The patron may and often does talk to waiters, women, or other patrons about very personal subjects and feelings, often becoming more intensely involved in an encounter than would be proper elsewhere. It would be fair to generalize and state that in the cantina self-involvement as well as involvement in a face-to-face encounter is permissible to both a greater and to a lesser intensity than would be the case elsewhere. Demonstrating lack of interest is an appropriate and useful means of avoiding or terminating undesirable encounters. Conversely, intense involvement can convey social intimacy in a way that would not be permitted in another situation. Alcohol can excuse much behavior which, but for the situation, would be regarded as impolite (cf. Cavan 1966). The women in the cantina command none of the respect that would be shown a "proper" woman. They may be teased, pointed at, and asked intimate questions in language generally reserved for men. Or they may be completely ignored. There is rarely any of the gallantry that is often exhibited toward women on the outside. Within the cantina the initiation of encounters does not re- quire an acknowledgment or excuse. The termination of encounters can be as informal as their initiation, and during an engagement the audience is usually of little importance. Strikingly characteristic of the cantina situation is the wide range of behavior that is permissible and not inappropriate. This lati- tude is highly relevant for the social distance and indeterminacy to be 73 discussed later. Despite the looseness of behavioral norms in the cantina, however, the situation is potentially disruptable by any of the following general situational transgressions: display of weapons; at- tempt to play someone else's situational role; trespass upon a region reserved exclusively for another status. This concludes the description of the structure of the cantina situation. This structure should provide a framework within which to discover and interpret at least a part of the meaning of any act that takes place inside of the boundaries of the situation. Banging on the table means that the customer wants a drink; a couple dancing intimately does not mean that the two are in love; use of the formal pronoun of address means that the speaker wishes to assert a social distance be- tween himself and the addressee; and so on. As these examples show, the situational context is more relevant for some cases than for others. The analytical concepts employed in the structural description appear to be adequate, for I am satisfied that any stranger with a know- ledge of the language could use the information in this description to appropriately perform any role in the situation (cf. Goodenough 1957). The concepts of action cycles and of the two types of rules have proven especially useful in this respect, and I believe that they should con- stitute a part of the structural description of any social occasion. RELATION TO MEXICAN CULTURE The cantina is important in Mexican culture because it serves many functions and because it is patronized frequently by many kinds of people. There are several ways in which the cantina affects the lives and thoughts of Mexicans. From a sample of respectable working class men and women come the following expressed attitudes regarding cantinas. Women say: 1) My father goes there almost every payday and sometimes doesn't show up at home for three or four days. 74 2) My husband spends too much money that I need for food on beer and women at the cantina. 3) The cantina is a bad place with women of no shame and awful men. It is very ugly. 4) I don't want to hear that my boyfriend goes to cantinas. But I know that he does. Most young men do. 5) The cantina is located in an ugly and dangerous area of the city. Men say: 1) It is a refuge. 2) The cantina is where I go to celebrate payday. One must drink to keep healthy in order to work. 3) The cantina is a place for men. 4) I sometimes go there with friends and we enjoy living it up together. 5) There are some who think that they are macho going into the cantina. Arimoni, in a psychoanalytic exploration of Mexican urban culture, generalizes on the level of the two sexes: Si las mujeres sufren penas intensas, van al templo y al con- fesor. Cuando el hombre esta decepcionado, porque un negocio le salio mal, o una mujer le dio calabazas y sufre pena muy honda, va a su templo predilecto- la cantina--; no faltara ah'i alguien a quien hacer participe del acontecimiento y veri- ficar la catarsis (1965:294) . Thus the cantina becomes a male temple or refuge for cathartic release of tensions. On this level of generalization, the cantina is a special region in the territory of the Mexican man just as the confessional booth is a functionally similar region in the Mexican woman's territory. The verbalized attitudes above reflect this sex difference in cognitive patterning. There are other forces at work attracting individuals to the cantina, however, and these will be introduced before a discussion of the cantina as a "temple" and as a system of value maintenance. First, the cantina has business potential for the individual entrepreneur who has some, but not overly much, capital. Ownership can 75 provide a steady and dependable income in addition to many opportunities to acquire money that is lost or spent rashly. Many patrons have just received their pay when they come to the cantina. Some are too depressed to care how they spend their money. Others are anxious to impress. Consequently markup can be significant on some items (e.g. rum) and over- head is reasonably little. The waiters are usually owners, part owners, or close friends of the owner. To the musician the cantina is a godsend. Little talent and experience is needed in order to geta job, and with the practice gained by playing for an audience and with a group, the musician can aspire eventually to a higher paying job with more stringent qualifications. The prestige that the musician has in popular Mexico gives no small ad- vantage to someone who can woo his girl with his own talents, and any professional musician is quite highly ranked in the cult of machismo. Of the men who patronize cantinas, some are peasants who are in the city to buy or sell at the market, and who have nowhere else to go. They sleep in the cantina, drink, or simply watch. The cantina is also a place where a large number of young men go to gain experience in drinking or with women. In this way the situa- tion functions as a rite for initiation into manhood or as a practice area. The cantina can be a place to demonstrate one's machismo, or maleness, in the aggressive sense of the word. Verbal duels, drinking contests, contests of attractiveness to women, and outright fighting are ways of proving manliness within the cantina situation. The potential for cantina fights is one reason for the large number of policemen stationed in the cantina sector of the city. An indication of the place of machismo in the cantina is the periodic entry of peddlers carrying a battery powered electric shock producing gadget that, for the small price of one to five pesos, can be applied to a person's hands. A man who can remain holding both electrodes as the current is increased to maximum is demonstrably macho. 76 Sex is another attracting force of the cantina. The women are prostitutes of a sort, and for a price of from twenty-five to two hundred pesos can be induced to leave the cantina with a man, bound for a hotel room. The price is variable with the age and appearance of the woman, with the norm of the specific cantina, and with the particular engagement. Because of this last variable, a man is considerably more macho if he can talk a woman into lowering her price for him. The vari- able price is probably one thing that allows the cantina to compete favorably with organized houses of prostitution (many of which are con- siderably cheaper and cleaner than the cantina) where prices are fixed, leaving the male's charm and manliness in the background. The ability to dance well is highly valued in urban Mexico, and yet dancing requires learning and practice. In the cantina a person can learn and practice dancing without being embarrassed about awkwardness or lack of skill, because it does not matter so much what a cantina woman thinks, for she is a stranger, and in any case not a lady. She is, how- ever, enough of a female to be danced with. It would look bad, of course, to practice dancing with a man, and there are not many other places in the city that provide dancing practice without presenting the risk of being evaluated negatively by women who count. The situation, then, has another important function in this respect. As a context for the revelation of inner feelings and frustra- tions to companions or strangers, and for giving vent to anxieties and aggressions brought about by daily problems relating to women and business, the cantina has few competitors. Alcohol is a partial escape by itself, but alcohol in the cantina situation can provide a real catharsis. In this light such elements of the situation as sex, dancing, liquor, noise, machismo, etc., can be seen as well integrated and important contributors to the functioning whole. The symbolic denial of womanly attributes by cantina women must be deeply satisfying to the patron who may have been driven in frustration to the cantina by the outside woman: the one that has been made austere, chaste, pure, and passive by the church and by 77 cultural norms; the one that has been placed high up on a pedestal and cannot be touched without defiling her. The cantina woman is attractive by her opposition to the ideal. Because she is earthy she can stand the indignities heaped upon women, and she can serve as a stand-in for the wife or girl friend on which to unleash hostility and aggression in many forms. The fact that she drinks emphasizes her role, for "drunken- ness in a woman has a much higher visibility than that in a man, which can be traced to the symbolic qualities of drinking and drunkenness in women in the past, when drinking customarily symbolized the bawd and the harlot" (Lemert 1951:353). The woman that one wants to marry has been said to be on a pedestal, a possessor of verguenza. Yet many women inevitably lose their position of purity and succumb to temptation in varying degrees, espe- cially in the anonymous existence within the city. With this happening ever more frequently, a redefinition of moral values will be ultimately necessary. Nevertheless some factors are at work that allow the change in norms to take place more slowly. There is the church, earlier referred to as an important region of the female's territory, steadily maintaining and inculcating values that came over with the Spaniards. There is also a socially recognized dichotomy between those with verguenza and those without, between a suitable marriage partner and a sin verguenza, or shameless woman. When a girl has lost her chastity she is theoretically out of the running for marriage. With human motivations as they are, it would seem that this system must fall apart. Part of the reason that it doesn't is that men have a physical outlet in the cantina and the house of prostitution, so that they are willing to perpetuate the ideal for a woman. This they do in a variety of ways. Even popular songs invariably emphasize in heart rending terms the romantic ideal. Mexico is still a man's country, and the man wants to and can afford to retain the status quo regarding the position of women. This is another reason that the cantina now functions and will continue to function as a source of em- ployment for unmarried women with children to support. 78 DISCUSSION: DETERMINACY AND INDETERMINACY I would like to preface this discussion with a few remarks on communication systems in general, bearing in mind the fact that not all communication is verbal. In any communication system, sender and receiver must share an understanding and knowledge of a code that is used with a given message channel. If any information is to be transmitted and re- ceived (by means of a signal through a channel), there is implied a necessary minimum of shared knowledge about the code. Conversely, neither sender nor receiver needs to know the whole of a code in order that a shared understanding of the meaning of a specific signal be reached. All that is necessary is a knowledge by both parties of that portion of the code that is relevant to the signal. The greater the shared understanding of the code, the greater the number of signals that can convey intended information to the receiver. In the interaction situation, specifically the cantina, not only is it true that no two individuals know the whole code, but also, and perhaps more importantly, there is no whole code. Unlike the chess game, in which both players learn all the code rules before they make a move (signal), participants in the interaction game learn and modify the code as they go along, in part by a process of trial and error. From birth to death each person builds up a knowledge of many rules pertaining to many codes employing many channels, and each person has a unique set of all those rules (belonging to various codes) that are within his com- mand. Since people do learn many of their own personal code rules from participating in situations, and since in a particular situation much of the objective reality is available to many of the participants, there is bound to be some overlap between the code system of one parti- cipant and that of another. This overlap is guaranteed from the stand- point of evolution by the need to communicate effectively in society in order to survive, and is in fact substantial. The concomitant need to miscommunicate will be discussed later. 79 The situation provides a context within which to interpret correctly many of the signals produced within the cantina, although only some of the rules are affected by this context. For example, the verbal signal "traigame una cerveza" (bring me a beer) will have much the same referent or meaning in the cantina that it would have in many other situ- ations in which the Spanish language is used. In this case the situational context is not very important. On the other hand if a patron were to bang his empty beer bottle on the table, this signal could only be interpreted as a proper request for beer in the context of the cantina situation. Because the social situation is a vitally important context within which to interpret at least some behavior (signals), it follows that the structure of the situation is worthy of description. In spite of the fact that a specific bit of behavior can be described and inter- preted using only that minimal amount of the situational context directly relevant to the specific signal, the total context is important in inter- preting a large class of signals within the larger set of signals that can be perceived in the situation. That is, not all of the possible signals received in the cantina depend for interpretation upon situational context, but many do. When the patron enters the cantina, he is in a position to re- ceive many signals that may or may not convey messages to him. A waiter will rush up to him, another patron may come to his table and start talking about his troubles, a woman might stare at him, a drunk could embrace him, and so on. For the patron to respond appropriately he must perceive meaning in these acts. This he does to a greater or lesser extent. In fact, to the extent that he is correctly able to determine the meaning as it relates to the codes of the sender of the signals, the communicative transaction is determinate. I would define determinacy in a communication system as the transfer of information from one point to another. In a completely de- terminate system all information will be transferred from the sender to the receiver. Where there is no communication system no information will 80 be transferred. Indeterminacy in a communication system, then, will be defined as impediments to the flow of information. Few people would disagree with the statement that among humans there is a social need to communicate if the individual and the society are to survive. Social groups are actually based upon the flow of infor- mation from one person to another, and organization implies some shared understandings. I have proposed to look at the other side olf the coin. From this standpoint, the argument is not that social integration is based upon information flow, but that it depends upon the control of informa- tion flow. It relies then on miscommunication as well as on shared meanings. According to requirements for social integration and to the above defi- nitions, determinacy and indeterminacy must be juxtaposed in the course of any symbolic interaction. The view of indeterminacy as a necessity rather than a problem in communication is not entirely novel. Murphy has discussed the functions of social distance for the viability of multiple roles and the promotion of autonomy of action, concluding that constancy of demeanor is charac- teristic of all social distance techniques. "The actor achieves a refuge by submergence in his social identity and, through uniformity of behavior, discloses the least of himself, while maintaining his relations" (Murphy 1964:1272; emphasis mine). Uniformity of behavior is information control in this respect. Fernandez finds indeterminacy to be necessary in a different way. Analysis of data from a Fang reformative cult "shows us one area in which effective communication is resisted, in this case, in favor of ongoing ritual activity" (Fernandez 1965:922). There are two kinds of indeterminacy in human communication systems. The first is built in, so to speak, and is due in part to the uniqueness of individual code sets, which leads to trial-and-error meshing of codes in interaction. Culture, situation, and intimacy help to make this meshing less arbitrary or accidental, but there is always some ele- ment of trial and error. The second kind could be termed information control and is more relevant to this paper. Information control consists 81 of conscious or unconscious manipulation of signals perceived by others. This manipulation involves the withholding of information, purposive transmission of ambiguous signals, and the giving of unambiguously inter- preted false information (misinformation). The withholding of information is the most relevant here because that is what social idstance essentially is. The degree of understanding and intimacy characterizing social relationships varies inversely with social distance. In other words, the more that one knows about another, the more intimate the knowledge, then the less social distance between the two. Conflict of interests, role strain, and role conflict potentialities assure not only that in any interaction some role distance is maintained, but that in general, as Park says, "the larger the number the greater the distances" (Park 1950:183). Thus the situation structure is an important determinant in the need for, the amount of, and the kind of social distance displayed. Accessibility of engagements in the cantina forces a condition in which the waiter cannot effectively segregate his role relationships by physical barriers to perception, so he compensates by maintaining a stereotype which gives him just enough of just the right kind of social distance in encounters to fulfill his obligations and desires in a manner appropriate to the situation, probably the work of a kind of natural selection. In order to please everyone he is allowed to be quite distant, showing favoritism in his intimacies in very subtle ways that should not be discernible to any but the individual recipient. So too the cantina woman displays a role constancy to patrons that puts her socially equidistant from them all, but because of the situation this constancy takes the form of attempting to communicate intimately with all. She cannot diminish the social distance on a per- sonal basis to any large degree until she is out of the situation, for if she did her purse would suffer. It was pointed out above that among the patrons a surprising amount of intimacy can be displayed. The situation permits this because 82 within the cantina there are a variety of methods for maintaining social distance to compensate for the breakdown of personal barriers. Outside the situation some of the maintenance of social dis- tance can be accomplished by physical distance, the formality of inter- action processes, the use of metaphors, the convention of the lying male, and the exaggeration of real feelings. All of these can be manipulated to indicate distance or intimacy as desired, and all can be used to maintain ambiguity of intent, so important in social relations. Within the cantina many of these forms cannot or should not be employed, or else they are manipulated primarily to indicate intimacy. Distance therefore is attained mainly by the constancy of behavior cycles in the situation, the excuse of being drunk, the monetary basis of the relationship with the women, the latitude for ambiguity created by the wide range of ap- propriate behavior, and importantly the fact that most of the other participants are strangers to the individual and will not intrude upon his roles outside of the situation. 8 Social distance, insofar as it preserves ambiguity of intent and withstands attempts to penetrate to the depth of the self, is a portion of the indeterminacy that must accom- pany interaction. Humans strive in communication for determinacy up to a point, and past that point they seek indeterminacy. The location of that point varies with the specific encounter and is influenced by participants, situation, and culture. CONCLUSION The cantina has been seen as an integral part of Mexican cul- ture and as a highly specialized setting for interaction. Behavior in the cantina has been related to the problem of meaning in interaction and to the enactment of social roles. Interaction has been shown to possess a determinate and an indeterminate aspect, and the attempt was made to delineate the structure of the specific situation. 83 The organization of concepts employed in the presentation of the structure of the situation has some implications for social theory, i.e., that any structural interpretation of a system involving the pat- terned changing of relationships between social units must make use of the concepts of action cycles and rules for action in some way. The question of determinacy and indeterminacy in interaction that has been dealt with here has a more specific application. Mecha- nisms for social distance maintenance are by implication mechanisms for manipulating the amount of indeterminacy in communication. Observation of these mechanisms being employed has led to the characterization of classes of people in terms of significant attributes. The attributes are believed to inhere in the person rather than his momentary role. In other words, some of what are thought to be individual personality attri- butes showing through in a person's performance of a given role are actually attributes of the role in that they are shaped or determined by the exigencies of the situation in which the role enactment takes place. The particular mechanisms for maintaining social distance in a given situation are selected from a wide variety of possibilities. Knowledge of the pool of potential mechanisms in the culture under consideration and understanding of the structure of the particular situation should make it possible to predict in part the kinds of mechanisms selected for use in a particular role performance. I believe that folk stereotyping (e.g., "the jolly waiter") can be partially accounted for on the basis of predictable manipulation of indeterminacy in interaction. NOTES 1 The fieldwork and write-up time on which this research is based was supported by the National Institutes of Health (National Institute of General Medical Sciences), Training Grant No. GM-1224. Field research took place during the summer of 1965 in Jalisco, Mexico. This paper had benefitted by critical comments from Gerald Berreman and May Diaz. 84 2 For the reader who is unacquainted with Mexico, the type of cantina herein discussed is a Mexican drinking place approximately com- parable to the saloon that once existed in the United States. 3 I have referred to the cantina as a situation and as a social occasion. Goffman defines "situation" as "the full spatial environment anywhere within which an entering person becomes a member of the gathering that is (or does then become) present" (Goffman 1963:18). A gathering is a set of individuals in one another's immediate presence, and "presence" is defined in terms of mutual monitoring capabilities. If "immediate" can be interpreted somewhat loosely, then it is in fact the situation that I am describing. However Goffman would prefer to call it a "social occasion," "the wider social affair, undertaking or event, bounded in space and time and typically facilitated by fixed equipment" (Goffman 1963:18), or a "behavior setting" after Barker and Wright (1955). I shall use the three terms interchangeably. 4 The study was carried out in Guadalajara, near the market of San Juan de Dios, where there are twenty-four cantinas within a four block area. Participant observation and a paid forty year old male informant supplemented information gained by asking numerous questions both inside the cantina and out. 5 This use of social distance can perhaps be profitably com- pared with that employed by a bartender or a hostess at a party, both of whom face similar problems. The desirable stereotype will differ in some respects from that of the waiter, and should fit the situation. 6 Goffman defines a back region for behavior as a "place rela- tive to a given performance, where the impression fostered by the perfor- mance is knowingly contradicted as a matter of course" (Goffman 1959:112). 7 This passage can be freely translated to mean the following: when women have worries or problems they go to the confessional in church. When a man is disappointed, because a business deal fell through, or be- cause a woman turned him down, and he is deeply grieved, he goes to his corresponding temple, the cantina, where there will always be someone with whom to share his experience and to confirm the catharsis. 8 This aggregation-of-strangers aspect of the cantina does hold for types of cantinas other than that discussed here. At least three types could be distinguished: 1) the small town cantina--no women; 2) the neighborhood cantina--few or no women, not located near the market centers of the city, serving primarily the same clientelle day after day; 3) the urban cantina (described here)--with women, steadily employed musicians, located near the market, open all night, and with a shifting clientelle. 85 REFERENCES Arimoni, A. 1965 Psicoanalisis de la Dinamica de un Peublo. Mexico: B. Costa-Amic. Barker, R. and L. Barker 1961 Behavior units for the comparative study of cultures. B. Kaplan, ed., Studying Personality Cross-culturally. Evanston: Row, Peterson and Co. In Barker, R. and H. Wright 1955 Midwest and its Children. Evanston: Row, Peterson and Co. Cavan, D. 1966 Liquor License. Chicago: Aldine. Fernandez, J. 1965 Symbolic consensus in a Fang reformative cult. American Anthropologist 67:902-929. Frake, C. 1964 A structural description of Subanun 'religious behavior.' In W. Goodenough, ed., Explorations in Cultural Anthropology. New York: McGraw-Hill. Goffman, E. 1959 The Presentation of Self in Everyday Life. New York: Doubleday. 1963 Behavior in Public Places. Glencoe: The Free Press. 1964 The neglected situation. American Anthropologist 66:6:2 (Special Publication, The Ethnography of Communication, John J. Gumperz and Dell Hymes, eds.):133-137. Goodenough, W. 1957 Cultural anthropology and linguistics. Georgetown University Monograph Series on Language and Linguistics 9:167-173. Lemert, E. 1951 Social Pathology. New York: McGraw-Hill. Murphy, R. 1964 Social distance and the veil. American Anthropologist 66: 1257-1274. Park, R. 1950 Race and Culture. Glencoe: The Free Press. Paz, 0. 1961 The Labyrinth of Solitude (trans. by Lysander Kemp). New York: Grove Press. 86 D L I J L I J JUKE6I XE LIZ LIZ H U~~~~~~~ LI j x SEX Cl~~X XLC '$TE P' D r I _ _ _ _ _ _ _ _ _ Fig. 1 Soma of the cantina situation. 87 MU:I CI AN] . 44 v q% I Lii Lii %P 2 w 2 a SER.VICE (DAWCE, ETC.) HoMEY I5EE.R _~~~~~~~~~~~~~~~- Fig. 2 Role structure of the cantina situation. 88 PAT2ON GrO ON Pu1'S GET 0 R OM RFTURN rO 1'APAE. WI1V eQINI(s) WAI TER DANCE W/ PATRIOJ VATRoN -4* I *Fig. 3 Action cycles of participant roles. 89 5 tv OOWN Cro ON Ca)rU WONAN 1ALI4 W/ WOMEN -1T AT PATrRoXS -Ai-E- 9I NI v 0 A * Ger PAI 12 -WC &LEAVF WI I