A Study of the Process of Affiliation vith Alaabolics Anonymous

Among the numerous resources available in the treatment of alcoholism, the voluntary fellowship of Alcoholics Anonymous occupies a prominent position. This nonprofessional group has apparently achieved a success as great, if not greater, than the efforts of medicine, psychiatry and psychology. Lacking any professional standing, it has merely appealed to problem drinkers to join with other drinkers of their kind in mutual support against the uncontrolled need to drink. For approximately 20 years this unguided form of alcoholic-treating-alcoholic has grown and spread on the American scene. Rarely are treatment activities described without a prominent place being accorded to “A.A.” It has been eulogized and analyzed, but seldom have its “negative instances been scrutinized. Emphasis has been placed on what happens after an individual affiliates with it; only infrequently has attention been given to the question of how affiliation came about in the first place. In the flush of favorable publicity and widespread recognition, one factor in this “group therapyu has been overlooked, namely, the alcoholic who is exposed to affiliation with this group but who, for some reason, “doesn’t take to it.” Usually it is maintained that some problem drinkers are to join while others are not. This immediately raises the question, what constitutes “readinessn? What explains the fact that among all the persons suffering from the same trouble, with many common experiences, some are able to align themselves with an A.A. group while others remain nonaffiliates? Do the experiences and attitudes of those alcoholics who affiliate with A.A. exhibit characteristics that set them apart from those who do not affiliate? The research effort to be described herein was aimed at this guiding question: What constitutes readiness to affiliate?

The available literature indicates a mild awareness of the problem of nonaffiliation, accompanied by incidental hunches as to its explanation. Bill W. (1), one of the co-founders, states that the majority of nonaffiliates “have powerful rationalizations to be broken down.” Wilson (2) is baffled by the inability of some to join: “With all the advantages A.A. offers, one cannot resist speculating on what characterizes those potential recruits who are unable to profit from membership,” but he offers no explanation. Often affiliated alcoholics are described as being more receptive to A.A. principles. Jellinek ( 3 ) , for example, speaks of some alcoholics as “being those types who are attracted to A.A.,” but does not indicate what constitutes these types. Ritchie (4)asks the question, “How does the factor of selectivity operate with reference to the types and classes of people who become A.A.ls?If but suggests no answer. Still others mention personality factors. Brown (5), after dividing a sample of alcoholics into neurotics and psychopaths by the Minnesota Multiphasic Personality Inventory, concluded that “the neurotic drinker may be more likely to respond to the supportive group measures offered by A.A.” Chambers (6) states that “most of those who become members have been lonely, isolated people who in the course of their abnormal drinking lives have lost their friends and contact with society.” Slight (7) believes that a l*secondarygl type of alcoholic is more disposed to affiliate. Finally there are those who suggest that nonaffiliates “have difficulty merging with a group pattern.” Jackson and Connor (8) go so far as to suggest that “A.A. may be more successful with the Skid Road alcoholic because it manipulates the pattern of group behavior which he already possesses toward a new end, sobriety.” ‘ As far as can be ascertained, no research has been done to test these speculations. The over-all picture concerning A.A. seems to be a widespread acceptance of its activity with no systematic consideration of the barriers that prevent many alcoholics from affiliating.

In the light of this meager literature regarding affiliation, a series of exploratory interviews with both affiliates and nonaffiliates was undertaken.

These were directed towards developing more precise “hunchesw that could be tested on a second group of alcoholics. Forty-six such interviews were held, lasting approximately 2 to 3 hours each. They were preceded by rapport-building efforts. The subjects were encouraged to talk about what their experiences and attitudes had been as they tried to affiliate with Alcoholics Anonymous.Groups in Madison, Wis., cooperated, while alcoholic patients at the Mendota State Hospital in the same city were available as nonaffiliates. These interviews produced three results. First, it was apparent that the subjects tended to think in terms of three phases regarding the affiliation process. These were experiences and attitudes (1) before going to any meetings at all; (11) at the time of initial contact with a group; and (111) after attending meetings for a few weeks. Second, they thought in “black and white” fashion. “1 joined because I admitted I was licked;” it was a matter of “yesn or “no;” “quit or continue;” “dry or wet;” “across the line” or “not across the line.” Third, a series of statements concerning the three phases was collected from the interviews, statements which lent themselves to the wtrue-false” response tendency just noted. In this form they could be suhnitted to a larger sample of both affiliates and nonaffiliates whose responses could be compared in order to see what, if anything, discriminates between them. These statements (Chart I), as classified under each phase, finally numbered 57.

8. The year bcfore going to any A.A. meetings at all I had at least two friends who would have laughed at me for trying A.A. as a way to stop drinking. 9. The majority of the people I had known in taverns before I went to any A.A. meetings at all tended to joke about A.A. 10. Bcfore going to any A.A. ~iwcti~igsat ;dl I had l ~ c a r dthat thc ~iiiwilwrslived up to it. 11. Before going to any A.A. nleetings at all I had a t least one relative whom I liked and who had quit drinking through A.A. 12. Before going to any A.A. meetings at all I had at least two good friends who would have admired me for joining. S o c u ~ISOLATIONEXPEHIEXCE 13. During the year bcfore I went to m y A.A. meetings ;at all I did niost of my drinking with at le.tst two or three other drinkers. 14. During the year before I went to any A.A. meetings at all I had not lived in at least two places other t l ~ a nthe one where I went to my first A.A. meeting. 13. Dcfore I went to m y A.A. meetings at sll I had hird very little e~pcrienccwith groups of any kind. GROUPEXPERIENCE LL.E A.A. 16. Before I went to any A.A. meetings at all I had belonged to at leist one group like the Eagles. hIwse, Elks, Masons, V.F.\V., American Legion, etc. 17. Before I went to any A.A. meetings at all I had tried at least once to get people to join a group of which I was alrttady a member. 18. Before I went to any A.A. meetings at all I had belonged to at least one group in which I was expected to give help to other me~ubersif they needed it.

P h e 11: lititid Contuct Experierrces Ex~~m.4mons 1. When I first went to meetings they were what I had expected. 2. \Vhen I first went to a few meetings I did not espect someone to tell m e how to be a controlled drinker. 3. The majority of the people in the comnunity where I first went to 1111 A.A. nieeting did not think much of A.A. CLOSESESSOF INITLALCONTACT 4. After I had first showed an intcrcst in joining A.A.. thc group kept a close watch over me. 3. Durir~gthe tinw of those Iirst fcw A.A. ~urrtingsI w;:s iponmrcd; in ot11c.r words, there was one certain ~ i i r ~ ~ i b wcl r~ ol~elpcdnw get on tlw progr,tm. 6. h number of old friends were in the group when I first went to ~ilcctings. I s c r n u ~ n o x.ro AFFILIATE 7. I thought I was ready to join A.A. when I went to those first few n~eetings. 8. \\’hen I first went to A.A. meetings I had decided thit the troubles from drinking outweighed the pleasures froin drinking. 9. After being in a hospital or treatment centcr you citrl scc what might happcn to you; this was not a reason why I first went to A.A. ~ueetings. CLASSS L Y S I T ~ 10. At the time when I first went to A.A. meetings I did not think that a lot of “big shots” u i t h nloney belonged.

11. When I first went to A.A. meetings I did not have a job, fairly decent clothes, and some money to keep me going. 12. When I first went to meetings I did not figure that the people who got on the program were “well fixed and well heeled.”

IMPETUS TO GO 13. No one urged me to get in touch with A.A. when I first went; I decided to do it myself. 14. I went to the first few meetings to please a relative or friend. 15. I felt that I was pressured into going those first few times. SPONSORSHXP EXPERIENCE 16. I like being with my sponsor as much as being with n good friend of mine. 17. He was about the same age as me. 18. He was not in the same social class 3s me. 19. H e was not the same religion as me. 30. I had not known him before I went to any A.A. meetings at d l . 21. I had not known him when he was a r e d “boozer.”

Plurre I I I : After Attending u Feu- ~ V C C ~ S

A~SOCIATIONAL PRESSURES( 1) 1. After I had been going for a few weeks I tliscovered that I came to like the cards and chit-chat after the meetings nlniost as much as the meetings thenlselves. 9. After I had been going a few weeks I discovered thnt it wtis easy to give up my drinking friends. 3. After I had been going for a few weeks I discovered that old friends in the group made it easier for me to keep on going. COSIPETITION: OTHERGROUPS( 1) 4. My wife (or girl friend) went to a few meetings with me at this time mtl took to the idea of A.A. 5. At this time my family did not back me up; I did not think that A.A. was necessary bemuse my fiimily helped me o r ~ tof lily drinking troubles. 6. My wife (or girl friend) was w t u controlled drinker and I did not h.~ve to choose between her and A.A.

7. After I had attended meetings for a few weeks I realized that the members were genuine alcoholics. 8. At this time I discovered thut I coulJ accept the spiritual ictm in .4.A. 9. After I had attended for a few weeks I did not decide that my drinking troubles were less than the others there. ASSOCIAT~ONAL PRESSURES( I)) 10. After going for a few weeks I did not get lonesome and have to drink to be with people. 11. At this time I discovered that I could get on the program and go around with the people I liked. IS. Alter I had gone for n few weeks I was not scared thnt someone w a d d “tell it around that I was going to A.A.”

1:;.I h r i n g this period I tlid ~ i a t11.tr.c. ;In t ~ n p l c ; ~ s m t 1-xpcrirncc with n mcnthcr that “left it bitd taste in n y mouth.” 14. During this time m y esperiencc with A.A. did not come largely from one or two niemhers. 13. During this tiuw I w a s not I)otltrred by thc- iclca;i o l b r i n g cnlled on to tell nly o\vn $tor\:

16. .+fti.r I hat1 :~ttcwtietl~ncvtingsfor a ftbw \vcc.ks I tlid not decide that my rcligion tool; the pl;tcc of A.A. rnet.tings I’or it few \vcvks I ‘lid not dccitlc that I \v;is too 17. ;\f’tc.r I Ii;~cl c ~ t t t v i t l ~ ~ t l invol\.c.d in I I I ~0\\.11 \\.orli to go to nit-ctings. 1s. .+t this tituc I W A S too iuvolvctl with relatives :mil fricncls to go to meetings.

The statements were next submitted to alcoholics who were asked to respond with true or false to each one. The sample of nonaffiliates of A.A. consisted of 141 male nonpsychotic alcoholic patients at the Mendota State Hospital. Due to earlier efforts to develop rapport the interviews could be conducted informally. Each statement was read to the subject and his response recorded. All male alcoholics “without psychosis” at the hospital during the months of November and December 1954 and January, February and March 1955 were covered. A collection of responses from a representative sample of A.A. members was not easy to achieve. The respondents had to be volunteers. Furthermore, the number was limited by the fact that, unlike hospital admissions, membership in the local A.A. groups had no constant accretion. Another complicating factor was that the total membership of any A.A. group was never available. From three groups in Madison, Milwaukee and Janemille, Wis., a total of 111 members cooperated. Informal discussions with group secretaries indicated that approximately 6 out of 10 members in the cooperating group responded. Among other relevant data gathered from these subjects was the rate of attendence at A.A. meetings during the past year (outside of a hospital or treatment center). All those subjects who indicated they had attended meetings at least twice a month during the past year were considered affiliates. Only 5 per cent of the total respondents indicated attending precisely twice a month during the past year. The bulk attended either “none” or “once a yearn at one end of the scale, or “twice a week” or “more than twice a week” at the other end.

The true-false responses of the affiliates to the statements were compared with those of nonaffiliates by clusters of three statants. Comparisons were made by Chi-square tests. If these tests showed significant differences, the degree of relationship between affiliation and the significant factors, both singly and as a cluster of three, were measured by McCormick8s (9) technique. Tables 1-3 show the raw percentage differences as well as the combined correlation of the three factors with affiliation. The percentages are in terms of the affiliates; for example, of those answering “truev to the first statement in Table 1, 73.6 per cent were affiliates; of those answering “falseu to the same statement, 30.7 per cent were affiliates.

Differential ~=i.ssocicrtioti 4 . No friend who quit by will po\vcr 5. Had already lost drinking friends (3. i’rcscnce of two A.A. f r i c d u 7. Ko relative who quit by will power 8. No friend who would 1;111gl1 9. T:wern friends joked illout A.A. 10. l I ; d Im1rc1 members lived 1111 to it 11. Re1otit.r who quit throrlgh A.A. 12. F’rirtds who would ;idwire Sociul lsolutioti Expericrtce 13. L ~ I I Id u~ n e 14. Livctl in one pl;~c.e 1.5. lfi~tllittle group experience Crorrp Experiences Like A.A. 16. One fr.~ternl~l order 17. Espect:~tion-new rnc-rnl~ers IS. Espectntion-lirlp others *Significant n l thr 5 yrr crnt Irwl of cri~lli&ncr. t Sot ,igrlificnnt.

TABLE2.-Experiences at Initial A.A. Contact Affiliates ( % ) Combined False Correlation 1. Slrctiogs what expected 3. Controlled drinking not expectrcl 3. So commtlnity ridicule 71.0 63.5 55.2 36.0 37.8 52.8 ] Closct~ersof I n i t i d Contact 1. Group kcpt close watch 5. \\’as sponsored 6. Old friends in group 32.0 29.5 45.6 } .6*

lrtclirrc~tio~~ to Affiliate 7. Felt was re:ldy to join 8. T r o u l h more than pleasures 9. Sot s(.:I~c~ by 1iospit.d experience 63.6 62.2 62.5 37.8 23.1 47.1 } Am

Clciss Sen iti icily 10. “Big Shots” did not belong 11. Did not have clothes, etc. 12. l’t-oplr in A.A. not well fised 67.4 66.0 58.6 30.5 51.5 44.4 } .*

13. lkcided to go myself 60.3 50.0 14. \!’cnt to please rel;ttive, friend 60.2 51.5 t 13. Felt was not pressured 50.3 51.7 Sponsorship Experience 16. Like lrinl as much as good friend 70.7 42.8 17. \Vas same age 71.9 59.6 IS. Sot same social class 71.2 64.6 19. Sot same religion 70.5 60.6 00. Not known before went to -4.A. 71.0 01.4 31. Not known when he was a hoozer 68.9 63.1 l Significant at the 5 per cent level of confidence. t Not siqnifirant.

Certain experiences before going to any A.A. meetings at all (Table 1) discriminated significantly between affiliates and nonaffiliates. Most prominent of these is self-conception as a person who could share basic emotional reactions with others (statement 1). Such a self-perception seems to be more meaningful for affiliation than self-definitions of being “licked” or labeling oneself an “alcoholic.” Affiliates apparently came from a background in which verbalizing emotional reactions to others was a norm of individual behavior. In turnt this self-conception meshes harmoniously with those norms of A.A. that emphasize “get it o’ff your chest,” “tell your storyt”and “coffee-clutch therapy.” In contrastt nonaffiliates did not perceive themselves as persons who readily shared emotional reactions with others. It is interesting to note that very little

discrimination between the two groups existed concerning the self-label “alcoholic” (statement 3). Apparently many nonaffiliates as well as affiliates applied this term to themselves before ever going to an A.A. meeting. Among the various associational experiences recounted by the subjects in the exploratory interviews, three seem to be highly conducive of affiliation. The absence of a close friend or relative who had quit drinking by his own will power was decidedly influential (statements 4 and 7). The nonaffiliates had a significantly larger number of associations with esteemed friends or relatives who, they thought, had stopped using alcohol by individual volition. Thus, they had a “will power model” in their immediate associational experience that competed strongly with A.A. As one nonaffiliate put it, “One of these days 1’11 do

TABLE 3.-Experiences After a Few Weeks of A.A. Meetings Affiliates ( 9 ) Combined A.ssociational Pressures ( 1 ) Trrrc Ftrlse Correlation 1. Liked ciirds, etc. after meetings 72.6 23.4 2. Easy to give up drinking friends 71.5 26.6 3. Discovered old friends 66.6 30.4 Competition: 0 t h Croups ( 1 ) 4. Wife went and liked it 61.0 5. No backing by family 63.7 6. \Vife not a controlled drinker 61.2 Attmctice Features ( I ) 7. Members genuine alcoholics 65.7 8. Accepted spiritual ideas 58.3 9. Drinking troubles similar 58.4 10. Not loneson~efor drinking situation 62.2 11. Coiild still be with friends 01.1 12. No fear of gossip 69.3 Attrnctice Features ( 2 ) 13. No bad experience with a member 61.9 14. Impressions not from 1 or 2 members 62.5 15. Not bothered by telling story 58.9 Competition: Other Croups ( 2 ) 10. Religion did not take place of .\.A. 58.1 17. Not too involved in work to go 57.1 18. Not too involvcd in family to go 57.6 Significnnt nt the 5 per cent level af confidence. t Not significant.

like my old uncle. One day he threw his bottle down on the floor and swore never to touch another drop, and he didn’t. The fact that his “old uncle” may not have been an alcoholic, or may never really have stopped drinking, is beside the point. The nonaffiliate is convinced his model did stop by his own will power. Thus when he comes to A.A. meetings he says to himself, “This stuff is all right, but why should I get mixed up in something that tells me to give up? One of these days I’ll stop by myself like old uncle Joe.” It must be borne in mind, too, that this reaction is highly approved by our society, which places a high premium on individualism, on independent action, on “standing on your own two feet.” With a model of will power in the immediate background, backed up by this potent social pressure, the action of the nonaffiliate is easy to understand. The affiliates did not have to overcome this barrier. They did not have to choose between a “will power modelu and group surrender. A second associational experience that differentiated sharply between affiliates and nonaffiliates is the fact that the former had, before going to any meetings at all, already lost their drinking friends, while the latter had not (statement 5). This suggests that the affiliates came from a background in which “symptoms” of alcoholism are readily stigmatized while the nonaffiliates came from social situations in which these manifestations are more readily accepted as “normalw behavior. It also indicates that group supports had been removed Erm t ? ~ drinking patterns of the affiliates before they approached A.A. In contrast, the nonaffiliates had made contact with A.A. while still active in spontaneous, informal drinking groups. This point raises a question concerning the significance of lllone’l drinking in the history of alcoholics. If the affiliates went through such a stage before going to meetings, and the nonaffiliates did not, perhaps it indicates that a “lone” drinking phase pves the way for more effective therapeutic efforts. In the past, “lone drinking” has been regarded (3) as a grave and portentous symptom of alcoholism. It may be useful to think of it also as signifying the removal of drinking-group supports, in the absence of which a full realization of the problem may develop more readily. A third factor in the statements classified as “differential association^ shows a clear-cut distinction between the two groups. The affiliates, in contrast to nonaffiliates, had experienced significantly more contacts with persons who described A.A. members as sincere and living up to the program (statement 10). If,

before going to any meeting at all, an alcoholic had been exposed to favorable descriptions of the sincerity of members, his chances of affiliation were increased significantly. Apparently those A.A. members who quietly and without fanfare live up to the principles of the group create impressions that become a part of favorable This, in turn, was often the precipitating factor that helped a nonmember become a member once he tried to “get on the program.” On the other side of the coin, nonaffiliates had been exposed to an excess of hearsay contacts that defied members as “going to meetings and drinking on the side.” Obviously an element of chance is implied here. It is quite possible that unfavorable hearsay was experienced by chance by some subjects while others came into contact with more favorable communications about A.A. equally by chance. On the other hand, it is quite possible that nonaffiliates have experienced a combination of barriers to affiliation, one of which is a rather uniform stigma of A.A. members as hypocrites. Regardless of these possibilities, one point is clear: Those alcoholics who had been exposed to favorable hearsay communications affiliated at a significantly higher rate than those who had not been associated with such definitions. The other cluster of statements in Phase I did not create significant differences between affiliates and nonaffiliates. Though the raw percentages show some variation, these factors cannot be regarded as definitely related to affiliation. In summary, certain antecedents seem necessary for affiliation, but do not assure it. These are: (1) self-definition as a person who shares basic emotional experiences with others, (2) the absence of behavior models of will power, (3) the loss of long-time drinking group supports, and (4) exposure to favorable comications about the sincerity of A.A. members.

Phase I1

At the time of initial contact (Table 2) the probability of affiliation is further increased if there is present in the expectations of the alcoholic an accurate conception of what will take place at meetings (statement 1). In significantly greater proportion, affiliates came from a system of communications that contained accurate descriptions of what A.A. meetings are actually like, while nonaffiliates comparatively did not share in this clarity of expectation.

These latter apparently had confused expectations; often, in exploratory interviews, they spoke of expecting a similarity to church, expecting more ceremony, more speechmaking, more prayer. At this timer affiliates experienced a closeness of contact with the receiving group that stood in marked contrast to the experience of the nonaffiliates (statements 4 and 5). Not only were the affiliates sponsored, while the nonaffiliates tended not to be, but the receiving group stayed in closer over-all contact with them. In other words, the entire group, or at least large segments of it, were involved in receiving the newcomer. Exploratory interviews revealed that the sponsor acted as a bridge to extend contacts with numerous other members who, in turn, initiated association with the potential member. Nonaffiliates do not report this experience. Sponsorship was typically absent as well as the wider contact with the group just after first going to a meeting. Sociologically this is what would be expected. The first step in affiliation was already present for affiliates. What was required next was a reasonable aggressiveness on the part of the receiving A.A. group. What was needed to push affiliation forward was the assertion by the group of its norms and controls. Coupled with this group experience of the affiliates was an inclination to affiliate that apparently arose from a conviction that life without alcohol could be more satisfactory than life with it (statements 7,8 and 9). It may be that affiliates and nonaffiliates are products of two different behavioral systems, each with its own values as to what constitutes “pleasure,” “fun” or “trouble.” The nonaffiliates seem to spring from subcultural groups that do not define as nsymptomsu those drinking experiences that the affiliates came to regard as usymptoms.” They seemed to associate the pleasures of living much more closely with these drinking experiences than did the affiliates. The latter had come to a “fork in the roadu- they realized they had to choose between home, job, c o m i t y esteem and continued drinking. They represented a value system that had more definite limits on what constitutes the rewards of drunkenness. A final attitude emerged from the comparisons concerning the initial contact phase. When social class sensitivity is present in the newcomer there are numerous opportunities to be repelled from the group (statements 10 and 11). The nonaffiliates differ strikingly from the affiliates in their sensitivity to social class symbols. They ranked the conviction that “big shots” belonged as the top sensitivity, suggesting that, along with the derogatory gossip about the

insincerity of A.A. members, they also were imbuded with the idea that members were of higher social status than themselves. It has already been seen that the background of the nonaffiliates differed sharply from that of the affiliates in that they had different definitions of llalcoholic symptoms,” different self-definitions and different hearsay descriptions of A.A. It seems plausible that this background sensitized them unfavorably toward symbols of social class, especially when they believed such symbols to be present in A.A. In summary, Phase I1 of the affiliation process can be tied to Phase I as follows: If before going to any meetings at all, a problem drinker possessed emotion-sharing self-definitions, if will power models were absent from his associations, and if he had experienced favorable hearsay about A.A., the probability of his affiliation was good. If, upon first going, he had clear expectations of what A.A. was like, if the group kept in close contact with him, if he was convinced that drinking troubles were greater than drinking pleasures, and if he did not have a social class sensitivity, the likelihood of his affiliating was closer to realization.

THe affiliation process moves towards completion in the experiences and attitudes that develop after attending meetings for a few weeks (Table 3). Affiliates, in impressive numbers, were attracted to the informal activities before and after meetings. This made it, compared with nonaffiliates, easy to give up their drinking friends (statement 1 and 2). The therapy of A.A. is thus revealed as one of informal, spontaneous groupings, centering around a formal base. Much of its effect lies in the unstructured networks of interaction in which an ease of discussion and a permissive atmosphere stimulate participation and self-involvement. The more formal program is important, but the newcomer who can become a part of the natural informal clusters of “chit-chat” has moved a long way toward affiliation. The group itself can make a positive effort toward incorporating individuals in these clusters. On the other hand, personality problems may prevent many newcomers from responding. It is clear from the data, however, that inclusion in these relationships operates strongly in favor of affiliation. Two aspects of family life tended to influence affiliation during this phase

(statements 4 and 5): the wife and blood relatives. In the case of affiliates, the wife (or girl friend) tended to go to meetings and support what she saw and heard. She may well have been the embodiment of those values that defined a sober life as more desirable than an alcoholic one. Furthermore, she tended not to be a drinker. Consequently, she did not force her husband into a position of choosing between her and A.A. Nonaffiliates were not so fortunate. Often their wives presented them with a hard choice: drink with the woman you love or abstain with a relatively strange group. Along with this dilemma the nonaffiliates apparently experienced, during this time, a great deal of support from kindred in their drinking behavior and in the problems deriving from the drinking. Affiliates did not experience this support in the intensity that characterized nonaffiliates. Another point that merits emphasis emerged in this period. Affiliates came to believe that members of A.A. were genuine alcoholics (statement 7). Nonaffiliates did not see this equally sharply. This seems to underscore the likelihood that affiliates came from a background in which the symptoms of alcoholism are regarded as pathological; in contrast, nonaffiliates reflect reference groups in which these symptoms are nothing unusual. This, in turn, raises the question of the definition of alcoholism. Current definitions have been developed by middle-class, clinically oriented persons who have paid only scant attention to the variations in emphasis placed upon certain “symptomsnby persons in other strata of American society. They have made much of the claim that losing control, experiencing “blackoutsM and utremors,uand the like, constitute indications that alcoholism is an illness. The data and interpretations of the present research indicate that there are subcultural groupings within which these wsymptomsw are not regarded with unusual alarm. Such groupings do not readily foster the decision that one is “lickedu by the presence of these features. “Will powerttis an assumed substitute for the problems of alcohol, if and when it becomes a “problem.” There exists in these groupings a tendency to see the pleasures of alcohol as compensating for these “symptoms;” the “problemsu of alcohol become, to them, necessary byproducts of a recreational outlet, not an illness to be treated. The affiliates were able to accept the spiritual ideas in A.A. more readily than nonaffiliates, although the difference was not as great as might have been expected (statement 8). This relates to the fact that affiliates had more regularly attended church in childhood than nonaffiliates. This was established through supplementary data, gathered from respondents, relative to church

attendance during childhood and during the 5-year period before first going to an A.A. meeting. It is consistent with the general finding that affiliates derived from an Over-all behavior system sharply different from that of the nonaffiliates.

By way of recapitulation, a tentative description of the process of affiliation with A.A. can be spelled out. But it must be emphasized that owing to the sampling problems which could not be solved, this process cannot be generalized to other samples. Nevertheless, some casual implications can definitely be drawn. The process begins before the problem drinker ever goes to a meeting. If, at this time, there is a self-definition of sharing emotions and no will-power model in the background, if long-time drinking friends are lost and there is exposure to favorable hearsay regarding the sincerity of A.A. members, a potential for affiliation has been produced. Upon attending meetings, this potential is brought closer to fruition if the problem drinker has clear expectations concerning what meetings are like; if he is sponsored and the group exerts a positive effort to keep in close contact with him; if he has decided that the troubles of drinking outweigh the pleasures of drinking: and if he is not sensitive to social class symbols. The affiliation process is on the way to completion if, after attending meetings a few weeks, the problem drinker can readily adjust to the small, informal, spontaneous groups that develop before and after meetings, if his wife (or girl friend) does not compete with A.A., if his kindred have refused support in problems that arise from excessive drinking, and if he has been raised in a system of values that recognizes the sign of having a drinking problem.

Stated negatively, this process consists of barriers to affiliation, to meet which, specific types of therapeutic action can be suggested. It seems reasonable for treatment agencies to develop pre-A.A. group therapy aimed at reducing the barriers that exist in any case. Specific sessions could be aimed, indirectly, at such items as will-power models, emotion sharing, the nature of ltslips,lt accurate

descriptions of what takes place at meetings. In such sessions the association between pleasure and drinking could be attacked, in an effort to create a “fork in the road.” Social class sensitivity could be explored and possibly reduced. Specific suggestions to A.A. groups can also be derived from the findings. The value of sponsorship coupled with a positive approach toward newcomers seems apparent. Specific efforts aimed at including him in the small, informal groupings before and after meetings seems to be especially important. But probably the basic significance of the findings in the discovery that factors can be isolated that discriminate between affiliates and nonaffiliates. If the search for differentiating items should be pressed so that generalized predictions might be extended to new samples, it seems possible that the rate of affiliation with this highly effective group could be materially increased.

1. W.W., Co-founder. THe Society of Alcoholics Anonymous. Amer. J. Psychiat. 106: 370-3731 1949. 2. Wilson, R. An evaluation of treatment methods in alcoholism therapy. Ment. Hyg. 35: 260-290, 1951. 3. Jellinek, E.M. Phases in the Drinking History of Alcoholics. Analysis of a Survey Conducted by the Grapevine, Official Organ of Alcoholics Anonymous. New Haven; Hillhouse Press; 1946. Also in Quart. J. Stud. Alc. 7: 1-88, 1946. 4. Ritchie, O.W. A sociohistorical survey of Alcoholics Anonymous. Quart. J. Stud. Ale. 9: 119-1561 1948.

5. Brown, M.A. Alcoholic profiles on the Minnesota Multiphasic. J. Clin. Ekp. PSYC~O~. 6: 266-269, 1950. 6. Chambers, F.T. Analysis and comparison of three treatment measures for alcoholism: Antabuse, the Alcoholics Anonymous approach, and psychotherapy. Brit. J. Addiction 50: 29-41, 1953. 7. Slight, D. Alcoholism and alcohol addiction. Proceedings of the 2nd annual conference on alcohol studies. (Mimeographed.) University of Wisconsin Extension Division; 1949. 8. Jackson, J.K. and Connor, R. The Skid Row Alcoholic. Quart. J. Stud. Alc. 14: 468-4861 1953. 9. McCormick, T.C. Toward casual analysis in the prediction of attributes. Amer. Sociol. Rev. 17: 35-44, 1952.