TO ALCOHOL SOCIOLOGY HOME PAGE
Citation: Roizen, Ron, "The Great Controlled-Drinking Controversy," pp. 245-279 [Chapter 9] in Marc Galanter (ed.), Recent Developments in Alcoholism, Vol. 5, New York: Plenum, 1987.
The Great Controlled-Drinking Controversy
Ron Roizen Alcohol Research Croup, Institute of Epidemiology and Behavioral Medicine, Medical Research Institute of San Francisco, 1816 Scenic Avenue, Berkeley, California 94709. It should be noted that the author had a small role in the debate this review examines, having served as one of the prepublication reviewers of the first Rand Report in 1975. His published commentary, on that report has been cited particularly by the pro-controlled-drinking side.
Abstract. This chapter reviews the controlled-drinking controversy. It presents cameo descriptions of the controversy's three major episodes--those occasioned by D. L. Davies' 1962 report, the 1976 publication of the first Rand Report, and the 1982 publication in Science of a paper by Pendery, Maltzman, and West--as well as a cameo for the long "interepisode" period between Davies' paper and the Rand Report. I argue that the controversy has emerged out of the failure of the "new scientific approach" to alcoholism, initiated a half century ago, to advance alcoholism treatment significantly beyond the point from which it began. Lack of progress, in turn, has generated tensions and reverberations along many of the normative dimensions that define scientific/treatment activity, both internally and in relation to the broader society. Some of the changing social and valuative forces at work in the controversy's history are examined.
The controlled-drinking question has provided the best-known controversy in the alcohol research and treatment world for a decade or more. By most accounts, it began in 1962 with British psychiatrist D. L. Davies' report1 of seven out of 93 alcoholism patients at the Maudsley Hospital who, at followup, were drinking in a controlled manner. Davies' discovery ran counter the traditional view, then almost universally held among treaters of alcoholism, that the alcoholic must abstain entirely and permanently from alcohol. Though a growing research literature--and a growing tension--accumulated around the controlled drinking topic between 1962 and 1976, this "interepisode" period saw no single, prominent outbreak ot controversy. In 1976, however, a second major episode erupted. The first Rand Report reported finding "normal drinkers" among a sample of American alcoholics who had received treatment at federally sponsored treatment centers at various sites around the country. A third major episode began in 1982, with the attack by Pendery et al.3 on treatment outcome reports by behaviorist psychologists Mark and Linda Sobell. Lesser battles have occurred as well, for instance, surrounding the second Rand Report,4 the Eagleville Conference,5 and, most recently, a follow-up study reported by Helzer et al.6 in the New England Journal of Medicine.
The controversy has occasioned publication of hundreds of substantive research reports, commentaries, and literature reviews--the order of magnitude of which is suggested in Heather and Robertson's7 extensive bibliography, numbering about 500 entries. But despite all the effort focused on it, the gap dividing the warring sides has not narrowed. Indeed, the controversy has grown fiercer and engaged a broader audience over the course of these three major episodes. Even the question of which side is winning is disputed. A number of controlled-drinking advocates--e.g., Miller,8 Marlatt,9 Heather and Robertson,7 and Sobell and Sobell10--have claimed a measure of victory for their side, arguing that controlled-drinking approaches have finally won a permanent place in the spectrum of American alcohol problem treatments. Some voices on the other side, however, see the Pendery/Sobell controversy as marking the effective death of controlled-drinking treatment approaches in this country.
It can fairly be said that this was a controversy that was never supposed to happen. The history of the past half century of American thought on alcohol-related problems has been predominantly that of the so-called "modern alcoholism movement." This movement, which began in the United States in the late 1930s, promised a new scientific and humanistic approach to alcohol problems, one replacing the passionate, moralistic, and legislative social controversies ot the hundred-year temperance-prohibition-repeal eras with the dispassionate, knowledge-cumulating, and benevolent approach of modern science and medical-style treatment. With the end of national prohibition in the United States, the citizenry was exhausted and fed up with social controversy over alcohol. Thus, one of the main arguments advanced by early advocates of this new scientific approach was that it would bring an end to protracted social controversy over alcohol.
A 1938 announcement of the formation of the Research Council on Problems of Alcohol (an organization chrysalis for the new scientific approach) published in Science magazine11 reported that this council would undertake a "thorough, unbiased and strictly scientific investigation of the problems related to the control of alcoholic beverages and to seek solutions through a program of unprejudiced research and education." Science, it was hoped, would remove the need for controversy, substituting a growing understanding in its place.The scientist believes that if an adequate body of factual data is made available and discussed fully, without heat or prejudice by men and women of intelligence, integrity and leadership, the right solution . . . may then be evolved (p.330).An outline for the Research Council's program, appearing in the then-infant Quarterly Journal of Studies on Alcohol,12 likened the current American perspective on alcohol problems to the country's prescientific perspectives on tuberculosis, syphilis, and cancer. Science, it was argued, had made great advances with respect to these problems, in large measure overcoming fears and prejudices associated with them. In due course, the text argued, science would also "prevail in respect to the problems of alcohol" (p.434). On the whole, the modern alcoholism movement can probably be judged a considerable success with respect to this controversy-reducing goal. Though the controlled-drinking controversy we will examine has been a heated and protracted one, and though it has spilled into the public press, nevertheless its societal scope and magnitude are tiny in comparison with the societal attention paid to alcohol issues in the hundred years prior to repeal in 1934. If nothing else, then, the new scientific movement has managed to shrink and compartmentalize social conflict over alcohol, making such controversy the province of a particular scientific and treatment community--though, of course, such shrinkage might have occurred in any case.
Yet, if we restrict our gaze only to this small alcohol community, the controlled-drinking controversy's protracted history can be regarded as a remarkable and intriguing puzzle. A cardinal rule of science, after all, is that meaningful controversies--controversies worth having--must concern matters that can ultimately be referred to nature or to empirical observations for resolution. By all rights, the controlled-drinking controversy seems one that ought to have been resolved in this way. Moreover, its 20-year-plus history and the growth of alcohol-related research over the period suggest that sufficient time and effort have been provided for its resolution. Why, then, has it persisted so long and so bitterly? My sense is that this dispute represents--now, a half century after the launching of the "new scientific approach" to alcohol-related problems--a kind of marker for the failure of science to advance the treatment of alcoholism significantly beyond the point from which treatment began at the outset of the movement.13,14 Put differently, it is unlikely that such a conflict would have occurred had the research and treatment effort devoted to alcohol problems in fact resulted in sure-fire treatment approaches.
In turn, this failure has placed considerable strain on the scientific/treatment institutions and their underlying value systems. Faced with an intractable social and scientific problem, that strain has illuminated both value conflicts in the internal norms defining scientific/treatment activity and important lines of connection and strain between scientific/treatment values on the one hand and the values and concerns of the broader society on the other. This chapter presents cameo descriptions of the controversy's three major episodes and the long interepisode period (1962-1976). Its goal is to provide a sketch of some of the changing social and valuative forces evidenced in the controversy's evolution.
2. Episode One: D. L. Davies' 1962 Paper
Though the controlled-drinking controversy is commonly dated to Davies' 1962 paper, if we recall that the proposition that alcoholics must abstain has a very long history in our society, stretching back as far as the beginning of the 19th century,15 it follows that the controversial question has been fought out countless times in the private lives of skeptical alcoholics long before Davies appeared on the scene. What was new in Davies' paper was that it lent a measure of medical, scientific, or official support to the possibility of controlled drinking. Davies' findings were not, however, entirely new. He himself cited similar reports by Lemere,16 Shea,17 and Norvig and Nielsen,18 Heather and Robertson,5 whose book Controlled Drinking is the most comprehensive source on the subject, cite five additional pre-1962 reports.19-23 In hindsight, however, it is not surprising that none of these sparked controversy. In most, the controlled-drinking findings were incidental and were not used to attack the abstinence requirement. Indeed, some of these early reports, read in ordinary day light, seem hardly to support the controlled-drinking idea at all. Their places in the history of the controlled-drinking controversy probably owe more to the thoroughness of the literature searches carried out by subsequent controlled-drinking advocates than to the intentions of their original authors.
Lemere,16 for example, drew his data from a 6-year inventory of the recollections of his psychiatric patients concerning the life histories of their "grandparents, parents, aunts, and uncles," finding that 3% ultimately became normal drinkers. As Kendall and Staton24 pointed out, such data were often based on long-past events, and the alcoholism diagnoses involved were unreliable; additionally, the reported proportion of normal-drinking outcomes was quite small. Shea17 reported on a single controlled-drinking patient and fashioned his case report into strong support for the abstinence treatment goal. "It is not the purpose of the present paper to challenge the truth of the general opinion [favoring abstinence]," (p. 595) he wrote. Shea concluded with three points: (1) alcoholism "must be tackled directly" and not merely as a symptom of underlying neurosis; (2) complete sobriety should be sought; and (3) psychotherapists should not "succumb to the temptation of therapeutic overambitiousness"--in other words, they should accept abstinence as a meaningful treatment goal (p.604). These were conclusions hardly likely to upset abstinence advocates. Pfeffer and Berger19 and Moore and Ramseur20 employed the clinical tradition of reporting "improved" as well as "abstinent" cases, from which their controlled-drinking findings derived. Neither paper gave more than passing attention to the salient findings, and both were subject to criticism for inadequate description of controlled-drinking outcomes. Though Selzer and Holloway11 asserted both that they had discovered moderate drinkers in a posttreatment group of alcoholics (13 of 83) and that this finding challenged conventional wisdom about abstinence, the main thrust of their article was on the worth of involuntary treatment for alcoholics in a state mental hospital setting. The controlled-drinking discussion amounted to roughly one twentieth of their text's length.
Davies' 1962 article,1 then, drew its notoriety from its direct attack on the classical wisdom. Davies had not, however, argued that the abstinence treatment goal should be modified. Indeed, he noted the contrary in his conclusion: "It is not denied that the majority of alcohol addicts are incapable of achieving 'normal drinking.' All patients should be told to aim at total abstinence" (p. 103). Instead, Davies questioned the classical wisdom underlying the abstinence requirement. Interestingly, Davies' paper did not attribute the paradigmatic basis for the abstinence requirement to Alcoholics Anonymous or Jellinek but instead to contemporary American psychiatric thought, citing Zwerling and Rosenbaum's 1959 article on alcohol addiction in the American Handbook of Psychiatry25 along with the similar views expressed by a number of other psychiatrists who had written on alcoholism. It is also notable that Davies regarded his controlled-drinking subjects as challenging not only the notion of alcoholism's biochemical irreversibility (as put forward by Williams26) but the separate notion of its psychological irreversibility as well (as put forward by Hoff27). Davies' objective in this paper was to encourage a broader, multicausal view of alcoholism's etiology.
On the whole, the Davies paper generated an episode less heated than we may remember it. Davies had initially offered his article to The Lancet, the British medical journal, which turned it down for lack of general interest.28 The American Quarterly Journal of Studies on Alcohol elected to publish it and in due course published 18 commentaries29-46 as well, perhaps suggesting considerable interest. Edwards,47 however, has rightly pointed out that the thrust of these commentaries was often "dismissive" in character. Commentators tended to look on Davies' finding as curious and thought provoking but, in the end, of little clinical value. Fox,29 the commentator representing the National Council on Alcoholism (a vigorous institutional proponent of abstinence in later historical episodes of the controversy), offered only a brief and bland response. One source of tepidity was that the responses were all authored by physicians/psychiatrists, not a profession whose interests were necessarily vitally challenged by Davies' findings. Nor did the episode draw the sort of attention from the popular press that would mark subsequent episodes. The New York Times indexes for 1962 and 1963 listed a total of 34 articles under the heading "alcoholism," but none concerned Davies' paper or the comments on it appearing in the Quarterly Journal of Studies on Alcohol.
Aside from their commitment to abstinence, there was little consensus across the commentaries on Davies. What most characterized this episode's debate was its scholarly content and tone. Commentators appeared to be secure in their commitment to the abstinence requirement and used their rhetorical opportunities to entrench it further. The commentaries were rich in conceptual content. Indeed, careful reading of them reveals that as a group they managed to touch on a very large proportion of the conceptual issues that would recur over and over again in the debate's full historical course.
One of the safest lines of defense for the abstinence requirement lay in questioning whether Davies' cases were true alcoholics. Many commentators considered Davies' diagnostic accuracy and his choice of diagnostic criteria, though their criticisms differed. Block,30 for example, argued that six of Davies' seven case descriptions did not include evidence of withdrawal symptoms and therefore were not true alcoholics. Lemere31 suggested that evidence of controlled drinking ought to imply diagnostic exclusion from alcoholism even if such diagnostic judgments had to be made post facto. Maidman32 argued that Davies' cases were not true loss-of-control alcoholics but "continuous or steady drinkers." Williams33 criticized Davies for having apparently diagnosed alcoholism on the basis of only a single patient characteristic, loss of control, and argued that four of the seven cases were better described as heavy drinkers, not as true alcoholics. Smith34 and Brunner-Orne35 suggested that Davies' case might fall along the hazy borderline between alcoholism and nonalcoholism. Lolli36 and Esser37 said they could not be sure of the diagnoses, arguing that the subjects were better described as "severe habitual excessive drinkers."
Kjolstad,35 on the other hand, conceded accurate diagnosis to six subjects, and Bell39 thought all seven had been accurately diagnosed. Armstrong,40 Bell,39 and Block30 noted that Davies' alcoholics were Englishmen and that there might be relevant cultural differences in alcoholism. Concerning Davies' choice of diagnostic criteria, Block30 criticized Davies for employing the World Health Organization (WHO) criteria for alcoholism. This definition, he wrote29 (p. 114), "does not define a true addiction." Esser37 argued that the WHO criteria were too broad and, echoing Jellinek,48 appealed that "the disease conception of alcoholism must not be extended to all forms of excessive drinking." Brunner-Orne35 (p. 732) suggested that Davies' cases might be "danglers . . . swaying around the hypothetical demarcation line between alcoholics and nonalcoholics." Smith34 proposed that there were "degrees" of alcoholism, implying a continuum rather than a dichotomy in the phenomenon. SeIzer41 thought the appearance of controlled-drinking patients suggested that there might be two alcoholisms rather than only one.
A number of commentators broadened their diagnostic comments to include the observation that all currently available diagnostic procedures were likely to misdiagnose some nonalcoholics as alcoholics. Block,30 for example, argued that it was not uncommon to have people drink very heavily over long periods of time and give every indication of being alcoholic without being true addicts" (p. 116). Maidman32 described his conception of "pseudoalcoholism," which, he wrote, may very closely resemble the real thing. Many pseudoalcoholics, Maidmen contended, "present themselves for treatment, either voluntarily or under coercion by family, employers, the law, etc. Some may even require long-term hospitalization and rehabilitation"25 (p. 733). Smith34 (p. 323) reported seeing two patients with delirium tremens whom he nevertheless did not consider alcoholics. Although such comments were not morally costly, they trod on thin ice with respect to the traditional mode of suasive alcoholism treatment. Davies' controlled-drinking challenge was necessarily premised on the assumption that alcoholism could be accurately diagnosed. Without that premise, controlled-drinking cases could be dismissed simply as the result of poor diagnostic criteria, and the search for better diagnostic criteria begun. However good the logic of this, this rhetorical direction harbored a perilous treatment implication: abandoning belief in alcoholism's accurate diagnosability in principle would fundamentally undercut the suasional stance of the therapist. The active alcoholic--someone who was thought to employ any means available to deny his alcoholism and continue drinking--could no longer be confronted with an absolutely certain diagnosis and its equally certain abstinence imperative. Lemere31 (pp. 727-728) very clearly saw this implication looming and made it the basis for his plea that controlled drinkers be excluded, post facto if necessary, from the alcoholism diagnosis:these few "exceptions to the rule" cloud the clarity of our definition of alcoholism and our exactness of treatment, for they occasionally make liars out of us when we tell a patient he is an alcoholic and can never drink again. . . . Those few patients who run counter to this rule should be classified as pseudo-alcoholics. Granted the diagnosis of pseudo-alcoholism may be difficult to make before the fact, it can be made afterward, thus preserving our basic concept of an irreversible process. This is the only way we can stick to our guns that a true alcoholic can never drink again. If we ever concede this dictum, we, as therapists, and our alcoholic patients, will be lost (pp. 727~728, italics added).Lemere's3l concern was one of several in the commentaries in which the interests of treatment and the interests of research were seen as at odds. Several commentators, for example, were concerned over the general danger of Davies' findings for alcoholic patients.39,40,42 Bell39 (p. 322), for example, wrote that "for every alcohol addict who may succeed in reestablishing a pattern of controlled drinking, perhaps a dozen would kill themselves in trying." Bell39 suggested that future research into controlled drinking "should be carried on with a minimum publicity" until potential controlled-drinker alcoholics could be clinically differentiated from abstinence-requiring alcoholics. A number of other commentators31,40,41 noted that as long as alcoholics with the potential to control-drink could not be differentiated from alcoholics requiring abstinence, Davies' findings could not be acted on. Brunner-Orne,35 however, suggested that a controlled-drinking treatment goal might be experimented with for two groups of patients, those who had failed to abstain in spite of favorable family and social conditions and those who were not really alcoholics in the first place.
Several commentators tinkered with conventional wisdom about alcoholism in order to better defend it. Such efforts also raised potential threats to alcoholism's tradition of suasive treatment and its conceptual underpinnings. Some addressed the loss-of-control element of alcoholism, namely, that the ingestion of alcohol by an alcoholic always set in motion a chain reaction leading the alcoholic to uncontrolled drinking. Several noted that true loss-of-control alcoholics did not necessarily get drunk every time they had a single drink. Glatt43 observed that inadvertent ingestion of alcohol by alcoholics did not inevitably lead to binges; Lolli36 noted the same for alcohol taken in "homeopathic doses." Esser37 said that even true addicts can control their consumption sometimes. Glatt,43 Kjolstad,35 and Bell39 suggested that there might be a "threshold effect" in alcoholism--only above a certain blood alcohol level would the loss of control be triggered. Kjolstad35 and Lolli36 used the term "nibbling" to describe this sort of subthreshold drinking. The threshold drinking notion also tended to raise the question of whether Davies' subjects should be regarded as "normal" drinkers simply bv virtue of the fact they drank moderately. Kjolstad38 argued that the definition of "normal" drinking should be that the drinker knows he can stop drinking no matter how much he has already drunk. This definition excluded Davies' cases who, it was argued, only maintained control because they kept their consumption levels below their thresholds and needed to be ever mindful of their consumption. Brunner-Orne35 also argued that treatment could never achieve truly normal drinking for alcoholics.
Glatt43 (p.116) modified the traditional paradigm by suggesting that "'loss of control' does not mean, as often wrongly assumed, that an alcoholic will invariably end up drunk on each and every occasion when he takes a drink, but rather that he can never be sure of not ending up drunk." Such a reformulation, however, harbored both clinical and conceptual difficulties. On a clinical level, it once again threatened to undercut the message usually delivered to alcoholics in treatment. On a conceptual level, it raised the fundamental question of the ultimate seat of alcoholism's causation. If, after all, the loss-of-control phenomenon was conceptually lodged at a physiological or biochemical level, then why did loss-of-control not consistently flow from alcohol's ingestion? An episodic or situational character for loss of control implied that other sorts of causal levels must be at work too. This sort of argument for alcoholism's multicausal character lay at the center of Davies' original paper.
The biochemical level of explanation might be preserved by arguing that Davies' seven alcoholics had somehow undergone a salutary change in body chemistry, either because of individual idiosyncracy or years of intervening abstinence. "Body chemistry," wrote Block30 (p. 115), "is such that it can change at any time. It is not rare to have spontaneous recoveries from very severe illnesses without the medical profession knowing the cause either of the illness or of its resolution." Block30 and Smith34 likened controlled drinking in alcoholics to spontaneous remissions from cancer, a rhetorical equation that served to defend the biochemical level of explanation by regarding such drinking as aberrational--freak anomalies without either theoretical or clinical implications. Thimann42 reported one controlled drinking alcoholic out of 25,000 seen at the Washington Hospital in Boston. Like cancer remissions, Block30 (p.116) wrote, "they are unique, they are different, and they are rare."
Tiebout,44 Smith,34 and Brunner-Orne,35 on the other hand, considered the possibility that years of abstinence might change body chemistry, eventually restoring some alcoholics' ability to control consumption. Brunner-Orne35 even suggested the novel idea that the future prospect of such chemical repair might encourage some alcoholics to commit themselves to longer periods of abstention. But if abstinence might account for newly controlled drinking, thus preserving the place of chemistry in alcoholism's causation, it also opened the door to the possibility that renewed drinking would in time decay body chemistry back to its former alcoholic character. Smith34 (p.323), for example, conceded that he had ". . . seen quite a number [of alcoholics] who, after an extended period of abstinence,, could partially control their intake for as long as a month; but their efforts at social drinking were never successful, and they always ended with another alcoholic bout." Esser37 (p. 121) wrote nearly the same judgment: "Once in a while a patient was able to control his drinking during a few weeks, but after that time he always relapsed." Lolli,36 who scorned the biochemical model, nevertheless noted that nibble-drinking eventually resulted in uncontrollable drinking episodes (p. 329). In short, there were several bases for anticipating that controlled-drinking alcoholics over time would decay into their former alcoholismic drinking (or even abstinence). Their abandonment of controlled drinking, then, might ultimately lend support to the notion that a biochemical difference had, after all, underlain their alcoholism. The problem of decay in control raised the enduring dilemma of the role of time in the debate. How long, after all, was long enough to call someone a controlled drinker? How would one know that a controlled drinking case would not later relapse into uncontrolled drinking?
Davies' commentators were prone to view alcoholism through the medium of their own alcoholic patients and evaluate Davies' report against the context of their own clinical experience. There was nearly no comment in their narratives on the circumstances or the "problem" of untreated alcoholics. Many commentators cited negative experience. Zwerling45 (p. 117), for example, wrote, "I have never myself met an alcoholic patient who then went on to become a social drinker." Fox29 (p.117) wrote, "My own practice covers many hundreds of alcoholics, and though I have never been in the position to do a follow-up, I do not know of a single patient of mine who has been able to resume normal drinking." Esser37 (p. 121) wrote that "in the last 15 years I have never seen one single 'true addict' who, after treatment, was able to return to regular normal drinking." Williams33 (pp.112-113) and Smith34 (p.323) made similar assertions. Such comments began a tradition of clinical-experience refutations by abstinence advocates that spanned the entire history of the controversy.
Clinical experience, however, was not always given the same weight and not always similar. Brunner-Orne35 and Fox,29 for example, noted that the practicing therapist might be in a poor position to observe controlled-drinking alcoholics, should such indeed exist. Brunner-Orne35 (p.731) wrote that the patients who may have "succeeded in drinking moderately are those we don't hear from." This stance favored more and better follow-up research. A number of commentators echoed Davies' report with controlled-drinking cases drawn from their own patients. Armstrong40 (p.118), for example, wrote, "I suppose any ot us who have seen many hundreds of cases can recall a few instances in which this phenomenon of return to social drinking has been demonstrated for a sufficiently long time to refute the generally held belief." Myerson46 (p. 325) wrote, "In my own clinical practice, I have found an occasional alcoholic who, under certain circumstances, controls his drinking for prolonged periods of time." Thimann42 (p.324) reported a single controlled-drinking case. Glatt43 (p.116) wrote that an occasional alcoholic can control his drinking when the situation is right. Kjolstad35 (p.729), writing from his Norwegian experience, suggested "that most alcoholics can take a pint of mild beer or a corresponding amount of alcohol, and not lose control"--the problem, he said, was that most alcoholics did not desire to drink in these small quantities (see also Smith,34 p.323). These commentators defended abstinence elsewhere on the rhetorical battlefield.
A few commentators raised the explicitly normative issue of whether Davies' controlled-drinking treatment goal had inherent value. Block30 (p. 115), for example, puzzled over the purpose behind attempting to prove that alcoholics could learn to drink moderately:Were alcohol a necessity of life, I would say that this might be an objective to be sought in order to give the individual something which is necessary to continue living. However, since alcohol cannot be considered a necessity, why attempt to prove that certain individuals diagnosed as alcoholics can return to drinking moderately?Smith34 (p.324) made much the same point by likening alcohol for alcoholics to penicillin for patients with an anaphylactoid reaction to the drug: "Certainly the possibility exists he will not have another [reaction] if he is given more penicillin--but then, of course, he may die, which is rather a drastic way of establishing his intolerance." Some noted that alcoholism's progressive character made abstinence the desirable treatment goal for alcoholics and non-alcoholics alike (e.g., Block,30 p.116; Maidman,32 p.734). Bell39 (p.321) cited a cultural dimension to controlled drinking's unacceptability as a treatment goal, noting that "in this country, where most people are conditioned to believe that controlled drinking for the alcohol addict is impossible, most attempts to drink would precipitate sufficient turmoil to assure that uncontrolled drinking would develop."
Two commentators, Selzer41 and Meyerson,46 differed from the rest by criticizing advocates of the traditional abstinence requirement. Their remarks concerned the social conduct of abstinence advocates and anticipated a plane of discourse that would become more frequently employed in later episodes. Selzer,41 for example, recalled his own controlled-drinking research21 and the negative response it had received. He wrote that "the data prompted the agency that provided funds for the study virtually to order us to omit these 'embarrassing' findings. At a subsequent psychiatric meeting where the paper was presented, a prominent colleague arose to state that the 13 subjects were probably not alcoholics. Similar episodes followed"41 (p.113). Selzer argued that abstinence advocates were much threatened by and hostile toward controlled-drinking findings. He suggested two possible sources for this: first, therapists might fear that controlled-drinking findings would "wreak havoc" with their treatment programs, and, second,many people working in this field are alcoholics themselves and are compelled to remain abstinent. It may be especially difficult for the alcoholic who must remain dry to accept the idea that others can recover and drink socially. To hear of the "success" of others may be frustrating--and those workers prefer not to hear about it since it also upsets their treatment concepts41 (p. 113).Myerson's46 comments had a similar ring. He argued that it had been "refreshing" to see the abstinence idea challenged, and he criticized "certain groups who deal with alcoholics" whotend to adhere to a rigid doctrine which they feel should be applied to the entire alcoholic population. They tend to preach rather than practice. They tend to respond with indignation if any of their ideas are challenged46 (p.325).Myerson called for renewed empiricism in the question of alcoholism--he found Davies' article useful, he wrote, "not so much from a clinical point of view but in the sense that it brings the problem of treatment down to clinical observations, which is where it always should be" (p.325). Both Selzer41 and Myerson,46 then, suggested that abstinence advocates did not play by the rules, were irrational, unfair, or laid claim to too much turf.*
* Edwards49 has recently published a critical report of Davies'1 seven controlled-drinking cases, arguing that five of them experienced "significant drinking problems both during Davies's original follow-up period and subsequently . . . and that the remaining two (one of whom was never severely dependent on alcohol) engaged in trouble-free drinking over the total period"49 p. 181).
3. The Interepisode Period (1962-1976): Controlled Drinking on the Offensive
Though the interepisode period between Davies' report in 1962 and the publication of the Rand Report2 in 1976 saw no major outbreak of controversy, an increasing tension grew up around the issue and the character of the debate shifted. This was a period of great activity, growth, and change for the alcohol field--changes that would alter both the social and scientific contexts of the debate. It saw the rise of a greatly increased U.S. alcoholism research and treatment establishment fed by a shift in public sentiment, by the creation of a National Institute on Alcohol Abuse and Alcoholism (in 1970), and by an expansive trend in federal government spending on social issues.9 Institutional build-up raised the material stakes surrounding the controlled-drinking issue. Growth also served to recruit a new class of therapists and researchers into the alcoholism field, one heavily infused with young professionals more or less detached from the alcoholism world and its conceptual commitments.50 They found places for airing their new views in an increasing number of journals servicing the alcohol field.
The controlled-drinking position established itself and built up a self-conscious constituency. Whereas Davies' patients had not been directed toward a controlled-drinking goal, the post-1962 period saw treatment specialists--especially behaviorist psychologists--attempting intentionally to train alcoholics into moderation. The controlled-drinking issue became something of a shibboleth dividing many new alcohologists from the older, more traditional treatment establishment. Traditional treatment interests grew too, and perhaps even at a greater rate. The treatment establishment that grew up in the United States in the early 1970s both built on and built up a great corps of traditionally oriented treatment paraprofessionals. Alcoholics Anonymous enjoyed substantial growth over this period, and the National Council on Alcoholism--the public relations wing of the old movement--maintained its considerable influence in national alcoholism affairs. A growing list of alcoholism treatment organizations51 often sided with the traditional abstinence requirement.
In the controversy itself, controlled-drinking advocates went on the offensive. Some of their argumentation addressed various abstinence defenses raised earlier in the Davies' episode. For example, over and against the negative clinical experience of proabstentionist clinicians, controlled-drinking advocates cited an accumulating body of supportive research. By 1974, Sobell and Sobell52 could count "60-plus" favorable controlled-drinking reports in the literature; 2 years later, Pattison53 described the figure as "now close to 100." Advocates pointed to the diverse research quarters from which supporting observations had come: from clinical follow-up studies (like those reviewed by Emrick54), from general population surveys (for example, Bailey and Stewart55), from the epidemiology of admission statistics,56 from social-psychological laboratory studies critical of the loss-of-control phenomenon,57 and from the successful treatment experience of behaviorist psychologists.55 (Several landmarks of this literature were conveniently collected by Pattison et al.59).
Controlled-drinking advocates also argued--as some of Davies' commentators had in 1963--that proabstinence clinical experience was badly biased by its observational position.60 Similarly, whereas proabstentionists had invoked alcoholism's progressive character to justify the abstinence requirement regardless of the patient's precise diagnostic status, controlled-drinking advocates argued that alcoholism's inevitable progressiveness was far from proved. Traditional wisdom was reexamined, and it was noted that Jellinek,45 himself, had allowed that a problem drinker might simply stop at one or another point along the way in alcoholism's famous phases. Widespread belief in progression was, once again, attributed to the poor observational position of clinicians. Prospective studies of untreated populations also argued against progression's inevitability.61 As Straus62 noted, data coming from new epidemiologic research posed significant challenges to the notion of inevitable progression in revealing both (1) much movement into and out of drinking-related problems in general population samples and (2) a concentration of drinking-related problems among younger men aged 21-24 rather than the older age groups commonly found in treatment samples--suggesting a good deal of natural remission in the phenomenon from youth to middle age.
Pro-controlled-drinking argumentation often took new directions reflecting important changes in the alcoholism treatment world. For example, Pattison's63 early critique ot abstinence, published in 1966, reflected the increasing importance of research in the alcoholism field. Pattison63 argued that the use of abstinence as the sole outcome measure in treatment evaluation research was too narrow and limiting. Alcoholics represented a diverse collection of patients, he wrote, and an equally diverse array of possible treatment outcomes. Treatment might serve to "improve" the patient's general wellbeing without necessarily achieving a complete abstinence. Though the quantity of treatment research was increasing, he argued, nevertheless, the absence of uniform outcome measures limited its value. The wide range of follow-up studies, Pattison63 (p.49) wrote, "has been so heterogenous that few reliable conclusions can be drawn about either treatment methods or results." Useful research would demand both uniform and multidimensional treatment outcome measures. In arguing for multidimensional measures, Pattison63 drew attention to the basic question of the relationship between an alcoholic's drinking or nondrinking behavior and his general well-being. This focus reraised an old conflict between traditional psychodynamic and traditional addiction models of deviant drinking.*
* In the psychodynamic model, deviant drinking could be regarded as a symptom of underlying psychopathology; thus, merely imposing abstinence on the alcoholic would not necessarily imply' improvement in his overall well-being--indeed, abstinence might harm the patient by depriving him of an important means of self-medication. This was the old "alcoholism as symptom" versus "alcoholism as illness" debate resurrected in a new form.
Contemporary interest was reflected in the publication of a number of reviews of the controlled-drinking debate in this period.65-70 These, in turn, provide convenient sketches of the debate's major rhetorical tendencies. Pattison's call for broader outcome measures and for a more diversified image of the alcoholic, for example, was joined by a variety of attacks on the abstinence treatment tradition. The most important of these was the claim that abstinence-oriented treatment had simply failed. The measure of this failure, it was argued, resided in the growing conviction that few alcoholics actually accepted the abstinence requirement.65-69 The magnitude of its failure was great--it included all alcoholics who never presented themselves for treatment, all alcoholics who rejected the abstinence requirement once in treatment, and all alcoholics who had attempted abstinence but abandoned it sometime after treatment.60,65 The value of a controlled-drinking goal was also tied to the argument that social drinking was conventional behavior in contemporary U.S. society. Thus, the alcoholic was confronted with numerous invitations and pressures to drink,60,70 and abstinence-oriented treatment might actually serve to heighten the stresses surrounding the alcoholic and promote excessive drinking by making sociable situations more stressful and by heightening the fear of the failure represented by taking a drink.71
Miller,69 citing Emrick's review,54 noted that in fact most patients improve rather than abstain as a result of treatment, and treatment goals might best be modified in that light. Behaviorists argued that merely demanding that the alcoholic abstain did little or nothing to help him achieve that goal. Pattison,65 Evans,68 and Verden and Shatterly72 argued that the abstinence tradition inhibited progress in the field by stifling research and experimentation. Hill and Blane73 criticized it for requiring a standard of conduct not required of normal drinkers, and Evans68 faulted the occasional clinical practice of punishing alcoholics for their drinking episodes, noting that medicine does not usually punish patients for showing symptoms of their disease.
Controlled-drinking advocates and reviewers often cited the ability of their approach to attract reluctant alcoholics into treatment.60,65-67,69,70 Controlled drinking, it was argued, would provide a more appropriate treatment goal for less impaired,65,67 younger,69 or prealcoholic69 drinkers. More alcoholics and problem drinkers would be attracted to treatment sooner in the course of their illnesses.*
* It also bears noting that controlled drinking was occasionally criticized because it might lead alcoholics to avoid rather than seek treatment.65,74 Untreated alcoholics, this argument went, might see controlled-drinking programs as signs that they could treat themselves, thus enhancing their denial of the need for treatment. Somewhere between the two poles of this attract-avoid dimension lay Miller's69 contention that alcoholism treatment might indeed be self-taught in self-help treatment manuals (see, for example, Miller and Munoz75).
This emphasis on the failure of abstinence-oriented treatment would have seemed quite foreign to most of Davies' original commentators. Their attention had been directed mostly to alcoholics who had already achieved abstinence, whereas the eyes of controlled-drinking advocates were fixed on a vast army of alcoholics who had not benefited from the traditional approach. This shift in emphasis reflected a number of important transitions in the alcoholism treatment world. Growth in the U.S. treatment establishment not only made more treatment possible, it changed the traditional relationship between the alcoholic and his treatment source. In the pre-1962 era, slender treatment resources obliged alcoholics themselves to seek help from AA, psychiatry, general medicine, or elsewhere. With the rise of a large, alcoholism-specific treatment establishment, however, the alcoholic was transformed into a would-be treatment recipient-one who has actively recruited into treatment and for whom something should be done once he was recruited. But what?
Outside the conceptual framework proffered by the Alcoholics Anonymous model of alcoholism, there was no "technical fix"--no specific, effective treatment for alcoholism available. A dearth of effective treatments made for the appearance of a great many treatment aspirants. Over time, more and more Americans became aware of the availability of the Alcoholics Anonymous program. This meant that public treatment institutions were often faced with alcoholics and problem drinkers who had by now either rejected or failed AA's methods. A patchwork of low-cost treatment setting--many involving paraprofessional therapists, themselves committed AA members and advocates-grew up in the United States in the 1960s and 1970s. They faced the difficult task of carrying the AA approach into a new treatment-dispensing context. The new corps of professional therapists faced the equally difficult task of mediating the dominant AA treatment ideology and resistant populations of patients.
The interepisode period also witnessed rapid expansion in alcoholism prevalence estimates--expansion that further emphasized and reflected the importance of the untreated alcoholic in public policy. Between 1962 and 1976, alcoholism's estimated prevalence virtually doubled--from 5 million to roughly 10 million76--while the U.S. population in comparison grew by only 20%. The change reflected a deeper shift in American attitudes toward alcohol-related problems and government involvement in them. By the early 1960s a quarter century had passed since Repeal, and an older generation of Americans--one with a deep aversion to government involvement in alcohol-related issues born out of the nation's prohibition experience--had waned, lost influence, or died off. Their passage, in turn, quietly dissolved formerly well-entrenched inhibitions in the rhetorical practices of early advocates of the modern alcoholism movement. Early advocates, as Keller77 has noted, tended to play down the scope and severity of alcohol-related problems in the country. Fiery and problem-magnifying claims would have run against the grain of the early movement's efforts to bring calm, scientific rationality to the territory and avoid identification with Drys. Alcohol's role in a variety of health and social problems was minimized,77 and even estimates of alcoholism's prevalence were kept modest. Jellinek's 1947 paper78 on then-recent American trends in alcohol consumption and alcoholism, for example, argued that although U.S. consumption had undergone a sharp increase between 1940 and 1945, American drinkers were becoming more, rather than less, moderate in their drinking practices, and alcoholism was probably not on the increase.
By the early 1960s a reversal in this direction of emphasis was underway. Later in the decade a new corps of alcoholism advocates preferred a rhetoric of problem magnification in place of minimization. Not only the increasing prevalence estimates but new methods for prevalence calculation hd important consequences for the alcohol field. Between 1947 and 1959, the "Jellinek formula" had been the preferred basis for calculating the prevalence of alcoholism in the country. This formula used cirrhosis mortality as the basis for its projection. By 1959, however, the formula had fallen on hard times, and even Jellinek79 called for its abandonment. Hence, by the late 1960s the burden of prevalence estimation was taken up increasingly by general population surveys of drinking practices and problems. Data from these surveys, however, revealed that the distribution of alcohol-related problems in the general population did not sort itself into two clear-cut populations, one problem-free and the other alcoholic. Instead, drinking-related problems appeared to be continuously and unimodallv distributed. Where. the cutting point between normal drinking and alcoholism should be drawn became the crucial question for prevalence estimation. That choice, moreover, in the absence of better knowledge, would have to be largely arbitrary, and prevalence estimates in turn could be made either very large or vanishingly small depending on how it was made.80 This was a weakness that could be exploited by parties desiring to amplify the alcoholism problem in the country.
Over this period, the country's alcohol-related problems became more broadly defined. The early alcoholism movement had deliberately restricted its attentions only to that segment of the problem-drinking spectrum whose alcohol-related problems could be said to be "essentially" alcoholic in nature. A restricted focus served a number of important rhetorical and institutional ends. It conformed to the general strategy of problem minimization by limiting the scope of the movement's efforts. It promised more effective scientific research by narrowing and more sharply defining the problem span it addressed. It also emphasized a hard and fast distinction between the true alcoholic (whose drinking behavior was involuntary) and the voluntary drunkard (whose behavior, it was thought, should remain under the control of conventional institutions of social control48). Such a distinction addressed the reservations of citizens who were skeptical of the "new scientific approach" to alcoholism and feared that treatment would be indiscriminately extended to those who did not "deserve" it. A restricted focus also served the new movement's political needs by helping to bar new alcoholism treatment specialists from encroaching on the territories of other, already existing treatment and social assistance institutions.81
The emerging conditions of the 1960s and 1970s reversed each of these utilities. The felt need for a clear distinction between true alcoholism and problem drinking (stemming from psychological or other nonalcohol sources) waned in importance as more and more Americans grew accustomed to the idea of psychological treatment for many sorts of behavioral problems. Researchers from a variety of quarters advocated the "continuum" as opposed to the "dichotomous" view of the relationship between normal and deviant drinking. The new federal interest in alcohol-related problems broadcast expanded claims concerning the severity of the nation's alcohol problems. Even the cultural demand for jailing drunkards (though not the practice of it) diminished.82
4. Episode Two: The Rand Report and the Abstinence Counterattack
The publication of the first Rand Report2 in mid-1976 brought the controlled-drinking controversy to a head. The Rand study derived from an ambitious ongoing evaluation system established by the NIAAA in order to monitor the performance of federally funded treatment centers. A preliminary report of treatment findings was available in 1973.83 The Stanford Research Institute (SRI), in collaboration with Rand, published a more complete account of monitoring system data plus a special 18-month follow-up study in which attrition from the sample had been reduced, in 1975.84 Even this earlier report stimulated controversy, no doubt for several reasons: (1) the study's scope was nationwide,85 (2) the authoring organizations were prestigious,85 and (3) the research project had been carried out on behalf of NIAAA and so promised to exert a direct influence on national alcoholism teratment policy. The Rand Report was a broadly conceived reanalysis of the same data SRI had used. Both the SRI and Rand reports had reported the presence of "normal" drinkers in the follow-up data. Rand reported 12% normal drinkers at 6 months following treatment and 22% normal drinkers at 18 months following treatment.2
This episode shifted the major terms of the controlled-drinking debate and broadened its audience. Whereas the Davies episode had been confined to the research literature, the Rand episode involved high participation from the public press. A great many newspaper reports and editorials addressed the controversy. By and large, these reaffirmed the prudence of the abstinence standard, suggesting, I think, that the abstinence requirement could draw on sources of support in the wider culture with little connection to the substance of the scientific dispute or competing paradigms of alcoholism. The Rand episode's primary challenge to the abstinence requirement differed from that laid down bv Davies.1 As noted earlier, Davies had not questioned the traditional clinical emphasis on abstinence but instead merely questioned the theoretical wisdom underlying it. The Rand authors2 took an opposite tack. They had few theoretical pretensions about alcoholism. Indeed, they had interpreted some of their own research findings to suggest that the crucial question in alcoholism treatment effectiveness probably lay in the alcoholic's "decision to seek treatment" rather than the actual treatment he received--a conclusion abstinence advocates, in different circumstances, would have little objection to.
The basic challenge laid down by the Rand authors was that "normal" drinking was a legitimate outcome for alcoholism treatment because it occurred almost as frequently as long-term (6 months) or short-term (1 month) abstinence in their data. This was a wholly descriptive assertion drawn directly from their outcome findings. The normal drinking they monitored, of course, might prove transitory. Thus, Rand authors examined the difference in likelihood of relapse into problematic drinking for abstaining and normal-drinking posttreatment alcoholics over a 1-year period. When they found that normal-drinking alcoholics were no more prone to relapse than abstaining alcoholics, they declared that controlled-drinking outcomes were as successful as abstinent outcomes. If Davies' commentators argued that controlled-drinking outcomes were too infrequent to be given serious consideration, Rand authors argued that particularly long-term abstinence was too infrequent to be made the sole focus and measure of successful treatment. In this way, Rand's findings raised the ancient "is-ought" evaluative dilemma in the debate. Their results, on the one hand, could be interpreted at "face value," implying that what "existed" also carried implications for what "could" or "should" be. The same findings, on the other hand, could be rejected or even made into rhetorical warrants for the redoubling of traditional suasional efforts associated with the abstinence approach. Rand's authors, of course, elected to take the face-value approach, thereby reflecting a connection between "is" and "ought" not always shared by their critics.
The content of proabstention responses to the Rand Report departed from that of the responses to Davies. Davies1 critics had devoted much attention to the question of Davies' alcoholism diagnoses and little attention to the quality of his outcome data. Rand's critics reversed the emphasis. Much commentary did stress that potential controlled-drinking alcoholics could not be diagnostically differentiated from those who must abstain. Fox's86 and Orford's87 comments on the Report, additionally, called attention to the alcoholism field's longstanding diagnostic limitation--Orford noted that Rand's large sample made accurate diagnosis problematic. Nevertheless, the main thrust of criticism lay elsewhere. In part, diagnostic criticisms were rare because the Rand authors had anticipated them and taken pains to address them in their analysis. For example, in salient analytical tables, Rand's authors had separated "definitely alcoholic" patients from other patients with less pronounced signs of dependence.
There were other sources for the critical focus on outcome data as well. The Davies episode had seen physicians/psychiatrists evaluating the work of a fellow psychiatrist. This, no doubt, brought a modicum of professional courtesy to the commentaries. More importantly, it implied that advocates on both sides of the debate were familiar with the methods, the circumstances, and even the limitations of Davies' kind of research. The Rand Report, on the other hand, represented a new and alien form of follow-up research on alcoholism. Critics noted that the research was "impersonal" or "statistical" or that a wide gulf separated the Rand authors from actual alcoholism patients or that the authors lacked personal experience and contact with the field.88-90 Such criticism also symbolized broader reservations being felt by the traditional alcoholism movement in the face of the field's recent growth and bureaucratization.
Emphasis on outcome data also tapped an important rhetorical resource proabstentionists could draw on, namely, survey research's dubious scientific reputation in the minds of many Americans. Criticism directed at the survey nature of Rand's data were legion and, notably, not always germane to the controlled-drinking question. For example, many criticized the representativeness and generalizability of Rand's sample, even though an assertion of the existence* of controlled-drinking alcoholics could logically be based on observations made in either a bad or a good sample. The Rand Report was criticized for a great variety of methodological biases: for its nonprobability sample of treatment centers; for big differences in patient characteristics across centers; for high attrition in its samples and the likelihood of respondent self-selection; and even for flaws in Rand's own analysis of nonresponse bias.88-95 The Rand researchers were chided that their data had not been collected by independent interviewers but often by treatment center personnel with a vested interest in favorable outcome results.91,92 Several critics noted that still-drinking alcoholics were notorious for misreporting their drinking behavior, a point that prompted some to comment on the Rand authors' naivete88,96 about alcoholism. Others noted the lack of corroborating evidence drawn from family or therapists.88,89,91 Pendery88 suggested that even the Rand Report's bibliography was biased in favor of pro-controlled-drinking literature.
*Bias in Rand's sample could, however, be employed to argue that Rand authors had greatly overestimated the true frequency of controlled drinkers. Wallace89 (pp.4-5), for example, used the sample's attrition rates to argue for a reduction in Rand's controlled drinking estimate at 18 months from 22% to 6% and the estimate at 6 months from 12% to only 2%.
Much of the remaining methodological criticism of the Rand Report focused on three charges in particular: (1) that too few cases were available in Rand's data for proper examination of the all-important question of the relative relapse probability of normal drinkers and abstainers87,91,93,94; (2) that the "normal-drinking" limits defined by Rand's authors were too permissive91; and (3) that the relapse analysis had not allowed sufficient time to go by to assess the long-term stability of normal drinkers.90 Less frequently mentioned criticisms included that the study's interview schedule was poor; that it gathered data on the wrong phenomena,88 allowing the clinical realities of alcoholism treatment to slip through its grasp90,97; and that data had been poorly handled,88 poorly categorized,94 and misinterpreted.91,94
Some criticism was procedural rather than methodological. The Rand authors were criticized on grounds they themselves had not carried out the study but were merely analyzing "print-outs of data collected in a routine fashion by NIAAA."91 The Rand authors were also criticized for publishing the report at all.88 In part, this line of criticism stemmed from longstanding fears that news of normal-drinking alcoholics from an authoritative source would tempt alcoholics still struggling to abstain. Rand's authors, for their part, had stated their controlled-drinking findings in a cautious and circumspect way, so much so, in fact, that Hodgson95 asked if the controversy might not be regarded as "much ado about nothing." To some commentators, however, Rand's mild tone did not neutralize the report's potential for hazard. "Even though such conclusions are stated tentatively," Blume97 wrote, "their effects on public policy and public opinion will be considerable." Pendery88 argued that Rand's authors had a responsibility to "less sophisticated" readers more prone to misinterpret their work. Seixas95 wrote that the Rand authors' preference for controlled drinking could be "read between the lines" of the report's text. Hingson et al.99 later reported that the Report, as it was interpreted in the mass media, appeared to have little short-term impact on alcoholics, treatment personnel, or the general public in a Boston sample.
Critics' regrets that the report had been published--as well as earlier efforts bv some to delay or cancel its publication82--could not, however, be justified on the basis of the fear of bad consequences alone. The critical rhetoric, instead, built on the contention that the report was poor science, even poor science masquerading as good science. The poor-science theme served a number of valued objectives for Rand critics. For one, poor science could be made an argument against NIAAA's future funding of similar research; for another, it could be used to counterbalance the oft-heard claim by controlled-drinking advocates that they, and not the proabstinence camp, represented the more scientific side91; for a third, it might oblige Rand's authors and supporters in the larger research and policy community to take proabstentionist objections and criticisms seriously. The Rand Report's scientific authority was also challenged in terms of the manner of the report's publication. Critics faulted Rand for the content of newspaper reports concerning the study that appeared prior to the report's publication.91 They also argued that the report had been given no properly scientific prepublication review. Rand's authors, moreover, were faulted for having staged a press conference to announce their report, in violation of scientific norms.
Parts of this episode's debate illustrated the crucial importance of evaluative judgments concerning not the correctness but the relative weight authors attached to the various critical observations and counterobservations crowding the debate. Wallace89 addressed this issue at a global level. He noted that the Rand authors themselves had conceded that their data were far from perfect but had argued, nevertheless, that some data were better than none. For his own part, Wallace89 responded that the report's potential for dire consequences implied the contrary--that no data would have been preferred. The starkest demonstration of the importance of the evaluative preferences underlying critical commentaries, however, can be found in two reviews: Emrick's85 review of the SRI Report84 and Emrick and Stilson's100 review of the Rand Report.2 Emrick's earlier SRI review had catalogued a variety of the study's methodological weaknesses, but his overall evaluation had been positive. He wrote,Despite the many methodological and reporting problems adumbrated here, the findings deserve serious attention. Consistent with the assertion that the findings probably reflect reality to a great extent, many of them are not unusual or surprising, being consistent with a vast body of alcoholism treatment outcome data85 (p.1905).Emrick argued that consistency with past research meant that SRI's report "should be treated with at least as much respect" as other studies in the literature. Emrick and Stilson's100 subsequent review of the Rand Report had much in common with Emrick's earlier review. Emrick and Stilson, however, assumed an opposing evaluative stance, this time arguing that Rand's data should not be relied on. The authors noted, as Emrick85 had before, that many of Rand's findings converged with past research, but this time convergence was cast in a quite different light: "While such agreement might suggest that the results are valid, this is by no means a certainty. The findings of one methodologically flawed study may agree with those of another flawed study, neither being correct."100 The authors also outlined the requirements for genuinely reliable studies, noting how difficult such would be to achieve. "But unless those studies are somehow done," they concluded," our only recourse is to more or less enlightened opinion, the force of which stems primarily from the articulateness of its proponents."100 One measure of the force of methodological preoccupations surrounding the first Rand Report can be found in Rand's subsequent research and second report.4 This report's attentions were locked firmly on methodological considerations--on achieving a more representative sample with less attrition, on collecting data from collaterals and checking self-report data against breathalyzer tests, and on lengthening the follow-up period and strengthening the interview to better collect information on the alcoholic's behavior over the course of the follow-up period.
5. Episode Three: The Sobells and Standards of Conduct
The third episode of the controversy again shifted the major terms of debate and greatly increased its moral stakes. This time, the debate placed its primary focus not on conceptual and definitional issues (as in the Davies episode) nor on methodological issues (as in the Rand episode) but instead on norms of scientific and professional conduct, including allegations of fraud.101 The episode's basic normative axis lay along the issue of whether quite poor outcome results (in either absolute or commonsense terms) deriving from a test of a controlled-drinking treatment approach nevertheless could be regarded and reported as favorable and promising in the light of even worse results deriving from treatment based on the traditional abstinence approach in a comparable group. It is a question that cannot be answered in strictly scientific or empirical terms since it pertains to the appropriateness of competing normative or evaluative judgments. Even so, disputants repeatedly attempted to recast this issue into empirical terms.
Behaviorist psychologists Mark and Linda Sobell had attempted to train gamma alcoholics to control their drinking in a controlled treatment experiment carried out at the Patton State Hospital near San Bernardino, California in 1970. It was called the "Individualized Behavior Therapy for Alcoholics" (or IBTA) study. They reported on this experiment in a series of publications including a 1972 monograph102 and papers on the 1-year and 2-year outcome results.58,103 By the second year follow-up, the Sobells could report103 that patients in the experimental group, which had been directed toward a controlled-drinking goal, functioned "significantly better" than patients in the control group, which had been directed toward a traditional abstinence goal. Caddy et al.104 similarly reported favorable outcome results for the experimental group at 3 years. As early as 1974, Madsen105 argued that Sobells' reports gave a false impression of their actual findings. Such doubts ultimately gave rise to an independently conducted and longer-term study of the post-treatment histories of the Sobells' experimental subjects carried out by a separate group of researchers. The results of this work were contained in a prepublication draft article entitled Controlled Drinking by Alcoholics? Refutation of a Major Affirmative Study, authored by Pendery, Maltzman, and West, which was circulated among a number of alcohol researchers.* The same authors3 later published a modified version of this paper in Science. They argued that the work of the Sobells and Caddy et al. had been widely regarded as establishing the plausibility of a controlled-drinking treatment goal for gamma alcoholics. Pendery et al. attempted to demonstrate that the Sobells' experiment, contrary to published reports, had been a failure. The authors argued, in essence, that in commonsense or absolute terms the results were very poor. The abstract by Pendery et al. charged that "most subjects trained to do controlled drinking failed from the outset to drink safely" and that "the majority were rehospitalized for alcoholism treatment within a year after their discharge from the research project." Ten years after treatment, Pendery et al. wrote, the Sobells' 20 experimental subjects had experienced the following outcomes: eight had continued to drink excessively, either regularly or intermittently; six were total abstainers; four had died from alcohol-related causes; one was missing, though he had been certified as gravely disabled bv his drinking 1 year after discharge from treatment; and one had achieved controlled drinking, though in this case, Pendery et al. argued, the subject was not a true gamma alcoholic in the first place.
*Unfortunately, I have not seen this.
The 10-year results Pendery et al. reported did not, of course, challenge the Sobells' published reports, since the Sobells had conducted no 10-year follow-up. These data in effect reraised the debate's longstanding call for longer follow-up periods. Concerning the 3-year time period covered by both the Pendery et al. restudy and the combined Sobell-Caddy reports, the fundamental charge by Pendery et al. was that the Sobells had misrepresented their data. "In our view," Pendery et al. wrote, "the references to hospital and jail incarcerations in the Sobells' tables and related discussion do not convey the relative that is evident when the actual incarceration records of each of the controlled-drinking subjects are analyzed individually"3 (p. 173). Their work3 (M. L. Pendery, I. M. Maltzman, and I. L. West, unpublished manuscript) made a number of additional criticisms as well: they questioned the Sobells' diagnoses of four controlled-drinking patients, the random assignment of patients into the two relevant study groups, data collection biases and thoroughness, and the handling of relations with the Caddy et al. 103 study.
The Sobells had taken not only a relativistic but an incrementalist perspective on their treatment outcome results. For example, they reported their results in terms of the numbers of proportions of days subjects "functioned well" in the study's follow-up periods. Such results, moreover, were often reported in the form of summary statistics and charts, a presentational style reflecting their underlying commitment to the notion that incremental and relative improvement (not solely abstinence) constituted legitimate measures of favorable treatment outcome. The Sobells10 responded to the Pendery et al.3 attack by arguing they had reported on the comparative and not the absolute success of their treatment program. Moreover, they demonstrated that the negative findings Pendery et al.3 had reported were fully accounted for in their own research reports. The Sobells justified their comparative orientation to treatment outcome in great part on the failure of traditional treatment approaches to help "chronic alcoholics who enter public treatment programs." Thus, the Sobells attempted to place the treatment-effectiveness onus back onto the traditional abstinence requirement rather than on their own approach. Their remarks on this orientation merit a long quotation:An often unspoken but important underlying reason for much of the controlled-drinking research is the marked absence of effective treatments for this widespread public health problem. . . . Thus, even today it is the case that there is no known panacea for alcohol problems, and particularly no scientifically validated effective treatment of chronic alcoholics who enter public treatment programs. In conducting treatment research on the latter population, the emphasis is on comparing the relative effectiveness of different treatment approaches. In this sense, an analogy can be made to cancer research. Treatment outcomes are a mixed bag, heavily speckled with morbidity and even mortality. Nevertheless, some treatments are associated with better outcomes than others. Such a comparative evaluation was the core research question of the IBTA study. The examination of long-term outcomes for only one group in a comparative study of treatments for chronic alcoholics (any comparative study) can easily be portrayed as tragic, as exemplified in the approach taken by Pendery et al. (1982). How-ever, for scientific purposes such an evaluation is meaningless. It fails to address the fundamental research question: is one type of treatment more effective than another? And in failing to address this question, it shields from discovery a major reason who the paradigm conflict exists--traditional treatments for chronic alcoholics in publicly funded treatment programs have little demonstrated efficacy. That is a central reason who the investigation of alternative treatment approaches 'vas, and remains, important; and the finding that nonconventional methods of treatment may yield superior outcomes gives further impetus to the process of paradigm change10 (p.413).Pendery et al.3 offered three reasons for focusing on the Sobells' experimental group exclusively. (1) The Sobells' control subjects were not true controls because they had been identified as appropriate for a treatment goal in which controlled drinking was attainable and so had been affected by expectations that cut into the traditional method's message. (2) The two groups differed prior to treatment--this contention, of course, raised the question of how the control and experimental groups had been assigned. (3) Finally, Pendery et al.3 asserted that they had an absolute, not a relative, objective in mind: "we are addressing the question of whether controlled drinking it itself a desirable treatment goal, not the question of whether the patients directed toward that goal fared better or worse than a control group that all agree fared badly"3 (pp. 172-173). Maltzman106 offered further justification by asserting that the Sobells, themselves, had made "absolute" statements concerning the effectiveness of controlled drinking--a charge the Sobells rejected.107
An important byproduct of the form of the debate was that it focused each side's attention on the failures of the other's approach--a consequence, Cook108 has argued, that caused both sides to lose ground in the struggle. The Sobells,10 for example, countered the long-term findings of Pendery et al. by arguing that the few long-term outcome studies in the treatment literature boded ill for traditional treatment approaches. They cited Vaillant's14 conclusion that long-term results were often seemingly no better than the improvements that might have been expected from the natural history of alcoholism. Among other discouraging reports, they cited the second Rand Report's finding4 that only 7% of its sample had maintained continuous abstinence over its 4-year follow-up period. The intensity of the Sobells' focus on both (1) a comparative framework for outcome evaluation and (2) the failures of the traditional approach was perhaps best evidenced in their handling of the mortality rates associated with the IBTA study.
Pendery et al.3 had reported from their long-term follow-up that four of the controlled-drinking subjects had died, all from alcohol-related causes. The Sobells addressed themselves to this long-term mortality issue in their response to Pendery et al. They noted, for example, that the deaths had occurred 6 1/4, 8 1/2, 10 1/2, and 11 years following discharge from the IBTA treatment program, a finding that stands in marked contrast to the expected [i.e., higher] mortality rate for this sample"10 (p. 433) based on the available mortality literature. The Sobells also argued that in fact only three of the four deaths were alcohol related and that they resulted from "accidental or intentional occurrences [i.e., suicide]" rather than "from chronic diseases well known to be related to alcoholism and more indicative of chronic heavy drinking"10 (p. 343). The Sobells also noted that the four cases of mortality ranked 15th, 16th, 17th, and 20th in the experimental group in terms of their numbers of favorably functioning days in the 2-year follow-up period. The time order of the deaths was also consistent with the 2-year follow-up data. Both findings were regarded as reflecting favorably on the measurement validity of the Sobells' follow-up measures. The Sobells also reported mortality information on the 20 control subjects who had been directed toward an abstinence goal and to whom the Pendery et al.3 report had not paid much attention. The Sobells10 reported that six of these cases had died--two before any experimental subject had died--and that four ot the deaths were alcohol related, often involving chronic alcohol-related diseases. The mortality discussion eerily dramatizes the gap between the comparative and the noncomparative outlooks of the two sides.
The normative character of the Sobell episode made the episode's audience wider and caused the debate's content to be more broadly defined. National television and radio coverage was added to the controversy's preexisting layer of newspaper coverage (see Dean109). CBS television's public affairs program The Fifth Estate, aired in September, 1982, interviewed the Pendery et al. authors as well as Halmuth Schaefer, who had supervised the controlled-drinking experiment at Patton and had since publicly dissociated himself from its findings.110 CBS radio's science program Quirks and Quarks interviewed the Sobells in November, 1982. CBS television's news magazine 60 Minutes aired a critical report on the Sobells in March 1983.111 Newspaper coverage shifted markedly in the character and scope of its content. With issues of personal integrity at stake, newspaper descriptions broadened to include matters of character, personal motivation, and interpersonal relationships.* Little of this sort of content could be found in newspaper coverage of the first Rand controversy.
*A feature article in the New York Times, for example, reported that the Sobells had circulated a letter among colleagues charging that Pendery was not qualified to serve as an objective. neutral observer because she is "'a recovered alcoholic' who is 'emotionally invested' in abstinence and is determined to ruin their [the Sobells'] careers because of her personal feelings about alcoholism and controlled drinking." It also conveyed allegations that Pendery was romantically involved with Maltzman, thereby "casting doubt on his [Maltzman's objectivity in endorsing her findings." The article reported that Pendery "has been deliberately sipping drinks with professional colleagues and journalists to prove she is no ideological teetotaler" and that she and Maltzman denied "any romance" and that Maltzman was calling the allegations "outrageous, irrelevant and scurrilous" and charging the Sobells in turn with "impugning motives in an effort to divert attention from the facts"112 (p C7).
The controversy's normative character sent it into new forums that were not concerned with the controlled-drinking issue per se but instead with the integrities of the researchers. After circulation of the draft paper (M. L. Pendery, I. M. Maltzman, and L. J. West, unpublished manuscript) but prior to publication of the Science article,3 the Sobells asked their employers, the Addiction Research Foundation (ARF) of Ontario, Canada, to empanel a committee to judge the scientific and personal integrity of their work. In June 1982 the ARF appointed an independent Committee of Enquiry for that purpose, chaired bv law professor Bernard M. Dickens. They submitted their report in October, 1982.113 The Sobells' research was also reviewed by the Subcommittee on Investigations and Oversight, Committee on Science and Technology of the United States House of Representatives, which conveyed its conclusions by letter to the Sobells in March, 1983.114 A special five-member federal panel, chaired by Robert L. Trachtenberg, deputy administrator of the Alcohol, Drug Abuse, and Mental Health Administration, reported its own review in September, 1984.115 The ethics committee of the American Psychological Association reviewed complaints filed against Pendery and Maltzman both by Caddy and by the Sobells, the latter alleging that Pendery and Maltzman had made "repeated, highly sensationalistic statements to news media" that "grievously and to some extent irremediably damaged our reputations" (quoted by Boffey112). Courts and lawyers also became involved in the dispute. The Sobells' request to Federal District Court that it block Pendery's use of their experimental subjects' names on grounds that such use violated the Sobells' guarantee of confidentiality to study participants was dismissed in 1977.112 On advice from lawyers, participants on both sides selectively participated in the deliberations of review forums or programs aired over public media, which raised charges and countercharges of selective candor.109 In July, 1984,a group of subjects who took part in the Patton study, and their relatives filed a $40 million claim against the State of California, alleging that the experiment had led to arrests, pain, public humiliation, and even the deaths of four participants; and that the Sobells' data were "negligently or intentinally [sic] misrepresented and falsified.109(Boffey115 describes this suit as involving $96 million in claims.)
Reviews of the integrity of the Sobells conducted by the Dickens Committee, the House subcommittee, and the federal review panel chaired by Trachtenberg exonerated them of misconduct, although the investigations unearthed a few methodological and procedural failings. These forums also noted that their own information-gathering efforts were often limited. Yet, despite the great effort associated with such reviews, the Sobells' exonerations harbored a serious dilemma. What was the relationship between charges of "misconduct" in professional or personal terms, on the one hand, and charges of "poor scientific methods or reporting practices," on the other? Walker and Roach,116 in their critique of the Dickens Committee's report, argued that these two levels of criticism were inextricably intertwined. The committee's exoneration of the Sobells, they wrote, was "likely to be interpreted by many readers as an endorsement not only of the Sobells' integrity, but also of the validity of their findings." Thus, a rhetorical gap had been created. The Sobells might be innocent of misconduct, but their research might nevertheless be sufficiently flawed in methodological terms to invalidate its findings and the conclusions drawn from it. The standards and substance of proof respecting these two planes of challenge were, of course, quite different. To observers fundamentally concerned with the scientific or therapeutic merit of the controlled-drinking approach, then, the Dickens Committee appeared to have defined its mandate much too narrowly.
Walker and Roach116 argued that if the Sobells' work had not violated norms of professional or personal integrity, it had nevertheless violated the norms of good scientific research. They criticized the Sobells' work for poor reporting of follow-up data, for inaccurate reporting of the frequency of follow-up contacts, and for the procedures for assignment of subjects to the experiment's study groups. But although Walker and Roach may have intended in this way to return the focus of discussion to the substantive and methodological issues raised by the Sobells' research, the Sobells' reponse117 regarded Walker and Roach's argument as raising issues on a normative dimension, one concerning the appropriate standards of scientific research in the contemporary alcohol field. The bulk of the Sobells' response to Walker and Roach117 addressed this new normative plane. In effect, they asked how scientific standards should be defined, and they argued that their own work measured up very well to prevailing methodological standards as such standards were evidenced in even the best of recent comparable research. The Sobells' text repeatedly voiced assertions that Walker and Roach "proposed standards of scientific reporting that any scientist would find unreasonable"117 (p. 158), that "hardly a published study exists which could not be criticized in some way" (p.158), that "the standards advocated by Walker and Roach are unrealistic and unworkable" (p.162), and that their study had been subjected to microscopic methodological analysis that no contemporary work--including that of Walker himself--could withstand.
Let us examine a few of the valuative themes and strains that span the three historical episodes and the interepisode period. Some of these themes and strains are easily observed--for example, the opposition between the lack of value attached to a controlled-drinking treatment goal by some of Davies' commentators and the positive value attached to the same goal (in light of the tact of drinking's prevalence and conventionality in American society) by controlled-drinking advocates in the interepisode period. Other normative dimensions are subtler and demand more work on our parts in order either to perceive them or to assess their importance. My sense, however, is that normative dimensions go a long way toward providing an account of both the controversy's intractability and its passion.
One readily apparent value conflict concerned the flow of information between the scientific/treatment community and the larger society. It pitted the societal and scientific premium on the free flow of information against the therapeutic premium on the management of information on behalf of hopes for the patient's betterment. The Davies episode saw commentators expressing views on both sides of this value conflict--from Bell's39 desire to modulate information by conducting relevant research "with a minimum publicity" to Selzer's41 objection to information control such as was evidenced in his account of the anonymous funding agency that asked him to omit "embarrassing" controlled-drinking findings from his report. In particular, the Davies episode threw into sharp relief the potential for conflict between scholarly candor concerning the field's fallible diagnostic ability and the consequences of such candor for the putative well-being of alcoholic patients. The Rand authors were criticized for inappropriate dissemination of their findings on several grounds--among them, lack of appropriate peer review for the report, improper prepublication news reports, indifference to "less sophisticated" readers, and improper announcement of their report--and they, in turn, criticized abstinence advocates for attempting to muffle the flow of research data. The Rand episode raised questions concerning not only the dissemination of information but the norms surrounding such irformation's proper claims to scientific authority.
The Sobell episode, in turn, further highlighted a number of strains in the social norms pertaining to information flow. Pendery et al.,3 for example, had based their critique of the Sobells partly on grounds that the Sobells' research was "widely regarded" as establishing the credibility of the controlled-drinking treatment goal. Similarly, Walker and Roach116 criticized the Dickens Committee's report on the ground that their findings "would be interpreted" as vouchsafing not only the Sobells' integrities but the validity of their scientific results as well. These criticisms, in effect, shifted the normative focus to questions concerning the impressions and consequences research findings might occasion in the minds of various social audiences--raising the normative issue of the proper scope of a researcher's responsibility for the interpretations and consequences placed on his findings, particularly in respect to audiences more or less distant from the scientists, therapists, and policymakers for whom their work may have been initially intended.
The controversy also repeatedly raised normative issues concerning the proper relationship between research findings about controlled drinking and actual alcoholism treatment. We saw that the fact of positive controlled-drinking results, by itself, did not constitute a direct warrant for revising the traditional emphasis on abstinence. Davies' commentators and, indeed, Davies himself emphasized a thick barrier between controlled-drinking research and traditional treatment. Only Brunner-Orne35 suggested that such results might justify controlled-drinking treatment goals for patients who had failed to achieve abstinence or who were nonalcoholic in the first place. The connection between research and treatment was mediated by a number of normative concerns. The interepisode period saw controlled-drinking advocates increasingly argue for a link between the existence of controlled-drinking cases and treatment aimed at a controlled drinking goal. That link was forged, however, as much in growing doubts over the efficacy of traditional treatment as in the accumulation of favorable controlled-drinking research. Whereas proabstentionists defined abstinence-rejecting alcoholics as strong evidence for the imperative of an unbroken and unyielding societal and professional consensus on behalf of abstinence's desirability, controlled-drinking advocates were more likely to view abstinence-rejecting alcoholics at face value. This face-value orientation underlay much of the Rand episode's rhetorical conflict and constituted an important, if tacit, normative premise for the Rand authors' interpretations of their outcome frequencies.
In the Sobell episode, finally, the face-value perspective on abstinence-rejecting alcoholics became a prominent rhetorical premise for controlled-drinking treatment experimentation. In the absence of the traditional school's emphasis on a profound change marking the commencement of abstinence, nontraditional therapists like the Sobells oriented their efforts toward achieving small or incremental changes in treatment outcomes. An incrementalist orientation, in turn, made possible reports of experimental successes that might appear to the traditional camp as abject treatment failures. So distant and mutually unintelligible were the orientations of these two approaches that they could raise the specters of bad faith and fraud between the camps.
What should define the treatment community's appropriate moral duty, on the one hand, to alcoholics who reject abstinence and, on the other, to alcoholics who adopted abstinence or at least have attempted to adopt it? I have argued that the controlled-drinking controversy grew apace with an increasing societal and policy emphasis on untreated or treatment-failing alcoholics in this country. Relatively little attention was focused on these alcoholics in the period leading up to the Davies episode. The interepisode period brought a variety of social changes and research findings that greatly enhanced their prominence in alcohologic thought. The Rand Report, it might be argued, symbolically represented the expansion of the societal vision of alcoholism to include populations who were often previously excluded from sight because of preference at the time for observations made on recovering alcoholics in Alcoholics Anonymous groups or treatment samples like those of Davies' commentators. The meaning of this broadening in perspective, however, was not merely empirical. I have argued that the expansion brought new ethical implications to the role of the alcoholism therapist. Whereas traditional treatment lodged the ultimate responsibility for embracing abstinence in the alcoholic's own hands, the social demand for treatments addressed to alcoholics who had often rejected or failed the traditional approach brought a felt duty to do something new and, it was hoped, helpful.
The controversy's intensity probably owes much to the competing and even incompatible moral "sympathies" aligned with these two alcoholic subpopulations. With their sympathies anchored in abstinence-rejecting alcoholics, controlled-drinking proponents inevitably drew in to question the rationalistic assumptions of and the place of will in the traditional paradigm. The experiential justification for this doubt lay in the belief that so many alcoholics had not been helped by the traditional approach. The traditional camp, on the other hand, anchored its sympathies in the abstinence-achieving or the abstinence-attempting alcoholic. This alcoholic symbolically affirmed both the possibility and accessibility of abstinence for all alcoholics and, by extension, the hegemony of "rational mind" over "alcoholic body chemistry" implied in the traditional paradigm's dualism. On the abstentionist side, the alcoholism therapist's duty lay in supporting the merit of the abstinent alcoholic's moral achievement and in offering prospects of such achievement to still-drinking candidates. Moreover, with their sympathies fixed thus, such therapists could view the emerging evidence of poor prognosis for abstinence-rejecting alcoholics--including its high mortality rate--as valuable rhetorical weapons in the argument for abstinence. The two sympathies tied therapists to the moral and social interests of the two alcoholic subpopulations, creating a struggle over each's appropriate social definition or status. For example, controlled-drinking advocates could argue that the traditional model--with its commitment to a willed acceptance of abstinence--had the effect of reintroducing and indeed heightening the social stigma surrounding active alcoholism. From a proabstentionist perspective, on the other hand, a deemphasis on will threatened a societal reduction in the moral credit due abstinent alcoholics.
The normative connections between the larger society and the scientific treatment communities evidenced in this debate provides a complex territory deserving much greater analytical attention that I can give it here. I have noted that the alcoholism movement over much of its early period preferred a sharply defined, dichotomous conception of the divide between normal and alcoholic drinking, in part because of the contemporary climate of opinion respecting both alcohol as a social problem and then-current conceptions of the proper social warrant and authority for science and treatment. Social considerations also played a part in the voluntarism/involuntarism attributed to alcoholism and the severity and scope attributed to the problem. In the absence of powerful theoretical tools for understanding and treating the condition, social forces held broad sway in defining the phenomenon at the focus of the new scientific movement. I have argued that many of the perceived societal requirements of the early movement gave way to changing conceptions in the period following Davies' report. In part, at least, the controlled-drinking controversy reflects a struggle between two competing historical images of the proper place of science and treatment with respect to alcoholism in American society.
To be sure, there are many other intriguing normative dimensions evidenced in the controlled-drinking controversy, some of which have been noted in the body of this chapter. The lack of significant scientific-treatment progress in the alcoholism field has occasioned tensions and reverberations along many of the normative dimensions that define scientific/treatment activity, both internally and in relation to the broader society.* And although such reverberations--over the course of the controlled drinking controversy--have often been experienced by the alcoholism field itself as angry, painful, and even futile exercises in disputation, the controversy's history nevertheless offers a strong beacon for guiding sociological analysis of the normative concerns that have animated and structured it.
*Such tensions, incidentally, suggest a line of analysis more akin to Fleck's118 than to Kuhn's119 picture of science and society, particularly because or Fleck's decidedly "externalist" emphasis.
ACKNOWLEDGMENTS. The author wishes to thank Jim Baumohl, Walter Clark, Tom Colthurst, Kave Fillmore, Barbara Owen, Robin Room, and Robert Straus for their helpful criticisms of an earlier draft and Andrea Mitchell and Craig Reinarman for their bibliographic assistance. Work on this review was supported bv a National Alcohol Research Center grant (AA 05595) from the U.S. National Institute on Alcohol Abuse and Alcoholism to the Alcohol Research Group, Institute of Epidemiology and Behavioral Medicine, Medical Research Institute of San Francisco.
References1. Davies DL: Normal drinking in recovered alcohol addicts. Q J Stud Alcohol 23:94-104, 1962.
2. Armor DJ, Polich JM, Stambul HB: Alcoholism and Treatment. Santa Monica, Rand Corporation, 1976
3. Pendeny ML, Maltzman IM, West LJ: Controlled drinking by alcoholics? New findings and a reevaluation of a major affirmative study. Science 217:169-175, 1982.
4. Polich JM, Armor DJ, Braiker HH: The Course of Alcoholism: Four Years after Treatment. New York, John Wiley & Sons, 1980.
5. Carroll JFX: How Eagleville came to this conference topic. Am J Drug Alcohol Abuse 5:257-275, 1978.
6. Helzer JE, Robbins LN, Tavior JR, et al: The extent of long-term moderate drinking among alcoholics discharged from medical and psychiatric treatment facilities. N Engl J Med 312:1678-1682, 1985.
7. Heather N, Robertson I: Controlled Drinking. London, Methuen, 1981.
8. Miller WR: Controlled drinking; a history and a critical review. J Stud Alcohol 4:68-83, 1983.
9. Marlatt GA: The controlled-drinking controversy: A commentary. Am Psychol 38:1097-1110, 1983.
10. Sobell MB, Sobell LC: The aftermath of heresy: A response to Pendery et al.'s (1982) critique of "Individualized Behavior Therapy for Alcoholics." Behav Res Ther 22:413440, 1984.
11. Report: The Research Council on Problems of Alcohol. Science 88:329-332, 1938.
12. The Research Council on Problems of Alcohol: An outline of its program, objectives, resources and progress to date. Q J Stud Alcohol 2:431-451, 1941.
13. Gordis F: Editorial: What is alcoholism research? Ann Intern Med 85:821-823, 1976.
14. Vaillant GF: The Natural History of Alcoholism. Cambridge, Harvard University Press, 1983.
15. Levine HG: The discovery of addiction: Changing conceptions of habitual drunkenness in America. J Stud Alcohol 39:143-174, 1978.
16. Lemere F: What happens to alcoholics. Am J Psychiatry 109:674676, 1953.
17. Shea JF: Psychoanalytic therapy and alcoholism. Q J Stnd Alcohol 15:595605, 1954.
18. Norvig J, Nielsen HA: A follow-up study of 221 alcohol addicts in Denmark. Q J Stud Alcohol 17:633-642, 1956.
19. Pfeffer AZ, Herger S: A follow-up study of treated alcoholics. Q J Stud Alcohol 18:624-648, 1957.
20. Moore RA, Ramseur F: Effects of psychotherapy in an open-ward hospital on patients with alcoholism. Q J Stud Alcohol 21:233-252, 1960.
21. Seizer NIL, Holloway WH: A follow-up of alcoholics committed to a state hospital. Q J Stud Alcohol 18:98-120, 1957. 22. Harper J, Hickson H: The results of hospital treatment of chronic alcoholism. Lancet 261:1057-1059, 151.
23. De Morsier G, Feldman H: Le traitement de l'alcoolisme par l'apomorphine: Etude de 500 cas. Schweiz Arch Neurol Psychiatr 70:434-440, 1952.
24. Kendell RE, Staton MC: The fate of untreated alcoholics. Q J Stud Alcohol 27:30-41, 1966.
25. Zwerling I, Rosenbaum NI: Alcoholic addiction and personality (nonpsychotic conditions), in Arieti S (ed): American Handbook of Psychiatry (vol 1). New York, Basic Books, 1959, pp 623-644.
26. Williams RJ: Biochemical individuality and cellular nutrition: Prime factors in alcoholism. Q J Stud Alcohol 20:452-463, 1959.
27. Hoff EC: Some principles of therapy for alcoholics, in Himwich H (ed): Alcoholism: Basic Aspects of Treatment. Washington, American Association for the Advancement of Science (Publ No. 47), 1957, p 1891.
28. Davies DL: Conversation with D. L. Davies. Br J Addict 74:239-249, 1979
29. Fox R: Comment on Davies DL: Normal drinking in recovered alcohol addicts. Q J Stud Alcohol 24:117, 1963.
30. Block NIA: Comment on Davies DL: Normal drinking in recovered alcohol addicts. Q J Stud Alcohol 24:114-117, 1963.
31. Lemere F: Further comment on Davies DL: Normal drinking in recovered alcohol addicts. Q J Stud Alcohol 24:727-725, 1963.
32. Maidman NIM: Further comment on Davies DL: Normal drinking in recovered alcohol addicts. Q J Stud Alcohol 24:728-730, 1963.
33. Williams L: Comment on Davies DL: Normal drinking in recovered alcohol addicts. Q J Stud Alcohol 24:111-113, 1963.
34. Smith JA: Comment on Davies DL: Normal drinking in recovered alcohol addicts. Q J Stud Alcohol 24:322-324, 1963.
35. Brunner-Orne M: Further comment on Davies DL: Normal drinking in recovered alcohol addicts. Q J Stud Alcohol 24:730-733, 1963.
36. Lolli G: Comment on Davies DL: Normal drinking in recovered alcohol addicts. Q J Stud Alcohol 24:326-330, 1963.
37. Esser PH: Comment on Davies DL: Normal drinking in recovered alcohol addicts. Q J Stud Alcohol 24:119-121, 1963.
38. Kjolstad T: Further comment on Davies DL: Normal drinking in recovered alcohol addicts. Q J Stud Alcohol 24:733-735, 1963.
39. Bell RG: Comment on Davies DL: Normal drinking in recovered alcohol addicts. Q J Stud Alcohol 24:321-322, 1963.
40. Armstrong JD: Comment on Davies DL: Normal drinking in recovered alcohol addicts. Q J Stud Alcohol 24:118-119, 1963.
41. SeIzer NIL: Comment on Davies DL: Normal drinking in recovered alcohol addicts. Q J Stud Alcohol 24:113-114, 1963.
42. Glatt MM: Comment on Davies DL: Normal drinking in recovered alcohol addicts. Q J Stud Alcohol 24:116-117, 1965.
43. Thimann J: Comment on Davies DL: Normal drinking in recovered alcohol addicts. Q J Stud Alcohol 24:324-325, 1963.
44. Tiebout HNI: Comment on Davies DL: Normal drinking in recovered alcohol addicts. Q J Stud Alcohol 24:109-111, 1963.
45. Zwerling I: Comment on Davies DL: Normal drinking in recovered alcohol addicts. Q J Stud Alcohol 24:117-118, 1963.
46. Myerson DJ: Comment on Davies DL: Normal drinking in recovered alcohol addicts. Q J Stud Alcohol 24:325, 1963.
47. Edwards G: Paradigm shift or change in ownership? The conceptual significance of DL Davies's classic paper. Drug Alcohol Depend 15:19-35, 1985.
48. Jellinek ENI: Phases of alcohol addiction. Q J Stud Alcohol 13:673-684, 1952.
49. Edwards G: A later follow-up of a classic case series. DL Davies's 1962 report and its significance for the present. J Stud Alcohol 46:181-190, 1985.
50. KaIb NI, Propper NI: The future of alcohologv: Craft or science? Am J Psychiatr 133:641-645, 1976.
51. Wiener C; The Politics of Alcoholism: Building an Arena around a Social Problem. New Brunswick, NJ, Transaction Books, 1981,
52. Sobell MB, Sobell LC: Alternatives to abstinence: Time to acknowledge reality. Addictions 21:2-29,1974.
53. Pattison ENI: Nonabstinent drinking goals in the treatment of alcoholics, in Gibbins RJ, Israel Y, Kalant H, et al (eds): Research Adcances in Alcohol and Drug Problems (vol 3). New York, John Wiley & Sons, 1976, pp 401~55.
54. Emrick CD: A review of psychologically oriented treatment of alcoholism I. The use and interrelationships of outcome criteria and drinking behavior following treatment. Q J Stud Alcohol 35:523-549, 1974.
55. Bailey MB, Stewart J: Normal drinking by persons reporting previous problem drinking. Q J Stud Alcohol 28:305-315, 1967.
56. Drew L: Alcoholism as a self-limiting disease. Q J Stud Alcohol 29:956-967, 1968.
57. Marlatt GA, Demming G, Reid JB: Loss of control drinking in alcoholics; an experimental analogue.J Abnorm Psychol 81:233~241, 1973.
58. Sobell NIB, Sobell LC: Alcoholics treated by individualized behavior therapy: One-year treatment outcome. Behav Res Ther 11:599818, 1973.
59. Pattison ENI, Sobell MB, Sobell LC (eds): Emerging Concepts of Alcohol Dependence. New York, Springer, 1977.
60. Llovd RW Jr, Salzberg HC: Controlled social drinking; an alternative to abstinence as a treatment for some alcohol abusers. Psychol Bull 82:815-842, 1975.
61. Clark WB, Cahalan D: Changes in problem drinking over a four year span. Addict Behav 1:251-259,1976.
62. Straus R: Problem drinking in the perspective of social change, 1940-1973, in Filstead WI, Rossi JI, Keller M (eds): Alcohol and Alcohol Problems: New Thinking and New Directions. Cambridge, MA, Ballinger, 1976, pp 29-56.
63. Pattison EM: A critique of alcoholism treatment concepts with special reference to abstinence. Q J Stud Alcohol 27:49-71, 1966.
64. Moore FA: The conception of alcoholism as a mental illness. Q J Stud Alcohol 29:172-175, 1968.
65. Pattison EM: Nonabstinent drinking goals in the treatment of alcoholism. Arch Gen Psychiatry 33:923-930, 1976.
66. Drewery .1: Social drinking as a therapeutic goal in the treatment of alcohol. J Alcohol 9:43~7, 1974.
67. Doherty J: Controlled drinking: Valid approach or deadly snare. Alcohol Health Res World Fall: 28, 1974; ALFAWAP 1:11-21, 1978.
68. Evans M: Modification of drinking. J Alcohol 8:111-113, 1973.
69. Miller W: Controlled drinking therapies: A review, in Miller WR, Munoz RF (eds): How to Control Your Drinking, Englewood Cliffs, Prentice-Hall, 1976, Appendix C, pp 194-228.
70. Pomerleau 0, Pertschuk M, Stinnett J: A critical examination of some current assumptions in the treatment of alcoholism. J Stud Alcohol 37:849-867, 1976.
71. Hamburg S: Behavior therapy in alcoholism. A critical review of broad-spectrum approaches. J Stud Alcohol 36:69-87, 1975.
72. Verden P, Shatterly D: Alcoholism research and resistance to understanding the compulsive drinker. Ment Hyg 55:331-336, 1971.
73. Hill MJ, Blane HT: Evaluation of psychotherapy with alcoholics: A critical review. Q J Stud Alcohol 28:76-104, 1967.
74. Penningroth P: Treatment goals for alcoholic--the great debate continues. Addictions 22:32-63, 1975.
75. Miller WR, Munoz RD (eds): How to Control Your Drinking. Englewood Cliffs, Prentice-Hall, 1976.
76. Keller NI: Problems of edpidemiology in alcohol problems. J Stud Alcohol 36:1442-1451, 1975.
77. Keller NI: Alcohol, science and society: Hindsight and forecast, in Gomberg EL, White HR, Carpenter JA (eds): Alcohol, Science and Society Revisited. Ann Arbor, University of Michigan Press, 1982, pp 1-16.
78. Jellinek EM: Recent trends in alcoholism and in alcohol consumption. Q J Stud Alcohol 8:142, 1947.
79. Jellinek EM: Estimating the prevalence of alcoholism: Modified values in the Jellinek formula and an alternative approach. Q J Stud Alcohol 20:261-269, 1959.
80. Clark WB: Operational definitions of drinking problems and associated prevalence rates. Q J Stud Alcohol 27:648-668, 1966.
81. Bacon SD: The administration of alcoholism rehabilitation programs. Q J Stud Alcohol 10:147, 1949.
82. Roizen R: Comment on the "Rand Report," J Stud Alcohol 38:170-178, 1977.
83. Towle L, Bosley J, Borgeson N, et al: Alcoholism Program Monitoring System Development--Evaluation of an ATC Program. Menlo Park, CA, Stanford Research Institute (Project 1357), 1973.
84. Ruggels WL, Armor DJ, Polich JM, et al: A Follow-up Study of Clients at Selected Alcoholism Treatment Centers Funded hy NLIAA, Final Report to NIAAA. Menlo Park, Stanford Research Institute, 1975.
85. Emrick CD: Review of Ruggels WL, Armor DJ, Polich JM, et al: A Follow-up Study of Clients at Selected Alcoholisni Treatment Centers Funded by NIAAA, Finial Report. J Stud Alcohol, 37:1902-1907, 1976.
86. Fox V: The controlled drinking controversy. JAMA 236:863, 1976.
87. Orford J: The future of alcoholism; a commentary on the Rand Report. Psychol Med 8:58, 1978.
88. Pendery M: Comments on the Rand Report. Statement provided in NCA press conference, Shoreham Hotel, Washington DC, July 1, 1976 (typescript).
89. Wallace J: Alcoholism and treatment: A critical analysis. Statement provided in NCA press conference, Shoreham Hotel, Washington, DC, July 1, 1976 (typescript).
90. Ewing JA: Statement on the Rand Report. Statement provided in NCA press conference, Shoreham Hotel, Washington DC, July 1, 1976 (typescript).
91. Seixas F: Comments on "Alcoholism and Treatment." Statement provided in NCA press conference, Shoreham Hotel, Washington, DC, July 1, 1976 (typescript).
92. Hyde LG: Statement provided in NCA press conference, Shoreham Hotel, Washington, DC, July 1, 1976 (typescript).
93. Adinolfi AA, DiDano B: Comment on the "Rand Report." J Stud Alcohol 38:169, 1977.
94. Crawford JJ, Pell J: The Rand Report: A brief critique. Addict Behav 2:141-146, 1977.
95. Hodgson RJ: Much ado about nothing much: Alcoholism treatment and the Rand Report. Br J Addict 74:227-234, 1979.
96. Pittman D: Critique of AlcohoILsir' and Treatment. Text of telephone critique on file in Rand Reaction, Library, Alcohol Research Group, 1816 Scenic Avenue, Berkeley, CA 94709.
97. Blume S: Comment on the "Rand Report." J Stud Alcohol 38:163-168, 1977.
98. National Council on Alcoholism: Press release, Shoreharn Hotel, Washington, DC, July 1, 1976. See Ref. 2, pp 232-234.
99. Hingson R, Scotch N, Goldman E: Impact of the Rand Report on alcoholics, treatment personnel and Boston residents. J Stud Alcohol 38:2065-2076, 1977.
100. Emrick CD, Stilson DW: Comment on the "Rand Report." J Stud Alcohol 30:152-163, 1977.
101. Boffey PM: Alcoholism studv under new attack. New York Times June 28, 1982, p. A12.
102. Sobell MB, Sobell LC: Individualized Behavior Therapy for Alcoholics: Rationale, Procedures, Preliminary Results and Appendix. (Monograph No. 13), Sacramento, Calif. Dept. Mental Hygiene, 1972.
103. Sobell MB, Sobell LC: Second year treatment outcome of alcoholics treated by individualized behavior therapy: Results. Behav Res Ther 14:195-215, 1976.
104. Caddy GR, Addington Jr HJ, Perkins D: Individualized behavior therapy for alcoholics: A third year independent double-blind follow-up. Behav Res Ther 16:345-362, 1978.
105. Madsen W: The American Alcoholic: The Nature-Nurture Controversy in Alcoholic Research and Therapy. Springfield, IL, Charles C. Thomas, 1974.
106. Maltzman IM: More on: "Controlled drinking versus abstinence: Where do we go from here?" Bull Soc Psychol Addict Behav 3:71-73, 1984.
107. Sobell MB, Sobell LC: More on Maltzman. Bull Soc Psychol Addict Behav 3:74-75, 1984.
108. Cook DR: Craftsman versus professional: Analysis of the controlled drinking controversy. J Stud Alcohol 46:433-442, 1985.
109. Dean L: Controlled drinking controversy: Review & analysis of the Patton Study conducted by Mark & Linda Sobell. Grassroots April: 11-15, 1985.
110. Researcher repudiates study on alcoholic social drinking. New York Times Aug 26, 1982.
111. Doob AN: Understanding the nature of investigations into alleged fraud in alcohol research: A reply to Walker and Roach. Br J Addict 79:169-174, 1984.
112. Boffey PM: Showdown nears in feud over alcohol studies. New York Times Nov 2, 1982.
113. Dickens BM, Doob AN, Warwick OH, et al: Report of the Committee of Enquiry into Allegations Concerning Drs. Linda & Mark Sobell, Berkeley, Addiction Research Foundation, 1982.
114. Jansen JE: Letter to Drs Mark and Linda Sobell, Mar23, 1983, copy in the author's possession.
115. Boffey PM: Panel finds no fraud by alcohol researchers. New York Times Sept 11, 1984.
116. Walker KD, Roach CA: A critique of the report of the Dickens' enquiry into the controlled drinking research of the Sobells. Br J Addict 79:147-156, 1984.
117. Sobell MB, Sobell LC: Under the microscope yet again: A commentary on Walker and Roach's critique of the Dickens Committee's enquiry into our research. Br J Addict 79:157-168, 1984.
118. Fleck L: Genesis and Development of a Scientific Fact, Chicago, Universitv of Chicago Press, 1979 (first published, 1935).
119. Kuhn TS: The Structure of Scientific Revolutions (ed 2). Chicago, University of Chicago Press, 1970.