Of the 456 core-city men, 110 became problem drinkers at some point in their lives; of the 204 college men, only 26 did. At the last assessment of them, Vaillant found 20 percent of the alcohol abusers in the core-city group to be drinking moderately and 34 percent to be abstaining. In other words, fully one out of five of Vaillant's own subjects was controlling his drinking—neither going on binges nor abstaining. What is more, Vaillant himself acknowledges that few abstaining alcoholics never drink again. Indeed, his own definition of abstinence is considerably more permissive than AA's: in Vaillant's study, abstinence means drinking less than once a month. Many might conside"!this to be controlled drinking.

All of Vaillant's case studies tend to make the same point: alcoholics must acknowledge that they have a disease and seek help (usually from AA). He finds, as did the Rand researchers, that those who manage long periods of abstinence tend to be people who stayed with AA for years. (The Rand reports note, however, that many who attend AA drop out and that those who become social drinkers rather than abstainers have little to do with AA.) Vaillant's subjects used AA in many different ways. Some attended only briefly and then went it on their own; others attended frequently without solving their drinking problems. Overall, among the core-city sample, he says, more than a third of the men who achieved a year of abstinence succeeded wholly or in part through AA.

We learn little from Vaillant, however, about those who dealt successfully with their alcohol abuse without help from AA. This was a sizable group: all the social drinkers and even the majority of the abstainers. For them, forces outside formal therapy—the influences of family and friends, and their own personal motivation—were most important in recovery. Thus Vaillant's own results disprove the common notion that alcoholics cannot change on their own. Though Vaillant acknowledges throughout the book the importance of forces outside therapy, "natural forces," he seems strangely reluctant to discuss them. Time and again he plays down willpower and self-reliance.

Vaillant's reticence is doubly disappointing in light of his report that the results of treatment at his hospital clinic "were no better than the natural history of the disease." That is to say, the patients he treated showed no higher rate of recovery than did alcoholics who never received any sort of treatment at all: Vaillant found that of patients at his clinic who received detoxification treatment and were required to attend AA meetings and keep in frequent touch with the clinic, 95 percent relapsed into periods of heavy drinking. Clearly, the admonition to abstain is hardly a cure-all.

Still, Vaillant's honesty in assessing the effectiveness of his clinic is laudable, and rare. Pendery and her colleagues, in disputing the Sobells' study, were not as careful. Though they tracked down subjects who had been taught to drink moderately and found that many had relapsed, they did not follow up on the subjects in the Sobells' comparison group, who had been treated with the standard goal of abstinence in mind. When reading their report or watching 60 Minutes, we were left to imagine that those who had received abstinence therapy experienced no such difficulties. In fact, the purpose of the Sobells' original study was to compare the rate of relapse for the controlled-drinking group and for the abstinent group. Six of the abstinent subjects had died in the period covered by the Pendery investigation (at least four of them from alcohol-related causes), while four died in the controlled-drinking group.

There have now been two investigations of the Sobell and Pendery claims and counterclaims. The Addiction Research Foundation, of Toronto, where the Sobells now work, appointed an independent panel whose chairman was a professor of law at the University of Toronto. The United States Congressional Committee of Science and Technology worked through its Subcommittee on Investigations and Oversights. The investigations have pointed out major problems in the Pendery follow-up study and have exonerated the Sobells of misconduct.

Though Vaillant's data show that alcoholics do return to social drinking, his data also show, like the Rand studies, that the more severe the degree of alcoholism, the harder it is to control or moderate drinking. Those who are heavily dependent on the drug are normally better off quitting entirely. But this applies to all sorts of addictions. Something in the nature of compulsive habits makes a return to moderation difficult: if a heavy smoker quits, he is unlikely to be able to have just a couple of cigarettes a day; if someone loses eighty pounds, he is unlikely to have just one cookie. A couple who have had an intense, unhappy relationship usually find it best to make a complete break. Whenever a habit has grown all-consuming, changing can become a matter of all or nothing.

This is not to say, however, that abstinence is a biological imperative. Expectations are crucial. The belief that one cannot moderate a habit makes it almost impossible to do so. Thus, problem drinkers who define themselves as alcoholics are those most likely to need to abstain. If alcoholism were inbred, mandated by genes, then moderating or reversing it would be impossible. Those who subscribe to the disease theory have yet to discover an inherited, metabolic mechanism that accounts for the alcoholic's loss of control of his drinking. Vaillant himself provides a good review of such metabolic theories, and shows them to be utterly inadequate.

One recent theory, proposed by Marc Schuckit, of the University of California at San Diego School of Medicine, holds that the alcoholic inherits an inability to discriminate the level of alcohol in his blood, and thus cannot tell when he's had too much. This may or may not be so; but it is a far cry from a genetic compulsion to drink. Why would such a person not simply limit himself to a single drink, for instance? He might get drunk inadvertently once or twice because of his insensitivity to the effects of alcohol, but presumably he would soon recognize the problem and, in the future, make sure he did not drink beyond his limit. Another recent finding, reported by David Rutstein, of Harvard Medical School, and others, is that severe alcoholics metabolize liquor differently from normal, social drinkers. But any apparent metabolic difference may well be a result of prolonged, heavy drinking, rather than a cause of the habit. Again, it is hard to see any evidence of a genetic compulsion. In Vaillant's study, a recovered alcoholic's choice of moderation or abstinence was not related to the number of alcoholic relatives he had. It was distinctly related, however, to the drinker's ethnic background. So, culture seems more important here than genes.

Ultimately, it is conceivable that the disease theory itself is contributing to the nation's skyrocketing rate of alcoholism. The very perception of the power of alcohol to corrupt and control may make a powerful difference in the way Americans deal with alcohol. Prohibition proved conclusively that we are not an abstinent nation. So the devaluation of moderate drinking as a treatment goal, and a social norm, is dangerous. Few reformed alcoholics, even "abstaining" alcoholics, never touch another drink. The typical American therapist reasons, "Why not tell everyone with a drinking problem to abstain? What can be lost with that approach?" What can be lost is the patient's resolve to cooperate.

Marty Mann, a prominent figure in AA, has estimated that there were three million alcoholics in America in 1943, five million in 1956, and six and a half million in 1965. Today, experts in the alcoholism field put that figure at ten to fifteen million. When we promote the belief that many people—in numbers that seem to be growing all the time—cannot taste alcohol without catastrophic results, we may be fulfilling our own prophecy.

next: What I Discovered Among the Aboriginals