One need not be a scientist to realize that correlation often has little to do with causation, but most journalists wrote as if this were dramatic proof of the highly addictive quality of alcohol, and of the dangers of drinking on the part of anyone under 21 years of age. However, to anyone who has paid attention to life in contemporary Europe where children are frequently introduced to drinking at an early age and where rates of dependency tend to be quite low the finding seems counter-intuitive. Similarly, when one thinks of Orthodox Jews anywhere in the world it is difficult to reconcile their famously low rates of drinking problems with the fact that male infants are given wine on the occasion of their circumcision (on the 8th day of life), and men, women, and children all drink at least twice a week to celebrate the beginning and end of the Sabbath.

In view of the vast panorama of human experience, the fateful findings about early drinking are in fact not only counter-intuitive but would be factually wrong in most of the world. However, in the contemporary US, there is little question that the findings are correct, and, when one thinks about it for a while, quite logical.

The difference is that part of the way in which members of the new temperance movement have tried to "protect American children from alcohol" has obviously been counter-productive. Given a setting where it is deviant (or even illegal) to drink at an early age, it is no surprise that those who do so are precisely those who choose to act in deviant, illegal, or other risky ways. By the same token, moderate individuals who are willing to conform to norms, obey the laws, and minimize risks are those most likely to defer drinking until later. The anti-drink constituency have succeeded in structuring the situation in such a way that the fateful outcome ("early drinking results in drinking problems") is all but inevitable!

By contrast, in most parts of the world, beverage alcohol has not been endowed with a mystical aura of "forbidden fruit" in such a way that drinking it is expected by young people to demonstrate their maturity, to make them more powerful or sexy or dynamic or sociable. Where children are not "protected from alcohol," they show no need of such protection. They learn to drink, usually at home and among their families. Drinking is a wholesome and enjoyable part of everyday life, rather than a risky act which they should hide from their elders, and learn from their ill-informed and inexperienced peers.

As an anthropologist, I often deal with patterns of small, isolated, or even tribal populations whose exotic patterns of belief and behaviour would be impractical in a modern urban setting in the industrial or post-industrial world. But this is not such a case. Some of the best illustrations of my point are the middle-class cultures of contemporary France, Italy, and Spain (among many others). It is in those well-studied and heavily documented contexts that I have repeatedly demonstrated, using their own official statistics, that the occurrence of so-called "alcohol-related problems" (whether physiological, psychological, social relational, economic, or other) is inversely related to both "age-of-onset" and to average per-capita consumption.

As you know, this is just the opposite of what is claimed by World Health Organization, US National Institute on Alcohol Abuse and Alcoholism, and other organizations and individuals that are primarily concerned with restricting or curtailing availability as a public health policy.

Incidentally, the same context that guarantees that early-onset drinkers be deviant and willing to flout the law, also guarantees that they will often do so surreptitiously, drinking too much and/or too fast, among peers who lack knowledge about the specific risks of drunkenness or chemical harm, and who may be reluctant to summon adults in the event that an acute problem should occur. Here again, it is the "just-say-no" approach that makes for problems rather than reducing the potential for harm.

In a "wet culture" where young people are early socialized to drinking, they simultaneously learn how to drink moderately, how and why to avoid drunkenness, not to expect magical transformations from drink, and to view excesses as inappropriate and illustrative of weakness (generally the opposite of what supposedly "protected" youths in the US learn).

In short, the "early onset" theory is accurate but only in those few parts of the world in which the legal and normative system makes it so. In the rest of the world, the opposite is the case!

For anyone who may not find such cross-cultural evidence compelling, I recommend a recent paper that referred to "age at first drink and risk for alcoholism" as "a non-causal association". It appeared in one of the most prestigious ("hard-core science") journals, and it has the advantage of using a statistical methodology similar to that used by most who write in support of the opposite view.

How ironic it is that an illogical scientific finding should hold, if only because an illogical legal context structures the situation in a way that assures that perverse outcome.

Dwight Heath is Professor of Anthropology at Brown University, Providence, Rhode Island and a member of the AIM Editorial Board.