Hans Olav drew a direct connection between the "disaggregationist" findings of early U.S. alcohol survey epidemiology and the emergence of calls for policy efforts to reduce per capita consumption. In this post I'm going to try to explain why the link Hans Olav posits is in fact highly problematic. I begin by defining "disaggregationism." 1. When U.S. alcohol survey researchers first braved to ask their general population samples about alcohol-related problems--i.e., ask respondents (Rs) questions about frictions with friends, relatives, and workmates connected with drinking, illness connected with drinking, heavy alcohol intake, evidence of alcohol dependency, etc.--the findings that came back from the field did not mirror the reality that an "alcoholism paradigm" might have predicted. Instead of finding that drinking problems were distributed into two distinct subpopulations-- one large normal group (with no or very low problem scores) and the other smaller group of alcoholics (with high problem scores)--the survey data brought the message that drinking problems were distributed in a smoother and more unimodal fashion: Lots of Rs reported positively on no problem measure at all or only one measure; but then quite a few other Rs reported positively on 2 problem measures; somewhat fewer, but still a significant number, reported 3 problems; and so on up the problem scale. 2. This simple finding had a number of corollaries: For one, it implied that individual problem measures were not highly correlated with one another. For another, it meant that prevalence estimation of something called "alcoholism" was going to be difficult to do with survey data because the unimodal distribution along the drinking-problems scale made the question of where one drew the line between "normal" and "problem" drinking essentially arbitrary. Survey data also revealed two additional embarrassing findings with respect to the then-prevailing alcoholism paradigm: (1) it was young men in their 20s (and not the middle-aged men populating most alcoholism treatment samples) who reported the highest rates of alcohol-related problems, and (2) longitudinal research conducted on these general population samples showed that such problems were often highly transitory over time, which ran counter to the image of high intractability that had grown up in clinical samples of alcoholics. Survey research, in short, painted a picture of alcohol-related problems sharply different from that derived from "clinical alcoholism." 3. I'm sure it was Robin Room who first coined the general term "disaggregation" to refer to this survey-based alternative picture of alcohol-related problems. To "disaggregate" is, of course, to unpack or separate out things that were once packaged in a single unit ("alcoholism") into a series of dissimilar (or potentially dissimilar), separate units ("drinking problems"). In a little 1972 paper that I've always regarded as the point of conceptual/policy departure for the "disaggregationist," "alcohol problems," or "public health approach" to alcohol problems (Room, "Notes on Alcohol Policies in the Light of General-Population Studies," _Drinking & Drug Practices Surveyor_, 6:10-12, 15, 1972), Robin elaborated a list of ten (listed "a" to "j") main implications and prospects of these simple but portent-laden survey-research findings. Robin produced a series of seminal papers further exploring the implications of this new perspective for alcohol-problems prevention (which he preferred to call "minimization") over the 1970s. 4. Item (b) on his 1972 paper's list read in part: "Our studies suggest that when the traditional unitary notion of 'alcoholism' is disaggregated, there are differences in correlates of different aspects of it....This suggests that a first requirement for any social policy having to do with alcohol is to define with some exactness what the target of the program is, since shooting at one target does not guarantee hitting all others" (p. 10). 5. Granting separate conceptual & policy status to individual "alcohol problemS," with an emphasis on the plurality of "problems," was one of the most important seeming implications of these survey findings. But it was not an entirely new finding and implication. For one thing, Harold Mulford was at about the same time contemplating similar survey data in his Iowa-based and national survey studies-- and coming to some similar conclusions. Moreover, Tom Plaut had elaborated a similar conceptual/policy perspective in in an influential report tellingly titled "Alcohol Problems [_n.b._]: A Report to the Nation by the cooperative commission on the Study of Alcoholism_ (Oxford, 1967). Plaut's dedication of this report to E.M. Jellinek--the celebrated scientific "father" of the modern disease conception of alcoholism --was entirely fitting not only because Jellinek had died while at work on the project (in 1963) but also because the alcohol problems perspective it articulated was in fact a re-expression of the framework Jellinek had worked out years before in the context of Yale's famous Center on Alcohol Studies in Howard Haggard's Laboratory of Applied Physiology. 6. There is something deeply intriguing about this continuity in scientific paradigms--from the Jellinek-crafted version of the "alcohol problems" paradigm of the early 1940s to its re-definition out of survey research findings in the early 1970s. My hunch is that it reflects some kind of deep, underlying, but tacit, affinity between Benthamite, secular, utilitarianism and the default stance of modern science when it is invited to address any complex social problem. Be that as it may, there is a single interesting feature of Jellinek's "alcohol problems" paradigm that I'd like to draw attention to here, namely: in Jellinek's hands the _plurality_ of alcohol problems provided a simple but effective rhetorical device for arguing against the Dry disposition to speak instead of "the alcohol problem"--i.e., in the singular--in making reference to temperance sentiment's chief problem focus, namely alcohol, per se. The late Mark Keller once recalled about this orientation: "We did believe in 'The Problems of Alcohol'--that was the title of Lecture 2, by E.M. Jellinek [to the Yale Summer School on Alcohol Studies]....It parted from older conceptions of 'the alcohol problem' which had filled the historical problem-oriented literature dominated by the temperance-antialcohol movement. We now were invited to consider not that alcohol was THE problem but that there were many problems in which alcohol was involved in a variety of ways" (quoted in Roizen, "Paradigm Sidetracked...," 1993, p. 16ff). 7. My point must be evident by now, Hans Olav: Ironic, isn't it, that the "same" alcohol problems perspective--with its pluralist and disaggregative approach--was _in Jellinek's hands_ an argument for _getting away from_ a focus on the control or proscription of alcohol and alcohol consumption whereas a half-century later and in the hands of public-health-model advocates, the "same" alcohol problems perspective provides the conceptual basis FOR refocussing attention on alcohol consumption and per cap consumption?! Indeed, how can such a circumstance be accounted for? 8. The answer to this puzzle, I think, is delightfully simple--it goes like this: It does not matter so much what the content or substance of the alcohol problems model is; it matters more what paradigm the alcohol problems paradigm is regarded as offering an alternative to at the time. In Jellinek's day, of course, the alcohol problems model was doing battle with the temperance paradigm (one that emphasized alcohol/alcohol consumption above all else) whereas in the early 1970s the alcohol problems model was doing battle against the alcoholism paradigm (one with a seemingly remarkable indifference to alcohol consumption in the nonalcoholic segment of the population). The important lesson in this, I believe, is that there is nothing inherent in the alcohol problems perspective that leads to great attention, interest, and policy focus on alcohol consumption and/or per cap alcohol consumption. 9. In fact, the alcohol problems model was very open-ended in this regard. Yes, it seemed in the 1970s to undercut the alcoholism model of alcohol-related problems (though even that simple assertion, in fact, was and is open to considerable argument!)--but beyond that, the disaggregationist perspective lent itself to a great many interpretive directions and possibilities. Even Robin, himself, may be credited with exploring and writing-about insightful analytical possibilities that LEAD AWAY FROM attention to alcohol, per se, consumption, or per cap consumption. E.g.: Robin elaborated an interesting interpretation of regional variations in alcohol problems in the U.S.--in which he argued that low-consumption normative contexts tended to produce lots of social disruption around alcohol but on the other hand relatively fewer long-term health consequences, whereas high-consumption normative contexts conversely tended to produce higher health consequences and lower social-disruption consequences. When it came to alcohol-problems and their relation to consumption, in other words, culture could take its choice! One of the most memorable of the lines of analysis that Robin (and others) explored in the early 1970s suggested that many of the things we had regarded as alcohol-related problems might actually be reduced by developing policy responses that would make the world "safer" for heavy drinkers. Robin's articulation of this perspective in a 1972 conference drew an unforgetable response from Griffith Edwards, who likened Robin's suggestion to (I paraphrase and approximate here, because I don't have the conference report handy) passing out earmuffs in Bedlam so as to muffle out the screams of the sufferers. To be sure, the same Robin Room went on to become one of the co-authors of the famous Bruun et al. report (_Alcohol Control Policies in Public Health Perspective_)--which we always called, simply, "The Purple Book"--in 1975. This work offered the foundation case for re-focussing attention on per capita consumption in coming years in the alcohol problems/policy arena--and a highly problematic argument it was, IMHO. The purple book's argument--with its emphasis on high-end alcohol consumers--looks somewhat dated nowadays. It was not an argument that relied directly on the disaggregative implications of the U.S. survey research findings. 10. The disaggregative survey implication offered still more conceptual possibilities, some of which lead even farther away from alcohol, per se, consumption, or per cap consumption as causal agencies in our minds. Some of these possibilities led in the direction of trying to hook-up with applicable or potentially applicable sociological theory. For example, two papers I wrote during this period offered basic attempts to come to grips with the theoretical possibilities of our survey data's disaggregationist patterning. The first was a paper with the unlikely title, "Deviance, Dogpounds, and Drunks" (_Drinking & Drug Practices Surveyor_ 7:21-25, 1973). In it I employed the metaphor of the problem of "stray dogs" to cast into sharp relief the differences between a structuralist/environmentalist model of "deviance" (in this case "problematic straying") and the individualist perspective proffered in models like the alcoholism paradigm (which would denote the "strays" problem in an individualist idiom, as "strayism"). My point was that it came quite naturally to us to explain the "stray dogs" problem in structualist terms, but came equally naturally to us to explain problems like "alcoholism" in individualist terms. The paper, then, tried to offer the metaphor of strays as a heuristic device for seeing what a structualist model of alcohol problems would actually look like. This, I note, was a model that tended to de-emphasize the factor of alcohol consumption, per se, and emphasize instead (for instance) the normative mismatches between the drinking norms of temporally adjacent social situations (e.g., the situation of a "party" followed by the situation of "driving home from the party"). (Not everybody in the office, incidentally, was altogether keen on publishing this strange, maybe even vaguely undignified-sounding analysis in our house journal--and I still admire its then-editor [namely, Robin] for braving to put it in nonetheless!) 11. In 1975 I presented a paper in Finland, the purple book's spiritual home, that used longitudinal data to explore the question of whether, and to what extent, we may actually have misnamed our "drinking problem measures" as these were used in alcohol surveys. My argument was that if the conceptual "glue" of an addiction model was, in fact, subtracted out of the conceptual rationale for drinking problem measures, then it wasn't altogether clear anymore why these problem measures deserved the name "drinking problems." I wrote, "many problematic behaviors which were labeled 'drinking' problems in alcohol surveys could be classified under many different headings by different observers: so, for example, a respondent who 'spends too much money on drinking' might be seen as mainfesting a 'budgeting' problem, an 'income' problem, a 'guilt' problem, or a problem finding other things to spend his money on" ("Drinking and Drinking Problems: Some notes on the ascription of problems to drinking," p. 3). I examined the longitudinal association between changes in drinking and changes in one drinking-related problem ("spouse problems") to examine the possibility of more _independence_ between actual drinking and this drinking-problem measure than our nomenclature at least tacitly implied. 12. There were other possibilities raised too: including that our lower-level problem scores were bascially trivial and unimportant--survey research "noise" obscuring the important "signal" of a rare few really problematic drinkers buried in our samples. Genevieve Knupfer, the intellectual source of much of the good thinking that went on in the early days of the Berkeley group, in recent years even painstakingly constructed an analysis addressing the altogether plausible possibility that top-group heavy drinkers in survey research studies were simply not drinking enough to really represent clinical alcohol dependence. She addressed this problem by aggregating the samples of many survey studies into a single super-sample that allowed her to look at the problem profiles and correlates of a very top-top group. 13. So, yes, Hans Olav, there is--and there was--a considerable leap between the basic disaggregationist implication of early survey studies of problem drinking and the (seemingly, according to your argument) derivative coceptual/policy interest in per capita alcohol consumption! Indeed, as I have mulled over your wonderful question when I had the chance over the past few days, I have come up with several additional points to offer you. For example, keep in mind that AS SURVEY analysts the variable "per capita alcohol consumption" would have been quite difficult for us to introduce into our analyses--for one thing, BECAUSE it was an aggregate measure and therefore wouldn't have fit well into the analytical habit we were familiar with. No, the purple book's consumption agenda--and its legacy of the Edwards et al. (Alcohol and the Public Good) volume in our own day--involves a much more complicated social history (and social construction) than the simple three-part argument your post presented. I hasten to add, Hans Olav, that it is you, rather than I, who is closer to conventional wisdom in this exchange. That's one of the reasons I'm frustrated with the current hegemony (or wouldbe hegemony) of the public health approach these days. There are a lot of critical shots that this new paradigm and the argument for this new paradigm really need to be peppered with. Thanks for providing a venue for suggesting just a few! Sorry this is so rambling & long--but it's late and I'm going to post it anyhow. Ron -- Ron Roizen voice: 510-848-9123 fax: 510-848-9210 home: 510-848-9098 1818 Hearst Ave. Berkeley, CA 94703 U.S.A. [log in to unmask]