Let's see--Where was I? Oh yes! The emergence of the survey-based critique of disease alcoholism in the mid-1970s. Let me backpeddle a little on what I was saying in Part I. Tho it's true that survey data--whether in RAND's clinical follow-up or Cahalan's general pop. samples--didn't seem to fit the clinical or disease-concept mold, it is *not* entirely fair to say that survey-research critique arose from a purely empirical, inductivist, and unthinking exercise. The "problem drinker" (PD) perspective that emerged from survey "epidemiology" (as the field was/is called) was often regarded as focusing attention on the not-quite-fully-addicted drinker--a sort of sub- or potential alcoholic--thus perhaps leaving the "real" alcoholic to continuing clinical jurisdiction. But the real difference between the "problem drinker" & "alcoholic" conceptualizations lay not so much in the *severity* or *degree* of putative dependence as in the eye of the scientific beholder. The PD perspective brought an inherent philosophical *nominalism* and a strong commitment to conceptual *operationalism* to alcohol-problems research. Both inclinations are in fact more or less inevitable in survey research work--where the enterprise itself tends to demand clear explication of concepts, translation of concepts into askable questions, and then analyses aimed at addressing the "lumping-togetherability" and relational patterning (or lacks thereof) of measured quantities. But even in the survey-research medium, the PD perspective can subtly reconverge with clinical & disease-concept platonism. All the survey-data analyst need do, for example, is regard a drinker's "problem-drinking score" as a *diagnostic* (rather than essentially descriptive/operationalist) exercise, and the trip back to an essentialism is half-completed. That sort of re-essentializing of the topic is apparent to some degree in almost all survey epidemiology of drinking problems but perhaps particularly so in Cahalan's *Problem Drinkers* (1970)--though, and ironically, that book contains one of the clearest explications of the key differences between the survey and clinical perspectives available in the alcohol literature.///Survey research's cardinal finding that alcohol-problems were not as "lumpy" as previously thought--i.e., (1) they didn't clump together in individual persons as much as clinical wisdom had led us to expect, (2) inter-problem correlations were lower than expected, and (3) (perhaps most importantly) even cross-temporal problem correlations were lower than expected--did not in turn create an "alcohol problems" perspective *de novo* in the alcohol research scientific community. On the contrary, such a perspective--which tended to foster a "disaggregated" view of "alcohol-related problems" (splitting these into different conceptual and policy domains rather than "lumping" them together under a single disease denomination)--already had a long history in post-Repeal American scientific thought. An "alcohol problems perspective" (APP) was richly evident, for example, in Thomas Plaut's (1967) report of the Cooperative Commission and even as far back as the early days of the Yale Center on Alcohol Studies (see *Alcohol, Science, & Society* [1945]). The emergent survey research critique of the disease conception of the mid-1970s, then, was of a piece with an older conceptual motif in American alcohol science--and by no means a wholly new thing under the sun. END OF PART II