GORDON E. BARNES, ROBERT P. MURRAY DAVID PATTON, PETER M. BENTLER, and ROBERT E. ANDERSON The Addiction-Prone Personality New York: Kluwer Academic/Plenum Publishers, 2000, xvi + 320 pages (ISBN 0-306-46249-4, us$65, Hardcover) Reviewed by ROBERT O. PIHL Searching for a relationship between personality and addiction is a bit like attempting to navigate between Charybdis and Scylla. Definitional whirlpools and methodological monsters have realistically, not mythically, engulfed many reviewers and researchers. The end result has been that for the most part, the determination of meaningful conclusions has been functionally obviated. This indetermination would seem predestined given the historical vagaries in defining personality and addiction. Seen positively, the area of personality has been evolving, from being conceptualized by broad speculative explanations, to concerns about the stability and predictability of any personality concept per se, to the present era where traits are defined factor analytically. Finally, some consistency and convergency is occurring in this arena, which allows for the repeatable specification of independent variables. Definitions of addiction, on the other hand, dominated by allor-none categories of the effects of drug usage, often ignore highly differential typologies and reasons for a problem with a specific drug. Thus, the predicament and challenge is apparent: how does one apply a selected set of predictor variables to the very noisy category of addiction? In the first section of this text, Barnes et al. embark on a review, almost exclusively of the alcohol-personality literature using 1979 as the point of departure. This date marked the publication of a previous review by Barnes, in which the traits of stimulus augmenting-reducing, ego strength, neuroticism, and field dependence were featured. The literature to that date, viewed sympathetically, was replete with untestable theoretical ideas, methodologically limited investigations, and suspect and contradictory conclusions. These facts, recognized by Barnes, nonetheless resulted in a conclusion emphasizing these four traits and the initiation of a 20-year journey chronicalled in this text. Thus, the present review, in part, represents an effort to evaluate this 1979 conclusion in light of more recent findings. As one might suspect, this attempt though laudable is only partially successful. For example, some parallels that are drawn are arguable, while others do bring a knowing nod. Importantly, what the review and this focus do provide is an organization for a complex literature which then provides a friendly consistency throughout the text. This is a significant accomplishment, as without a workable framework attempting to comprehend this literature is akin to drinking from a fire hose. Also of note is the breath of literature reviewed with sections on high-risk and longitudinal studies particularly relevant because of their obvious methodological advantages over clinical data. Sections II and III present results and analyses from the ambitious Winnipeg Health and Drinking Survey (WHDS), with the first representing a longitudinal panel survey of personality and substance-use and the latter an "add-on" comparison to a clinical population. In the survey, participants were stratified by age, and considerable care was displayed in their selection and the choice of measures. Significantly lacking, however, is information about when these data were collected, although the reader finds generally favourable comparisons with census information of the late 1980s. Like most studies of this kind, the inclusion of certain measures is time-limited and thus there is some dissociation between current measures discussed in the previous section and those included in the survey. The use of multiple measures of drinking, however, is a strength, as is the proliferous use of structional equation modeling. The strongest personality predictor for alcohol problems (continuously defined) was a psychotism factor which also predicted problems at the second sampling, as it did the number of problems at the first sampling and alcohol use between samples. …
Read full abstract