Sellman's provocative paper makes several excellent (e.g. ‘different psychotherapies appear to produce similar outcomes’) and sometimes contentious (e.g. ‘compulsive drug seeking is initiated outside of consciousness’) points [1]. In this commentary, we offer some general reactions to Sellman's paper and then discuss an important area in the addictions field about which little is known.

One of the most important points raised by Sellman, also discussed by Orford [2], is that rivalries between explanatory models seem to have been more important to practitioners and researchers than to clients. Another important point raised by Sellman is that therapists can play a critical role in increasing a client's motivation for change. We found it puzzling, however, that Sellman's paper focused almost exclusively on substance abusers whose problems are severe. Such a focus raises two concerns. First, the references Sellman cites about alcohol problems are decades old, and there are no references to the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC). Data from the NESARC survey suggest that long-term stability of outcomes is achieved by many alcohol abusers, with about equal proportions of abstinent and low-risk drinking outcomes [3]. Secondly, Sellman's focus on substance abusers who present for treatment ignores the majority of individuals with alcohol [3,4] and other drug problems [5], as they do not seek treatment. In this regard, any comprehensive conceptualization of substance use disorders must apply to all cases, not just those in treatment. This point, made repeatedly by others, including Orford & Edwards [6], Vaillant [7], Cahalan [8] and Robins [9], is best captured in the following statement: ‘addiction looks very different if you study it in a general population than if you study it in treated cases’ ([9]; p. 1051).

Reference to the well-documented process of self-change or recovery without formal help or treatment [10] is also absent in Sellman's paper. The concern here is that many substance abusers who do not enter treatment recover on their own [10], and typically their problems are less severe (amount of substance used, consequences, or both) than those in treatment. Not including such individuals in a paper on ‘things known about addiction’ presents a skewed and incomplete view of substance use problems. From the standpoint of the health care system, recognizing the process of self-change and targeting individuals with less severe problems is a very important priority [11].

An important area in the addiction field where we feel rigorous data are lacking is about the effects of psychoactive substances on clinical populations. Our experimental intoxication research in the early 1970s provided us with an exceptional opportunity to observe and interact with individuals who had severe alcohol problems when those individuals were under the influence of alcohol. That experience greatly influenced our thinking both then and now. At that time, within a short period of time a handful of laboratories produced a great deal of very important information about the nature of alcohol problems (e.g. [12–14]). Today such research is lacking. Consequently, critical scientific information in several areas is lacking. For example, we do not know how decision making capacities are affected when substance abusers are under the influence of alcohol or other drugs. Unfortunately, our main source of knowledge about many important effects of drugs remains subjective, retrospective reports.

Declaration of interest