The definition of disease is quite slippery, and changes over time. Did you know that medical science once recognized an illness called "drapetomania", the inexplicable desire of slaves to run away from their owners? Or that there is no definitional characteristic that you can give for the term "disease" that would not cover pregnancy as well?

Several previous posters have pointed out some of the problems with the authors' approach - you can tell someone with a mood disorder that they're just weak, or tell them God can save them, but that doesn't do much to help. 12 step programs are no better, in controlled studies, than no treatment at all. My own take on the authors' dismissal of brain imaging is that they just don't understand the science very well (my bona fides: PhD in neurophysiology, MD specializing in radiology, doing research on imaging of cerebral blood flow at a top hospital).

Aside from these definitional difficulties and flaws in the authors' argument, I think that there is one major positive aspect of calling addiction a disease. Public health approaches to dealing with drug abuse are enormously more effective, both in absolute terms and on a per-dollar basis, than the criminal justice approach (see Holland and Switzerland for models). Defining addiction as a disease is a step towards shifting scare national resources towards the much more effective public health approach. I'm glad that this shift is happening.