There are varying definitions of 'addictive drugs', one can find one definition from the AMA, another from the WHO, still others from various researchers & scientists. Some make a very broad definition such as 'any substance that can cause addiction', thereby defining the substance by the associated behaviors (or potential associated behaviors, or psychological reactions such as withdrawal). Other definitions (for instance, the one I referenced when I rather broadly stated alcohol wasn't per se addictive) are based on studies that have determined what percentage of casual users will continue to seek the drug, or what percentage of users (whether legal, or illegal) will experience withdrawal - physical or psychological.
I was referring to the fact that of all the substances typically thought of as 'addictive drugs', alcohol will only affect 10% of the users - a very low percentage similar to that of food, or sex - two things which most people couldn't;t define as 'addictive' - they are normal substances which, when abused, become addictive.
A casual user of alcohol has a very low risk of becoming psychologically addicted - in comparison, 90% of tobacco users become addicted within a few cigarettes - 80%+ of meth, cocaine, or heroin users become addicted. 40-50% of processed opiates (hydro and oxy codones) will - simply as a consequence of taking the drug - experience 'dependence' leading to addiction. By the definition I am using, neither LSD nor Pot are addictive, even though both are associated with dependence.
In terms of our current discussion, alcohol doesn't meet this level - oh it's certainly capable of becoming an addictive substance but this is behavior dependent, while in contrast, 90% of tobacco users will become addicted, regardless of behavior or mental state.
Taking this subject further would be an excellent topic for another post, where one compares varying levels of addictive potential, the difference between addiction and dependence, pseudo-addiction, and other behaviors.
So, in summary, I am stating that alcohol isn't 'addictive', chemically/behaviorally, in the same sense that, for instance, opiates and tobacco are - it's addictive in the same sense that food, or sex, or sky diving is - it can be used in an addictive fashion, it can certainly cause both physical and psychological withdrawal, but it is not medically 'addictive' in and of itself - many heavy drinkers can simply stop or moderate, however the same cannot be said of opiates, or tobacco, or cocaine compounds, for instance.
I realize my statement puts me in a minority position, but I do believe it's supportable (certainly debatable) by the evidence & scientific studies.
Does that help clarify what I meant?