I begin with a conundrum—freshly devised, but I hope adequately old-fashioned in spirit. What’s the difference between guns and cigarettes? The answer is that guns fire and kill first, and then they smoke, whereas cigarettes smoke first, and then they fire and kill. But if this contrast is right, then gun smoke has an evidential value that cigarette smoke does not immediately have, as the casualty from smoking is yet to come and might therefore be, to some extent, a matter of conjecture. This leads to the further question that must be dealt with in one form or another in a world conference like this one: how can we sensibly assess the harm that smoking does? Only after coming to grip with this question can we go on to decide what needs to be done here and now.

This deliberative perspective may seem a bit too hesitant, especially in a gathering of this kind where the participants are basically convinced that the facts and the values involved in assessing smoking as a practice are all fairly well sorted out, and all that is needed now is urgent action. The connection between smoking and morbidity (including cancer, cardiovascular diseases and other ailments) is indeed well established, but the idea of social harm raises, as the critics of interventionist policies have discussed, other issues as well. It is important for tobacco control enthusiasts to recognise that a huge proportion of the population at large as well as a great many caring and kind intellectuals are not at all convinced yet that extensive tobacco control is the right way to go. The counterpoints are many, and they all need to be dealt with, for the battle for tobacco control is not just for new laws or new government policies but also for the …