Sir, - You report (July 10th) that a tobacco company executive, Mr Ian Birks, Head of Corporate Affairs with the Gallagher Group, recently gave testimony to a Dail committee which included the claim that no causal link has been established between smoking and lung cancer.
It is an time-honoured stratagem of the cigarette lobby to affect a semantic scrupulousness (which turns out on examination to be conceptually bankrupt) when it comes to the use of the concept of causation in contexts where a lot of money is at stake.
The totally inapplicable deductive standards of mathematical inference are suddenly applied to one of epidemiology's highly-confirmed causal hypotheses, and we are reminded by Mr Birks that of course no causal connection has been proven between smoking and lung cancer.
Such mathematical standards of rigour can undeniably furnish us with a sense of proof which makes such a claim come out true. But if this is the sort of proof at issue then it should be noted that it is also lacking in the case of the hypotheses that overexposure to the sun sometimes causes cancer, that cancer sometimes causes death, or that the moon influences the tides.
Another favoured gambit of the tobacco lobbyist is to be found in Mr Birks's testimony. It is the surreptitious assimilation of the concept of "cause" to the much narrower concept of "direct cause". It is true that smoking most certainly doesn't directly cause cancer. In cases where smoking does have such an effect, it starts a chain of such direct causal links, a chain which leads from the inhalation of cigarette smoke to intermediate events involving changes in the chemistry of cells, culminating in the development of the cancer. Mr Birks's assimilation is fraudulent because the epidemiological, and in this case, jurisprudential claim at issue is that the connection between smoking and lung cancer falls under this broader concept of a causal relation, not the narrower one.
Since the time of Hume, it has been an increasingly widely-held belief among scientists that causal connections are not directly observed at all. The scientist, on this very popular account (which is not to say that other accounts make the tobacco lobbyist's claims any more tenable!), sees only a succession of events that may eventually lead her, after controlling for as many factors as possible, to posit a causal relation between two distinct events or event-types under investigation.
Given that physicists, biologists and chemists continue to come up with highly confirmed, though never incorrigible causal hypotheses of the most diverse sort on this methodological foundation, I trust that Mr Birks's implied rebuttal of epidemiological theories linking smoking and lung cancer will be interpreted by the Dail committee as more of the same expert-sounding opportunistic drivel that tobacco companies have been spouting for half a century. - Yours, etc.,
Desmond Hogan,
Department of Philosophy,
Yale University.
New Haven, USA.