Among the comments to the last post, there’s this from Parmenion59:

So…if a cure for lung cancer is found, and the study has been funded through money from a tobacco company…the BMJ won’t publish said study?
Way to go BMJ.

Hmmm.  At least on the face of it, this looks like an important point – one that deserves a bit of unpacking.  We can begin by distinguishing between responses to this particular point, and responses to the general idea behind it.  First things first.

I’m willing to bite the bullet and admit without worrying too much that the policy of not accepting papers funded by the tobacco industry may mean that some research is not publicised.  There’s a small handful of reasons why I’m willing to do that.  One of them – admittedly the weakest of the lot – is based on the idea that it’s not wholly clear that much tobacco money really is directed at finding a cure for lung cancer, rather than firefighting other research about the detrimental properties of tobacco.  But that, as I say, is weak, based on suspicion rather than anything enormously substantial; and even if the hunch is correct, it’s merely empirical rather than anything conceptual.  Still, even if the hunch is wrong, it shouldn’t matter, because there’re stronger reasons.

One is based around the idea that there’s a special providence in the fall of a pipette – or, put another way, you can’t keep a good truth down.  If something is there to be discovered and is worth the effort, then it’ll be discovered sooner or later; if not by Smith, then by Jones.  And, because scientific progress is invariably a matter of the accretion of the work of several teams, all working independently and making minor discoveries, rather than one heroic person who would be solely responsible for The Cure For Cancer ™, the loss of one paper here and there probably won’t make all that much of a difference in the grand scheme of things. If that’s correct, then the idea that we might lose the cure for cancer is not all that compelling – not one about which we should worry too much.

A final reason is that, as I’ve said before elsewhere, I’m not persuaded that research is obligatory: it’s admirable, but not required by duty.  There’s a range of second-order arguments one might present here, but most relevant has to do with the benefits that research might generate.  If conferring benefit is a duty, that might (might at most, but I’ll let that pass for the moment) pan out as meaning that research is; but if it’s not, it won’t.  And while I think that some kinds of beneficence might be obligatory – rescuing people in the face of acute and immediate danger, for example – research isn’t going to create benefits of that sort.  What it can do is help with an incremental improvement in affairs; but I’m not sure that that’s obligatory.  Making people better off than they would have been is good, but not self-evidently so good that it’s blameable not to.  (I’ve got a fuller version of this argument coming out in a few weeks: hold on to your seats.)  Correlatively, a piece of research not being published represents at most the loss of an opportunity to confer benefit of one sort or another; but – again – unless that benefit is required, then a loss of opportunity is all there is.

Sp much for the particular point.  I don’t doubt that others’ll take issue with me on it, but there we go.  What about the more general idea?  What I mean by this is the idea that counterfactuals are relevant when it comes to solving moral dilemmas, and that the bigger the counterfactual difference, the greater the impact it ought to have on any given decision.

The question posed is, in essence, one about the opportunity cost of the policy.  Such questions are important in moral reasoning, especially if you’re a consequentialist.  Given a choice between doing A and B, what do you lose by rejecting the one you do reject.

Now, of course, there are certain counterfactuals that can be dismissed straight away.  For example, we don’t have to worry about

What if Smith’s tobacco industry-sponsored research found a way to save Bambi’s mother?

because it’s absurd.  Neither do we have to worry about

What if Jones’ tobacco industry-sponsored research found a way to reverse anthropogenic climate change?

because, though it’s not absurd, the chance is vanishingly small.  By contrast, a questions like

What if Brown’s tobacco industry-sponsored research found a way to neutralise the carcinogenic effects of smoking?

or

What if Robinson’s tobacco industry-sponsored research found a way to reduce the carcinogenic effects of smoking?

– both of which are questions the like of which Parmenion59 might have in mind – is probably not one that can be dismissed so easily.  This is not to say that a desirable outcome is particularly likely (we need take no position on that) but it doesn’t take such a huge leap of the imagination that it might be possible at least to mitigate some of the effects of smoking, and it’s not crazy to suppose that that’s the kind of research that tobacco companies would be interested in sponsoring, or that some progress might be made.

The general puzzle is this, then: given that it’s built into the nature of research that a given programme might come to nothing, a refusal to publish that leads to research not being done might make no difference at all to the welfare of the world.  But, at the same time, it might; and if there’s a decent chance that a given piece of research will provide valuable insights… well, what’s the standard of decency that applies?  How can we decide whether the gamble is worth the risk?  I’m assuming there must be literature on the subject – maybe people could make suggestions in the comments here.

Note that non-consequentialists have a different problem to solve.  Given that tobacco industry-sponsored research is, indirectly, funded by sales of a product that’s both deadly and addicted, there’s a “dirty hands” dilemma.  Would such research be so contaminated as to mean that we ought to forego any potential benefits?  Might rightness demand sacrificing potential goodness on the altar of integrity?  If you think that there’s a beneficence-based duty to research, that’d be potentially really tricky.

Again, suggestions in the comments about good papers on the topic would be most welcome.