Hostage authorship and dirty hands: A reply to Tang

Idealism and lack of realism?

Starting with the third criticism, we found it a bit surprising that Tang does not acknowledge that we agree that the research community should take serious measures to counteract the prevalence of hostage authorship. In our original article, we explicitly argued that although researchers may face situations where they should be prepared to get their hands dirty, they also have a shared responsibility to discourage all forms of dishonest behaviour in science. We concluded that ‘the research community must make the effort to strike back against fraudulent authorship practices’ (Bülow and Helgesson, 2018: 8).

That being said, this is compatible with there being scenarios where researchers are faced with the dilemma of either refusing to include X, at the cost of not being able to carry out important research, or to unwillingly include X, against the recommendations of widely accepted authorship guidelines, thereby securing important benefits of this research. Whenever collective measures of the sort Tang envisions are lacking (which too often is the case), there will be situations where this gives rise to a moral conflict that should not be taken lightly. Doing so would be overly idealistic and would ignore the real-world circumstances in which researchers often find themselves: if researchers are seriously expected to pause their research until the hostage-takers have been rebuked by appropriate authorities, they will likely have to wait for a very long time. This is so not least because there often are no authorities with the relevant power over the hostage-takers (who, for instance, may be private companies with a monopolistic control over crucial reagents, without which the research cannot be carried out), but also because of limited resources in most research communities to deal with this kind of issue with any force. So while we agree with Tang that giving in to hostage-takers like X, because there is no better choice, is a failure on the part of the research community to uphold ethical values in research (Tang, 2018: 6), the problem remains what the individual researcher(s) should do in such less than ideal circumstances (Bülow and Helgesson, 2018: 7).

Beneficial research

Much of Tang’s criticism concerns our use of the ‘greater moral good’ as a justification for sometimes getting one’s hands dirty and giving away undeserved authorship positions in hostage-like situations. Tang’s point is that we never clearly define ‘greater moral good’, and as a result our defence of dirty hands risks inviting self-serving justifications and moral disengagement. We grant that we should have been clearer in this part. What we wanted to say was simply that giving away unjustified authorship positions in hostage-like situations can be justified by reference to the principle of beneficence, as discussed in biomedical ethics. In other words, the expected benefits of the research to participants, science or society ¹ may be important enough to justify the granting of hostage authorship, if this is unavoidable in order to carry out the research.

Even if it is granted that biomedical research can promote important moral goods, it can be utterly difficult to determine how much good an individual research project will bring about, not least because the outcome of the study is normally not known beforehand. This point is stressed by Tang, who holds that ‘it is premature for even the most wishful to compute, with any confidence, that the good at the end will justify its unethical means’ (Tang, 2018: 3). In developing this point, he writes:

In politics, ‘dirty hands’ may well be justified when human sufferings could be conceivably alleviated by somewhat unethical means, with the masses benefiting at the expense of a few. How this might be possible, as far as research is concerned, is difficult to envisage in practice. Even research work that represents a significant breakthrough rarely, if at all, provides a direct and immediate impact on the alleviation of human suffering. Moreover, in the scenario provided by Bülow and Helgesson (2018), and for many cases conceivable with a situation that culminates in ‘hostage authorship’, the research in question, or at least the critical part of it, has usually not yet been performed (with X withholding critical resources until a promise of authorship is made).

The complaint, then, seems to be that since it is impossible to determine beforehand what the benefits of a study will be, and thus whether they are sufficient to outweigh the moral costs of providing authorship to X, one should not accept devious demands that are the result of hostage-like situations. This is worth taking seriously. Yet, we believe that Tang is overly pessimistic about the chances of making this sort of estimation. Surely, for many studies, neither costs nor benefits can be well specified beforehand – there are uncertainties at both ends. But this does not mean that there is never any acceptable basis for estimating whether the benefits will outweigh the moral costs. To the contrary, careful estimations about the benefits of individual research studies are often made, and indeed required, for example by ethical review boards when approving research with human participants. In this case an expected positive balance of benefits over costs in terms of risk relating to participation is required for ethical approval. But if these estimations are deemed possible, why should we believe that similar estimations are impossible in hostage-like situations? If it is premature to ever estimate the positive impact of a study, it will also be hard to decide if, for example, the risk of harm to research participants is outweighed by its estimated positive effects. On the other hand, using history as a guiding-light, there seem to be good reasons to permit research on humans, considering the progress made in, for instance, the area of medicine during the past century.

Therefore, recognizing both the difficulties underlined by Tang and the progress made in many areas of life based on the benefits of research, we can sum up the situation by saying: we have good reasons to believe that there is a lot of benefit from research, but it may be genuinely difficult to see beforehand which projects will be beneficial and which projects will not. But we also have good reasons to believe that if we stop all research where it is not clear beforehand that it will be beneficial, we will most likely forgo most of the benefits that research may provide. One response to this decision situation would be to say: I see, but we nevertheless know too little to be able to make an adequate evaluation of the individual study – human research should therefore be stopped. Another response would be to say: since we have good reasons to believe that the collective efforts of many research areas will be beneficial, we should generally approve the research. This does not hinder us from turning down some projects because the benefit–cost ratio in those individual cases is too disadvantageous (per se or in comparison with similar studies). We suggest that the latter kind of reasoning is the most adequate, and that reasoning of that kind would be relevant also in the hostage-like situations we discussed in our paper. The fact that proceeding with the work in those cases would involve the allocation of undeserved authorship should then be an additional (negative) consideration relevant for the decision.

Practical necessity and moral disengagement

Tang also makes remarks about our notion of ‘practical necessity’. To be clear, what we meant by necessity was that the good effects of being able to carry out the research cannot be achieved in any other feasible way but the one involving getting one’s hands dirty. Tang’s remarks here are mainly off target. For example, he argues that our argument would allow a senior researcher to justify the inclusion of undeserving authors by arguing that it is a ‘practical necessity’ since it would promote the writing of a good article, which could prove critical in order to provide the group with means to obtain subsequent funding. Such behaviour has most likely nothing to do with practical necessity, but rather concerns bad excuses for disrespecting authorship guidelines. So while there indeed might be something to the general critique that moral sloppiness and disengagement might be a consequence of an arbitrary or lax use of terms like ‘greater moral good’, ‘proportionate to the end’ and ‘practical necessity’, this seems to be of limited relevance to our specific proposal, although a worthwhile standing warning against moral dilettantism.

Rule-following and best consequences

When trying to understand what, more specifically, Tang finds unappealing about our proposal, an idea that comes to mind is that it has to do with our idea that it might be right to overrule an (authorship) rule for the greater good. One way of thinking about this is that it concerns the well-known problem of rule-following and the breaking of rules. If the rule is there for purely consequential reasons, then the entire matter is one of consequences. If it is established that it is beneficial that the rule is respected, then there is a potential cost of breaking the rule in the individual case, even if the benefit–cost calculation for that immediate situation comes out positive, namely the risk of eroding the general acceptance and abidance of the rule.

Authorship rules, such as the ICMJE Recommendations (ICMJE, 2016), can be defended, at least partly, on consequentialist grounds: the overall consequences for science and society will be better if the rules are followed. We cannot spell out in any detail here why that would be so, but one important aspect is that since authorship is strongly tied to scientific merit, based on which researchers compete for positions and research funding, a consequence of rule abidance (the argument would go) is that academic research positions and research funding to a greater extent will be allocated to those most deserving in terms of research accomplishments and therefore likely to be those best fit to do research. If this were the only value of following the rules, it would not be particularly upsetting to see them disrespected for consequentialist reasons on some occasions. As already suggested, the remaining worry would then concern the risk of eroding acceptance and compliance.

Authorship rules may, however, be defended also in another way: based on some idea of fairness and justice. This idea is not primarily consequentialist, but rather deontologist. Here it might be seen as wrong as such to disrespect the authorship rules since they identify the deserving authors, entitled to authorship. Any misuse would, according to such a view, be bad regardless of consequences. Even if accepting this as a relevant ground for a defence of respecting authorship rules, this moral consideration might nevertheless clash with that of considering the outcomes of our actions, such as in a hostage-like situation. Again, a moral balancing of the relevant considerations might end with the solution that the rules should be disrespected in some situations. If the value of rule abidance is considerable, and the benefit in the individual case of breaking the rule is also considerable, then it might be a delicate matter to decide what ought to be done. In our article, we never said this would be easy. However, we suggested that disrespecting the authorship rules might sometimes be justifiable in the hostage-like situations we described. We still take this view to be correct.