Luc Bovens has written an interesting article entitled The rhythm method and embryonic death (alas not freely available, and sometimes inaccurately summarized in the press, so I will be quoting from the original) in the Journal of Medical Ethics (JME) on a hypothetical side-effect of the "rhythm" method of contraception. The article has numerous flaws, but the interest remains even when these errors are removed. (And since the JME has a detailed review process that it indicates can take up to two months, one wonders why the article was accepted a mere two days after submission. One fervently hopes that any particular political appeal of the article did not sway the Journal's editors from their scientific duty.)

The first three words of the title make one immediately wonder exactly what the author thought he was writing about: "The rhythm method…". That method was devised back in the 1930's, but is currently not much used — for example, no organized programs promote it in North America. Instead, various other forms of Natural Family Planning (NFP) are promoted (sympto-thermal, mucus, etc.), which have effectiveness rates of 98-99%. Did Bovens actually understand this, and intend his remarks to specifically apply only to the out-dated rhythm method, or did he use "rhythm method" as an inaccurate way of referring to any NFP method? I am unable to confidently determine this from the article. (And if the article was actually peer-reviewed, the peer-reviewer(s) also seemed not to be aware of the necessity of being correct and clear in this matter).

The abstract promises that the article will show that:

Given certain plausible empirical assumptions, the rhythm method may well be responsible for a much higher number of embryonic deaths than some other contraceptive techniques.

The embryonic death rate under consideration are those deaths (mostly unnoticed) which occur within a matter of days after conception. To be precise, the article actually shows that that death rate may be larger for that method than some others. But the author does not look at the size of the populations actually using the techniques he mentions, and so does not calculate any estimated numbers – just rates.

After a good description of the ways in which the contraceptive pill can actually directly cause the early death of conceived individuals, Bovens then says:

A concern for consistency has pushed advocates of the pro-life position into opposing all contraceptive techniques that cause embryonic deaths.

It is highly dubious that a concern for consistency has caused this. The pro-life position is based rather on not wishing to murder anyone at all. Bovens provides a reference to the website of a Christian evangelical church, EPM. The article referenced there shows no sign that consistency was a driving issue there — rather, the position was taken by that Christian author when he unexpectedly discovered that the contraceptive pill had more than a contraceptive effect, and could avoidably directly cause the death of a conceived individual.

Catholics might welcome this, since the official position of the church is that, aside from the rhythm method, no contraceptive techniques are permissible.

Catholics have had the same reaction that the EPM author had, since they also see the abortifacient action of the pill as directly causing death. (And there's another example here of Bovens' unclear intention behind his use of the term: "rhythm method". The method accurately called the rhythm method is certainly not the only method of spacing births permitted by the Catholic Church.)

What has gone unnoticed is that, if one is willing to make a few relatively innocent assumptions, then the rhythm method may well be responsible for massive embryonic death and the same logic that turned pro-lifers away from morning after pills, IUDs and pill usage, should also make them nervous about the rhythm method. 

The intention behind the author's remarks becomes a little murky here. The position taken by EPM — and many pro-lifers — that the pill is a seriously immoral choice of contraceptive (a position shared by Catholics) is not based on how large the numbers might be. Rather, the position is based on the observation that whenever an avoidable death of an individual is directly caused, this is murder. Just one murder is enough for an action to be regarded as a seriously immoral choice. So why does the author focus almost exclusively on the largeness of the numbers, and not on the ethics of single events and single individuals? (One would expect more attention on ethics in a journal of medical ethics.)

Bovens then makes a rough calculation (necessarily rough because the relevant parameters are just not well understood, or sometimes even completely unknown), and essentially shows that it is possible (given a suitable selection of parameters) that the average number of conceptions (per successful full-term pregnancy) that die before reaching full term could be larger when employing the rhythm method than when using some other methods — a conclusion which relies for its shock value on its counter-intuitiveness. Bovens supposes that if this were so, then this could be a reason for pro-life supporters to question whether the rhythm method should be used at all. Bovens somehow expects that pointing out this counter-intuitive conclusion will cause pro-lifers to question the rhythm method.

But he never clearly states the principle behind his reasoning. In fact, a radically vague assumption is made by him. He says:

If one is concerned about minimising embryonic death [...]

At first glance the goal of minimizing embryonic death seems obvious. Bovens certainly seems to think that all pro-lifers should accept that as a goal. Who could fail to support such an apparently so reasonable-sounding a request? But in fact it's an ambiguous statement, capable of being interpreted in a variety of ways. Bovens provides no statement of the way he thinks it should be interpreted. Every minimization is subject to constraints, but Bovens provides neither constraints that he expects should be followed, nor clearly states the constraints that he expects pro-lifers should follow.

For example, one could reduce the total number of early deaths of conceptions by refusing to engage in reproduction at all! That would be insane, and one could hardly expect pro-lifers to accept such a "solution". Alternatively, rather than total numbers, one could decide to reduce the rate of early deaths per successful pregnancy. But why? What is the ethical principle from which we should conclude that a successful pregnancy and one early death is somehow automatically better than a successful pregnancy and two early deaths? (Note carefully that pro-lifers would always expect that not even a single death would be directly caused — that is the mimimization that they would follow.) We are discussing here those early deaths for which the cause is not currently understood, and for which we have no means of prevention. Since we have no means of preventing such early deaths, they are not being directly caused. In fact, the only way of preventing the early death is to prevent the life from coming into existence at all.

So we come to a rather important point — why should an early uncaused death be seen as a bad thing? If the early death of the individual could be prevented, that would surely be something different, but that is not the situation being discussed. A life that comes into existence and dies an early but uncaused death is better than no life existing at all.

Bovens leaves no clue as to the mimimization he has in mind. His conclusion that:

One could simply conceive of this whole argument as a reductio ad absurdum of the cornerstone of the argument of the pro-life movement

is left void, since the views he ascribes to pro-lifers aren't held by them.