One example of implicit reference to this is in a discussion of Boorse by Elselijn Kingma in Analysis. She comments:
A final interpretation of design as Nature’s intent is closely related to ‘natural’. This should capture the idea that Nature intended there to be men and women, but it did not intend blind people. The latter are an accident, perhaps a ‘freak of nature’. Since I, and I suspect Boorse, reject an appeal to intelligent creation, the most obvious place to justify an appeal to design or Nature’s intent is evolutionary biology. This is not an attractive position for Boorse, however, who explicitly rejects the idea that evolution is relevant to physiological function and health (1976: 85). To evaluate this solution in detail goes beyond the scope of this paper, but if, as I suspect it must, this solution appeals to a difference between polymorphisms that are maintained by natural selection, such as eye-colour and sex- differences, and polymorphisms that are not maintained by natural selection, such as heart-defects, then it must at the very least dispose of the following problem. It must give a non-question begging account that explains why certain traits that are maintained by natural selection, such as sickle cell-anaemia, are nevertheless diseases. Since natural selection can enter into the explanation of both diseased and healthy traits (Sober 1980), this seems neither easy nor obvious... [Kingma 2007 italics added]
One of our students expressed the following eminently reasonable worry (in fact three good concerns) about the appeal to sickle-cell anaemia as a potential counter-example.
The reference to sickle-cell anemia isn’t entirely convincing to me. For one, this seems to be a fairly unique example, which is often the sole reference (at least that I’ve heard) for this type of argument about nature selecting disease. While, yes, one example is enough to disprove a general theory, are there any other examples of this conflict? I have a hard time thinking of conditions exactly like sickle-cell in this way, though there are certainly no shortage of traits originally selected by evolution which have become maladaptive in our present environment (insatiable hunger, metabolic problems, our capacity for salt-retention and resulting hypertension). The other argument I can see is that it’s actually sickle-cell trait (heterozygous allele) which has been selected by evolution; sickle-cell disease (homozygous allele) may be understood as an unfortunate consequence of that selection.
The problem for an evolutionary-selective account that Kingma flags stems from the thought that some things are maintained by natural selection but are nevertheless diseases. And hence the norm of maintenance by natural selection cannot be the norm whose failure is being tracked by, and hence sheds light on, intuitive judgements of illness, disease and disorder.
It seems to me, however, that this is rather the point of spandrels in thinking about natural selection. Spandrels are features maintained by natural selection. In other words they are selected. But they are not selected-for: the spandrel does not possess a function that would be appealed to in giving a natural selective explanation of fitness. This is implicit in our student’s comment that it is the ‘sickle-cell trait (heterozygous allele) which has been selected by evolution; sickle-cell disease (homozygous allele) may be understood as an unfortunate consequence of that selection’. It seems to me that this isn’t quite right. It is the sickle-cell trait (heterozygous allele) which has been selected-for by evolution. The sickle-cell disease (homozygous allele) is the spandrel: a feature selected but not selected-for. It occurs and is maintained because of the former and some biological laws and laws of probability, I imagine. Maintaining is thus disjunctive: occurring as either what is selected-for or what is selected but not selected-for.
This suggests that the misstep was to think of the relevant norm for assessing disease as being maintenance, instead of function or selection-for. But it leaves open whether, nevertheless, a spandrel such as sickle-cell anaemia serves as a counter-example to a general evolutionary-theoretic approach to disorder. It would, if there is no way to answer the challenge Kingma raises: It must give a non-question begging account that explains why certain traits that are maintained by natural selection, such as sickle cell-anaemia, are nevertheless diseases.
But it seems to me to be plausible that there are resources available to evolutionary theorists if they alter their gaze. Sickle-cell anaemia (homozygous allele) counts as a disease because it produces dysfunctions in the cardiovascular system. Short lived, too large red blood cells cannot perform the functions that explain, in natural-selective terms, the existence of the cardiovascular system. It is because of its effect on the performance of this that sickle-cell anaemia counts as a disease. Hence there are two natural selective characterisations in play: the disease is maintained as a spandrel because it rides piggy back on the heterozygous allele, which confers selective advantage in, let’s call it, the malaria-protecting system but itself, in the form of homozygous allele, is biologically dysfunctional in the cardiovascular system. One explains its prevalence; the other its disease status.
If so then it seems that spandrels can be accommodated without counting as counterexamples. The norm of disease might still be natural selective dysfunction as far spandrels are concerned. It is another question, however, especially in the case of mental illness whether there is a principled and non-question-begging method to pick out functions, spandrels and dysfunctions without merely relying on antecedent assumptions about what is and is not an illness, disease or disorder.
Kingma, E. (2007). What is it to be healthy? Analysis, 67, 128–133
