« Tip of my tongue.... | Gene Expression Front Page | Balancing selection in color blindness? »
July 08, 2004

When axioms attack!

I've been poo-pooing the importance of axioms, explicit principles, in shaping societies recently. But I think there are clear cases where principles have been important in differentiating societies. One of the crucial matters is about HIV-AIDS in Africa. Christian blogger Martin Roth points out that "in general the HIV/AIDS rate is highest in those countries where Protestants and other non-Catholic Christians predominate" (see his graphs). Why might this be? I think there are issues of correlation and conflation of various causes, but, here are two points that are axiomatic differences between Christianity and Islam:

1) Islam accepts polygamy, traditionally Christianity does not, or at least the European derived Christianity that was introduced into most of Africa.

2) Muslims generally circumcise, while Christians often do not.

There are issues with these "axioms," in that the Koran does not sanction circumcision, but rather, it is an Arab tradition that has become synomous with Islamicization. Similarly, from what I gather the monogamy of early Christianity was the reflection of its growth in a Greco-Roman mileu where monogamy was the norm among the gentile peoples (European Jews accepted polygamy until the 9th century and some non-European Jews still do). African Christians point out that issues like monogamy are not as cut & dried in their Biblical grounding as Europeans like to assert, but, operationally, these issues are closely tied with adherence to these faiths, that is, they are close to being axiomatic and entail a change of behavior or act upon conversion (nevertheless, many African Christians do practice polygamy).

How does this relate to AIDS? In the case of circumcision, Christian groups that practice it have lower rates of HIV infection than those who do not practice circumcision, while Muslim groups all practice it. There is some evidence that circumcision in certan contexts can reduce the spread of HIV.

On the issue of polygamy, the evidence is thinner and the speculation more extravagant. This article in The Economist paints the general thesis:

In many places men used to be polygamous but faithful, albeit to several wives. In northern Mozambique, which is much more Muslim than other parts of the country, this is still typical. It almost certainly contributes to the much lower prevalence of HIV in the north. In most places, however, monogamy now supposedly obtains, though men persist in having several sexual partners. Since they marry only one (at most), the other women they sleep with must make ends meet as best they can.

Here you have a case where axioms and traditions are in conflict, the solution is to have a public face that is possibly at odds with the private reality (which has to change in reaction to the new principles). Martin Roth notes the high rates of infection in countries with "evangelical Protestants," likely the group most stringent about "keeping up appearences" and least sensitive to cultural differences.

Posted by razib at 06:30 PM