Religion science: if you pay the piper, do you call the tune?
A hot debate has been taking place these last few days, in the comments section of Harvey Whitehouse's recent post on religion. Part of the dispute has to do with the way cognitive scientists working on that topic might be influenced by the money they get, particularly from a Christian foundation that hopes to promote a more favorable view of religion by funding research in that area, albeit in a nonintrusive way. What, everyone wonders, does funding of this kind do to the work it finances? Is Christian-funded research biased? Is it more likely to present religious people with a rosy mirror?
This question has been adressed systematically by a recent paper looking at broad trends in the sociology of religion (found via The Immanent Frame). The authors, David Smilde and Matthew May, looked at thirty years of religious sociology in five high-profile social science journals, and (among other things) they looked for correlations between funding types and 'pro-religiousness'. Articles were classified as pro-religious when they considered a religious independent variable and a non-religious dependent variable (say, how being baptized affects your likelihood of being in jail), and concluded that the religious variable had 'positive socio-evaluative effects' (baptized people are less likely to go to jail). As for funding type, they looked more precisely at the papers whose authors were funded by foundations with obvious Christian sympathies like the Pew Foundation, the John Templeton Foundation, the Metanexus Institute, etc. – compared with papers whose authors had money from other sources, and with papers not funded at all.
Bottom line: authors financed by Christian foundations are more likely to write pro-religious papers than authors who declare no funding at all, but the same applies to all financed authors, wherever their money may come from – governments, or non-religious private foundations. This is just one of many surprising findings.
The author's most important claim is that the last 30 years have seen the rise of a Weberian view of religion: religious factors are more and more treated as independent variables, with researchers looking at the way they influence social factors rather than the other way round. This trend seems to be driven by American research on Protestant Chrisitanity. It coïncides with a rise in 'pro-religious' papers (partly because making religion an independent variable is necessary for being 'pro-religious', so that an increase in Weberian papers automatically brings about an increase in pro-religious papers), but also, more recently, with a rise in 'anti-religious' papers (papers reporting unpleasant social effects for a religious variable). I am strongly tempted to see all this as an aftermath of the demise of Marxism, but the authors are unwilling to jump to that conclusion. Articles financed by Christian Foundations seem a bit more likely to treat religion as an independent variable – as indeed the official guidelines of many foundations (Templeton being one) encourage them to do. But stressing the autonomy of religion does not seem to lead these authors to a prettier view of it.
What about 'Religion'?
In his comments, Maurice Bloch is concerned that talk of religion as an abstract object might "confort advocates of soft religiosity". It turns out that one third of the papers under study dwelt on religion in the abstract (and not on a specific religious tradition or a mixture of specific religious traditions). Were researchers more sympathetic to 'Religion' than they were to Christianity, Islam, or New Religious Movements like Scientology? No, and 'Religion' was not more likely to be considered an independent variable than, say, Islam or Buddhism. Obviously this won't allay all of Maurice's concerns, but it is a beginning.
Will the Templetons be happy about these findings? Should they? I don't know. Anyway, the paper is a nice example of cool-headed sociology of sociology.