Ofterdingen and Kropotkin

Very often one acts as if only that were evident what one can claim as evidence. This is mostly caused by the formation (such as it occurs in the infamous methodology courses at American universities) of a pre-given and learned cultural template. The evidence fetishism results in the fact that empirical works that do not automatically accept the same defintion of evidence as given meet with rejection. […]
Characteristically, entire “schools” or traditions can form around a single method of collecting data. For instance, certain contemporary ethnomethodologists will not admit anything else but conversation analysis – an analysis that they’ve reduced to a de-contextualised text – while they completely ignore all the data connected to the immediate context […] This “data” that is subsequently taken for the concrete in itself, is in reality the product of an immense abstraction […] but in this case an abstraction that is unaware of its own abstractedness. And this quickly leads to the monomania [of methods] of statistical distribution or of discourse analysis or of participant observation or of the depth-interview or of ethnographic description and so forth. The rigid subscription to a certain method subsequently defines ones belonging to a certain school in which its participants recognise one another, for instance, in their cult of “ethnography”, e.g. the ethnomethodologists in their exclusive love of conversation analysis. And if anyone combines ethnographic analysis with discourse analysis, this is seen as a flagrant break with methodological monotheism.
The same analysis would have to be done with regard to data evaluation methods, such as multi-variation analysis, regression analysis, path analysis, network analysis, etc. Here also pure monotheism dominates. It is possible that this is due to the fact that it provides an apparent methodological foundation for the arrogance of ignorance: yet even the most elementary sociology of sociology teaches that methodological condemnation verdicts are very often only a way to turn adversity into virtue […]
But I want to try to transform all these points of critique into a positive rule which quite simply says: Be wary of all sectarian rejections mantled in exclusive confessions of faith; instead aim to mobilise for each individual case all the methods that may appear relevant or practicable. […] I would almost like to say “prohibitions are prohibited” or: be wary of methodological guard dogs.

Pierre Bourdieu, Reflexive Anthropologie, Suhrkamp, Frankfurt/Main, 2006, pp.260-61, transl. mine.