"The 'data' (given) of research are not so much given as taken out of a constantly elusive matrix of happenings. We should speak of capta rather than data."