One of the great joys of academic life is learning from others, and here I have in mind other academics as well as respondents in ‘the field’ in India. I have been lucky enough to work in several great universities, but I will always look back with affection on the London School of Economics (LSE) and in particular my time in its Department of International Development. Two colleagues, Tim Allen, an anthropologist, and Diana Weinhold, an economist, put on a course each year for our Masters students, which really encouraged everyone to think critically about some of the different methodologies they held dear. And it was fun in my own seminars to see really smart students with an economics background critique books which maintained that the failure of aid project A was evidence in itself of the failure of ‘Development’. Of course it wasn’t. Generally, the author was describing a principal-agent problem in a small project that mattered hardly at all to long-run income trends, even at a local level. Development is largely about the accumulation and distribution of capital. But it was also exciting to see the table turned, especially when a student with many and sustained experiences in ‘the field’ challenged the plausibility of some of the data sets that were being taken for granted in large-n statistical studies.