The name ‘Foucault’ was first spoken to me in dark, conspiratorial tones, as if he were a threat to the then-alluring project of combining Althusser's ideology-centred thinking and the British culture-and-hegemony thinking. Foucault, along with Weber, Popper, Berlin, and many others (the list was a tiresomely long one) had to be rejected, or so I was told. My mind was soon changed on that score. The exciting work of Barry Hindess and Paul Hirst (see esp. Hindess and Hirst 1975, 1977), who had worked through the Althusser and British Cultural Marxist possibilities more thoroughly than anyone else I had then read (or have read since), indirectly opened up the idea that Foucault was not only not a threat to the best-alternative project I shared with hundreds of others, but was the key to that project's success.At last, here was a thinker who could treat power seriously yet undogmatically, someone who could relate power to society without making it read like the script of a prison movie. I was hooked. I tried my best to understand (or to sound like I understood) all the methodological innovations that came with the Foucault package – ‘archaeology’, ‘genealogy’, ‘discourse’, ‘episteme’, and so on. My excitement reached its peak when, using these tools, Foucault appeared to have succeeded in crafting an entirely new approach to the study of government, under a term of his own invention, ‘governmentality’. But, as so often happens in life, the peak of excitement turned out to be the moment when doubts emerged. These doubts became stronger, eventually leading me to think that Foucault's works from this period too often pronounce and too rarely argue from the historical evidence.