The diversity and scope of the issues discussed at the millennial ASA conference – including those collected in this volume – defy a comprehensive treatment in a mere ‘afterword’. Certainly, no useful purpose is served by reiterating and annotating the contents in the way in which Paul Sillitoe has already done most excellently in chapter 1. I shall, therefore, confine myself to four themes. The first concerns how we understand and reconstitute the boundaries between different kinds of knowledge; it is preceded by a historical preamble and concludes with a reflection on the relationship between science and indigenous knowledge.The second theme explores the advantages and disadvantages of contextualization; the third participation and power; and the fourth, the relationship between academic anthropologists and development work. These are, to some extent, selected idiosyncratically, mainly because they continue to pose puzzles for me personally, but also because they do seem, serendipitously maybe, conveniently to represent penetrating threads connecting some otherwise very different concerns amongst contributors. Even though the conference was large by comparison with most of its predecessors, there remained noticeable gaps in its coverage of relevant issues. Compared with their prominence in recent literature, there were, for example, no feminist issues or perspectives as such, there was surprisingly little about indigenous knowledge in green arguments, on intellectual property issues, on the transportability of knowledge, or on participatory mapping. If there is a substantive bias in my treatment, it is in the direction of natural resource management and local environmental knowledge, which seems to me to have forced the pace of recent developments.