Kristian Kristiansen, cogent critic though he may be, commits a category error of a depressingly familiar kind. It is a confusion which has led distinguished scholars such as Dumézil into error, and which, by allowing the conflation of such categories as language, ethnicity, race and institution, worked to the detriment of many groups and nations during the twentieth century, and now, no doubt, also in the twenty-first.

Nowhere does he define precisely what he imagines the term ‘Indo-European’ to mean. Following the perspective agreed by most historical linguists I take it to be a linguistic term, pertaining therefore to languages, members of the language family first recognised by Sir William Jones in 1786, and then further analysed and defined by subsequent generations of linguists. Through examination of the phonology, the morphology and the lexicon, all of which are well-defined and well-understood, it can readily be decided and demonstrated whether a specific language belongs to this family or not. So that when Hittite emerged in the early twentieth century from the archives of Hattusa, and later when documents in Tocharian were discovered and deciphered, the place of both those languages within that family could readily be agreed. Such a methodology is clearly not applicable to social institutions: it only works with words.