Let me begin with a confession. When I see that the latest issue of a law journal contains a reply to an earlier article, I tend to be overcome with a mixture of irritation and fatigue. On the occasions on which I peruse the offending article, the effect is generally similar to that attributed by Emile Zola to a bedtime reading of the Code Civil. Such rejoinders rarely illuminate the basic issues which the original writer tried to address, and all too oiften consist in very detailed points which can only be grasped by a reader assiduous enough to return to the earlier text. When a reply is in turn met by a counter-reply, the reader frequently has the frustrating feeling that she is witness to what is essentially a private debate which has somehow found its way onto the public page. In the light of such a confession, I clearly owe MLR readers some explanation for inflicting a member of this doubly unfavoured species upon them. 1 Jeremy Horder's rejoinder to my article on intention is an unusual member of the genre.2 First, it gives a clear and (generally) fair statement of the portions of my argument with which it is concerned. Secondly, and more importantly, it raises some significant points about criminal law doctrine and scholarship in its own right. Indeed, it does not really constitute a 'reply' to my article, for in relation to the two main theses which I was advancing, Horder is, as far as I can tell, in partial agreement with one and does not address himself substantially to the other. What he does is quite properly to point out that the relevance of intention in criminal law is broader than I explicitly acknowledged. In doing so, however, he in fact considerably expands the terrain over which my arguments can be played out. The purpose of this comment, then, is both to acknowledge (with some modifications) the legitimacy of Horder's assertion of the multiple roles of intention in criminal law, and to show how that recognition fails to support the conclusions which he draws from it in relation to my argument. My aim is not so much to reiterate or defend my arguments (something which in any case would not be possible in the space which the MLR has generously accorded me) as to clarify the points of genuine difference between Horder and myself so as better to allow our readers to judge the relative merits of our positions. As Horder explains, the main purposes of my article were, first, to illuminate the ideological role of conceptions of intention in legitimating criminal legal practices and, second, to argue for a loosening of the boundaries between criminal law and criminal justice studies. I shall return briefly to this second point below, but as it is the first argument on which Horder concentrates, this will be my principal concern. In essence, Horder accuses me of focusing unjustiE1ably on only one of the roles of intention in criminal law that of grounding and shaping attributions of responsibility and judgments of culpability. He argues that intention has at least three other significant roles: constituting wrongs, protecting autonomy