Nicholas B. Dirks. 2001. Princeton: Princeton University Press. A richly researched account of the political structures (and historical structuring) of the "caste system" in India, Nicholas B. Dirks' Castes of Mind: Colonialism and the Making of Modern India is both more and less than an argument about the solidification of caste under British colonialism and the ethnographic imagination. On the surface, Dirks argues that caste is not the unchanging, historically frozen structure that we present it as (though it is not always clear who this "we" is). Through extensive archival work-that both reminds us of the critical players in the formation of modern caste ideology and asks us to question their assumptions-the author echoes the post-modern (and post Said's Orientalism) reflexivity of the last 30 years of anthropology, and compiles years of archival work into a comprehensive archeology of caste consciousness and its modern usurpation by nationalist Hindutva politics. Castes of Mind provides a helpful archival addition to contemporary post-colonial explorations of the anthropology of caste. But Castes of Mind also represents a political subversion of disciplinary genealogy, and a subtle catachresis of power and knowledge in the academy. By largely ignoring widely accepted definitions of varna and jati in the anthropological literature as changing, multiple, and contextual instances of the local and lived reality of social organization in recent decades, Dirks illustrates the ways in which editorial power can marginalize and create knowledge. Similarly, he reiterates the well-argued reflexivity of post-colonial critiques of anthropology by collapsing the colonial enterprise and the anthropological/ethnological/sociological project. In doing so, he traces this enterprise through ethnographic methods and anthropological knowledge production by colonial servants-in effect creating a kind of "anthropology of mind." At this level, Dirks' most recent book is a study in the symbolic and discursive politics of anthropological genealogy and the elusive boundaries between history and ethnology. Taken in concert with McKim Marriott's recent critique, the resulting argument is reminiscent, though at present not quite so prolific, as the Sahlins/Obeyesekere debate of the last decade. The book is useful in explorations of anthropological canons, maintenance of disciplinary boundaries, and an archeology of disciplinary knowledge/power in general. The book's 15 chapters (14 and a coda) are divided into four sections whereby Dirks explores the colonists' gaze through the archival record. In the introductory section he sets up his major claim as well as the confines of his critique. In short, he argues that caste is a politically modern construction that served to categorize and delimit previously more fluid social organization throughout India. While he does not argue that cast was a British construction, he does maintain that British colonialism "made caste what it is today." Here also, Dirks outlines the early contribution of Christian missionaries to the colonial project, and sets up his critique of mid-20th century anthropology of India-in particular its construction of caste-with commentary on major works from this era (Louis Dumont's Homo Hierarchicus, and early works by McKim Marriott). In a critique of anthropology and the ethnographic method, Dirks suggests that the discipline remains a handmaiden of British colonialism. Yet, his construction of anthropology is unclear, and he does not address the more contemporary work of post-colonial anthropology in his critique. Section two illustrates the ideology of caste prior to the 1857 uprising and weaves missionary and colonial reports together to argue that previous "ethnographies" of caste were not only more rare, but also less cohesive. Missionaries perceived both Brahmanism and the caste system to be a major hindrance in the work of converting souls to Christianity, yet their contribution-in terms of caste politics-to the colonial archive remains marginal. …