This ambitious, but ultimately slight, book asks some interesting questions and attempts to stake out some important points about post-Mao Chinese state practice and social life. Its premise is that a “revolution” in the field of social knowledge from the 1980s onwards has led to the hegemony of modern statistical (rational) thinking in China today, a hegemony that despite being state-created has become the common sense of late 20th and early 21st-century Chinese and contemporary world social life. The author, Xin Liu, writes that the book intends to develop an “anthropological analysis of the a priori categories of statistical reasoning reborn and relived in the People's Republic [of China]” with the object of providing a “descriptive account of an epistemological rift – in both the mentality of governance and the sentimentality of life – that has produced a different outlook for the People's Republic” (p. xi). Liu's historical point is to establish an absolute break from Maoist to post-Mao China, a rupture that encourages one to see this moment as an entirely different epistemological and social era; meanwhile, his anthropological point seems to be to stake out a new ethnographic subject, not a tribe or a segment of a culture garden, but rather a body of knowledge, here government statistics glossed as “statistical thinking.” Both of Liu's claims could be interesting if properly presented and argued; neither is persuasive as articulated in this book. Indeed, the book quickly breaks down into chapters whose relation to one another appears juxtapositionally notional rather than narratively or conceptually logical.

Part of the problem with this book taken as a whole is the unacknowledged tension between the historical and the anthropological claims; part is the lack of a concept of historicity that might complicate his presentation of rupture; and part is how much arbitrary analysis and sheer verbiage the author throws at the discussion. To take the last issue first: quite bluntly, this is a poorly written and poorly edited book. The proliferation throughout the text of italicized phrases meant to convey importance betrays laziness; cute phrases and common clichés compound the effect. While isolated portions of the text present interesting discussions and materials, overall the book does not hang together as a coherent whole. More substantively, the major problematic of the book is derived most obviously from Foucault's analysis of the episteme and the genealogy of knowledge. Yet, Liu never gives an account of Foucault or of why and how he is relevant. The naturalization of Foucault prevents the author from considering weaknesses within the Foucauldian argument itself, leading him to adopt a superficial theoretical framework as common sense. Liu then throws many names at the text – Deleuze, Durkheim, Althusser, Lacan, Weber, Parsons, Popper and others – in what becomes a nominalist mash-up of 20th-century theorists. (There also are token references to Hegel and Marx, accompanied by a peculiar conflation of their respective dialectical theories.) None of this adds up, as the theoretical profligacy conceals a lack of rigorous theorization that forestalls discussion of the tension between the theory and the history or within the theory itself. Most disconcertingly, the theoretical claims come to descriptively reinforce the very common-sensical historical hegemony – statistical thinking as rationality defined – that Liu treats anthropologically.

This is achieved through the repeated invocation of the “revolutionary ruins” and the “Maoist nightmare” against which the post-Maoist moment of rational statistical thinking is presented. Despite the attempt in chapter six to denaturalize the before-and-after historicity by addressing the interiorization of knowledge regimes, for the majority of the book Liu takes the chronological comparative as truth. This makes the contemporary period the only possible moment from which to produce knowledge of the present (and indeed, his sources for the Maoist period are primarily oral histories collected in the late 1990s by Chinese sociologists). Accepting such a condition merely reifies the simplistic chronological historicity at the foundation of the production of the state-sponsored hegemony itself. Indeed, Liu's comparative – the Maoist past vs. the post-Mao present – not only naturalizes chronology from the perspective of the state (and, as he correctly acknowledges, sociology is a hand-maiden to the state), it also leaves out the large swathe of modern Chinese history (the Republican Era), during which statistics and sociology were first related to one another and worked out in the Chinese context. Not so new, then: the contemporary reinscription of this old knowledge regime as truth in the post-Mao present could be said to be a repetition rather than a rupture. What might this recognition do to complicate the historical genealogy woven by the author?

The anthropological claims also present problems. Marking out a body of knowledge (“statistical thinking”) as a new anthropological subject could be an interesting exercise; however, it requires some specification of theoretical and empirical boundaries, lest it become arbitrary. When “statistical thinking” and “the sociological fact” are linked to the problem of scientificity, the arbitrary nature of Liu's specification becomes apparent. Indeed, the author is acutely aware that the claim to science was as ubiquitous in the Maoist period as it is today. Clearly, the kind of science to which each appeals is quite different. However, without a direct discussion of scientificity as a major intellectual problematic of the 20th century, the lines Liu draws between Soviet-Mao era statistical thinking and that of today are unclear in their historical significance. This is primarily because Liu takes history as a chronology not as an explanatory logic and he takes the anthropological as reinforcement to an already-known historicity. This renders his history and his anthropology bereft of dynamism, even while they are populated by stock characters. Here, the anthropological becomes a conceit through which an already-known historical chronology is verified. The emplotment of Professor Dai as the hero of modern statistics against all odds in the 1980s (chapter three) with Professor Ma as the Marxist doomed to be a tragic figure in the 1990s (chapter four) rather than add flesh to the story merely succeeds in bloodlessly renarrativizing capitalism's own account of its triumph over socialism through its supposedly superior ability to fathom reality “objectively.” Talk about mirages! This is one to which the author never addresses himself.

In the end, all of this rehearses a familiar drama. That the author insists on telling this story in the idiom of the transition from “the management of people to the administration of things, which has required the statisticalization of society” (p. 177) sheds new light neither on China nor on Foucault. This is unfortunate as the topic is inherently valuable and the potential insights are numerous; however, Liu falls short of fulfilling his announced ambitions in this book.