This deeply researched book, Convict Politics: From Utopia to Serfdom in Early China (221 BCE–23 CE), analyzes several interrelated injustices and inefficiencies in the early imperial Chinese legal system: overreliance on rigid performance standards and instruments of collective punishment led to the conviction of far too many people, which in turn led to a need for frequent general amnesties, which in turn led to proliferating harm to victims as scofflaws and abusers would cynically wait for an amnesty and have their criminal liability vacated. Although the argument builds (responsibly) on prior scholarship, I have never read a sustained and compelling critique like this one. Cai’s copious translations, from both paleographical and conventionally transmitted texts, are consistently accurate.

A marked strength of the work is that the author adduces dozens of passages from contemporary voices who understood all too well how their legal system was going awry. In the process, she shows that she is not naively applying modern concepts of justice. As an example, consider these observations on the failure of the amnesty system, by Wang Fu 王符 (ca. 90–165):

Quoting this complaint at the end of her chapter on amnesties is a deft rhetorical move on Cai’s part, for it well exemplifies the dysfunction that she has sketched in the preceding pages. If readers have been harboring the suspicion that her case might be overstated or anachronistic, they are now disabused by Wang Fu himself.

Another strength is Cai’s navigation of early imperial legal language, which is frequently opaque. What are the underlying Chinese terms for “convicts” and “convict laborers”? The answer is not simple; as Cai explains, the relevant terms overlap, but are not interchangeable. Lichen 隸臣 (for males) and liqie 隸妾 (for females), often translated as “bondservants,” refer sometimes to convicts, sometimes to slaves (p. 26, n. 9); similarly, tu 徒 and tuli 徒隸 often refer to convict laborers such as wall builders, but “could also encompass slaves who were traded between the government and private parties” (p. 34; see also pp. 17–18). One reason for the confusion is that the early imperial state would typically seize the family of a defendant and enslave his dependents upon conviction (pp. 90 and 98).Footnote ² Servile status was understood, in a fundamental sense, as a reflection of some formally adjudicated wrongdoing, whether one’s own or that of a husband or father.

Hence, one must always take heed of the context when analyzing such documents, and, in many instances, the precise source of a labor squadron is not clear. In addition to the convicts at its disposal, the state presided over a vast supply of statutory laborers (yaoyi 徭役), who were legally obliged to work under conditions set by the state, but were neither conscripted soldiers (zu 卒), who could also be mobilized for large projects, nor criminals. Then, in extraordinary circumstances, members of the free population could be levied for ad hoc service beyond their statutory duty (one term known from excavated documents is xing qianshou 興黔首, p. 32), although the state recognized the resentment fostered by such measures (e.g., p. 109). Cai handles this terminological farrago adroitly—relying, it should be noted, on a robust bibliography of East Asian scholarship. Due attention to her footnotes is essential.

Another difficulty posed by the sources is that one cannot always tell when they are being prescriptive and when they are being descriptive. That is to say, were the stipulated punishments applied regularly and literally? Take the case of Tui 㢈, a magistrate newly appointed to a territory facing a serious insurgency.Footnote ³ The troops sent to quell the rebellion had been routed, and Tui was expected to apprehend and prosecute the many deserters (who feared punishment for their defeat). When Tui requested leniency for them, citing the exigent circumstances, he himself was charged with “releasing and letting guilty persons go” 釋縱罪人 (Cai’s translation, p. 134). His defense and the subsequent judicial review are recorded in detail; in the end, he was convicted and sentenced to shaving and a relatively mild form of penal labor. Did this really happen? Cai is forthright about the problem: “The veracity of Tui’s story, whether it is a precise reflection of historical events or an embellished narrative designed to instruct officials in the spirit of the law, remains uncertain” (p. 136).

The clash between officials like Tui, who believed that the law is but a tool for the higher purpose of implementing humane ideals, and the inflexible bureaucrats who indicted him for what they regarded as connivance anticipates the last two chapters of the book, which address the rise of Confucian governance in the Han dynasty. To the ongoing debate over how to define Confucianism in this period, Cai adds the observation that ru 儒 officials (whom she calls “Confucians”) are distinguished, in multiple sources, from li 吏 (whom she calls “technical bureaucrats”): the former are characterized by their moral and literary education, the latter by their lawyerly and clerical skills (pp. 277–94). Thus, the conception of two mutually antagonistic administrative cohorts, motivated by incompatible premises, is not a figment; many people at the time must have accepted the picture as accurate, and to balance the commonplace charge that the li were mere automatons who applied the law without concern for mitigating factors, one should point out that the self-righteous Confucian practice of judging cases according to one’s private moral sense led to its own brand of abuse: now subjects could be punished without having violated any law whatsoever.Footnote ⁴

As Cai concedes, the extant sources on the hostility between ru and li belong almost entirely to the ru camp (p. 277). Historians today are not afforded much insight into the esteem of the two parties among the millions under their rule. The ru have the privilege of telling the story, but at the cost, in Cai’s harsh account, of bearing the most responsibility for the failure to achieve much-needed institutional reform (pp. 332–40).“The daily routines of a bureaucrat forced Confucians to compromise the principles and ideals they held dear” (p. 305), she writes. Perhaps this criticism is unfair because there were other, more structural, impediments (lucidly summarized on pp. 342–44). But the larger point is valid: whoever is to blame, reform did not come.