Charles Hartman has produced a broad and deeply researched study of Song court politics and procedures, unequaled in any Western language. In contrast to the chapters on imperial reigns in the Cambridge History of China, which attend to both personnel and policy, Hartman's goal is to show how the tensions between two modalities, Confucian institutionalism and technocratic governance, evolved and were managed during the Song period, in order to arrive at a correct judgment of the Song's place in China's history. His research yields many more valuable findings than this review will enumerate. Instead, I will consider the categories—the two modalities—that undergird Hartman's analysis, an analysis which too easily becomes procrustean, and which is foreshortened by a principled disregard of the society that his state sought to govern, the changing connectedness of one modal group to that society, and the real differences within that group over policy and values.

Hartman builds on his preceding book on Song historiography. There he argued that the Southern Song historians created a “grand allegory” that posited that the true nature of dynastic government was based on benevolence, which flowed from the character of the founding emperor and his successors but was thwarted by nefarious ministers. This metanarrative served the interests of Confucian institutionalists—literati who identified with the Qingli-period (1041–1044) minor reforms led by Fan Zhongyan and Ouyang Xiu, who shared the Yuanyou-period (1086–1094) rejection of Wang Anshi's New Policies, who were briefly accommodated in early Southern Song, who were attacked by Han Tuozhou at the end of the twelfth century—and who cast their opponents as self-interested petty men (xiaoren). A primary goal of Structures of Governance is an “analysis [that] strips away the moral pejoratives that Confucian historiography, as demonstrated in The Making of Song Dynasty History, attached to the disfavored member of each dyad—the xiaoren, talent, the inner court, and the imperial channel—and reframes each as a contributor to this alternative, technocratic modality of governance” (115). The Southern Song historians were thus the spokesmen for Confucian institutionalism against the technocrats.

Hartman structures his analysis sociologically, into three ideal types: the technocracy, headquartered in the monarchy (including rulers, females, and eunuchs); the Confucian literati who advocated an institutionalist approach to Song government; and the technocratic-Confucian continuum (13). Although the first two can be ideal types, I do not understand how a continuum can be. From a historical perspective, ideal types are not real. As Hartman notes, “a purely Confucian or a purely technocratic governance are theoretical abstractions … that never existed in actual practice” (133). Nevertheless, the book goes to some lengths to show that they did exist, and that we ought to analyze Song political culture in this—in effect—binary framework.

The traditional distinction between the inner court and outer court is a large part of this binary framework. The inner/outer court distinction was partly a literal one, as the two were spatially defined (Chapter 3). It also involved, as Chapter 4 explains, two distinctive avenues for making and communicating decisions. The public channel maintained by the outer court was the complex process by which imperial edicts were formulated and promulgated. But there was also the imperial channel of the emperor, which produced inner directives through which the ruler offered opinion and advice and attempted to steer bureaucratic action. The throne was crucial to both channels, but the outer court bureaucrats were limited to the public channel.Footnote ¹ For autocratic chief councilors, holding power required both controlling the public channel and forging ties with the inner court, but Confucian institutionalists aimed to limit the influence on policy of the imperial channel and bolster the public channel. Successful rulers had to accommodate the demands of the civil officials of the outer court without losing their own prerogatives.

Who were the “agents of technocracy” (Chapter 6) who worked through the imperial channel? There was the imperial family, 20,000 in the clan by the thirteenth century, who were 17 percent of military grade officials in 1213; the imperial women, not only empresses (nine of whom were regents in the Song) and consorts but also the female service personnel and female inner court secretaries; the eunuchs, in both military posts and infrastructural projects; the military servitors, who typically held posts in the civil administration but were supposed to do a tour in the armies as well; and finally the clerks, who totaled 440,000 at all levels by 1081. But would we expect all these groups to have a shared interest in supporting emperor-led governance? Taking the clerks as an example, why would their interests not lie with local society rather than the monarchy? After all, before the New Policies “hired service” policy, the provision of clerical services in the field administration was a service obligation of the local wealthy.

Who, then, were the agents of Confucian institutionalism? That “is a more complex and difficult concept” (17). Structurally they are the opposite of the technocrats: they are not aristocratic; they are officials not clerks, civil not military, court not palace officials, and male not female (17). Examination degrees are not a criterion: Hartman contends that most degree holders were not committed Confucians, defining “Confucian literati” as those men recruited through the exams “who professed a personal commitment to the precepts of Confucian moral education” (17). These are the same as “a committed, activist Confucian” who contributes to Confucian scholarship by commentaries on the Classics or prose showing “a serious intellectual engagement with Confucian ideas, principally as they relate to governance” or who make a “personal commitment to these principles as manifested in the official's own career choices” (121). The new Confucian ethic that emerged in the mid-eleventh century required “three fully unrelated skills”: knowledge of the Classics, Ancient Style writing, and administrative talent (230). Some literati thought these were closely related, but irrespective of whether literati thought of themselves as Confucian or not, hundreds of thousands took part in an examination education that included the Classics, Ancient Style writing, and state policy choices. It is not clear why they should not all be labelled as Confucian institutionalists.

Generally, Hartman treats the Confucian institutionalists as generalists guided by ideology; they think that, as gentlemen (junzi), their kind should dominate government, whereas the technocrats prefer talented specialists capable of practical accomplishment, what some Neo-Confucians belittled as shigong 事功. But Chapter 7 on the Northern Song technocratic state argues that “the Song technocratic state did not lack ideological, philosophical, and religious foundation,” most clearly evinced by Zhenzong (r. 977–1022) and Huizong (r. 1101–1125), the two emperors most associated with Daoism. A “structural coherence and continuity can be plotted in a straight line from the Zhenzong to the Huizong era” (185). Perhaps, but as regent in the 1030s Empress Dowager Liu turned away from Daoism, and Empress Dowager Gao in 1085–93 aligned herself with Confucian institutionalists. Hartman does not track into the Southern Song the coherent and continuous ideology he posits. Contrary to Hartman's treatment, it is not obvious that Zhenzong's patronage of Daoism—his letters from Heaven and the Feng and Shan sacrifices—in the face of his disadvantageous treaty with Liao and possibly his own succession at the expense of the founder Taizu line (it did not revert to the Taizu line until Xiaozong [r. 1162–1189]) was the same as Huizong's, whose self-glorification enhanced his authority over a civil service deeply divided over the New Policies.

The historical narrative in the Structures of Governance is focused on the rivalry between the technocratic and Confucian modalities. The Song begins as a technocratic state, with the founder increasing imperial authority at the expense of chief councilors. Under Renzong (r. 1022–1063), Confucian literati challenge this with some success (Classic Mat lectures, Imperial University, Bureau of Policy Criticism, academic institutes, etc.), but their agenda stalls under Yingzong, and Shenzong's “deputation of Wang Anshi to devise and implement the New Policies revealed the frailty of the institutional model” (207) as Wang had worked around established institutions in pursuit of the New Policies. “[U]nder the young and politically detached Zhezong, factional struggles consumed” the court (208) and any restoration of Confucian institutional norms was negated by the return to the technocratic mode under Huizong.

Hartman's discussion of the Southern Song is largely about court politics during Xiaozong's reign (1162–1189) within the framework of rivalry between Confucian institutionalists (beginning with Zhu Xi's memorials to Xiaozong in Chapter 9) and the technocrats. Chapter 10 is particularly interesting for its discussion of Xiaozong's shifts between the two camps and his effort to keep the support of both sides. In this account, politics plays out along the continuum between technocratic governance and Confucian institutionalism. Chapter 11 on the deeper structure of governance discusses a variety of issues involved in applying the technocratic–Confucian continuum to the Xiaozong reign.

Structures of Governance begins with a claim that “the scholarly community still struggles to achieve a consensus on the fundamental nature of Song dynastic governance.” If the consensus we should now reach is that Song at times achieved a kind of a balance between the technocratic and Confucian modes, and of inner and outer court interests, the reader is left to wonder how this makes the Song different from the Han, Tang, Ming, and Qing. Hartman blames this lack of consensus on “the general disinclination of social and intellectual historians to consider the state a significant actor in their recreations of Song society” (3)—a misrepresentation of Song social history. Indeed, contrariwise, this book simply ignores social history. A social-historical perspective does not ignore the state, it sees it in a larger context. The political scientist Yuhua Wang has written that “we need a better understanding of how society works before we can grasp how politics works.”Footnote ² When court officials have kinship ties to local elites rather than being embedded in national networks, he argues, they are disinclined to strengthen the state. The shift from national to local networks has been demonstrated in multiple ways. This does not ignore the state, as Robert Hymes's survey of Song social history in the Cambridge History makes clear.Footnote ³ Song Chen has documented the same shift by showing that in the 1040s prefects were embedded in a national kinship network, whereas in the 1210s there were multiple regional networks.Footnote ⁴ Ultimately, the examination system made it impossible for all but the highest civil officials to make government service the family business, whereas the inner court was populated by people whose family fortunes depended on the court rather than on their local lineages. Although it is true that some Daoxue philosophers were not interested in the state, I don't think intellectual historians ignore it.Footnote ⁵ An understanding of Song political culture is enhanced, and complicated, by asking why Sima Guang, whom Hartman sees as a Confucian institutionalist, proposed limiting access to the examinations to those who had the requisite number of recommendations from court officials,Footnote ⁶ whereas the New Policies agenda of Wang Anshi, whom he locates at the technocratic end of the spectrum, favored expanding educational opportunity (in 1109 over 167,000 students were being supported in state schools).Footnote ⁷

The binary framework of technocrats versus Confucian institutionalists leads to labeling rather than explaining, and lumping rather than differentiating. To say that Fan Zhongyan's reformers were Confucian institutionalists does not explain why that view emerged and gained sway. Because Hartman takes Sima Guang, the great defender of imperial prerogatives, whose chief councillorship was engineered by Empress Dowager Gao, and who called for fast-tracking the appointment of specialist officials, to be a leading Confucian institutionalist, Wang Anshi as his rival has to be labelled a technocrat. For the Xiaozong era historians, Sima was right and Wang was wrong, but the issue for them was not technocratic versus Confucian. Rather, they pursued the question that consumed Northern Song political thought: how far should the state intervene in society, economy, and culture? Wang represented the activist and expansionist policy to which Sima objected. Shenzong recognized that the two were the intellectual leaders of their generation and in vain tried to persuade them to serve together on the Council of State.

In the Southern Song, Zhu Xi sought to address local welfare with elite voluntarism and political culture with moral self-cultivation. As Hoyt Tillman has shown, Chen Liang found neither to be adequate.Footnote ⁸ His utilitarianism stemmed from his answer to the major question facing the Southern Song court: when to fight and when to make peace. The key issues in the Northern and Southern Song were different. Chen claimed to be just as much of a Confucian as Zhu Xi, and Wang Anshi claimed to be following the way of the first kings just as much as Sima Guang. There was not one Confucianism in the Song. Evaluating the Song state's place in China's history needs to recognize that the political does not have to be divorced from social transformations and intellectual trends.