For perhaps two-thousand years the Han shu,Footnote 1 whose authorship is generally ascribed without question to Ban Gu 班固, his father Ban Biao 班彪 and his sister Ban Zhao 班昭, has formed an exemplary model of clearly written historical writing. China's officials and writers, teachers and students, have been proud to admire the style of this book and have often taken it as a model for their own writings, whether for historical works or other types of literature. In particular the compilers of the standard histories for the subsequent dynasties tried not only to follow the form of the Han shu but also to imitate the style of clear, formal prose that Ban Gu and his colleagues had adopted. We may ask to what extent the Han shu may be considered an original creation, how far the compilers deserve credit for their initiative, and how far they were willing, or even anxious, to incorporate the writings of others. In doing so, we may remind ourselves that the concept of a writer's personal and unique writing, for which he claimed complete responsibility, is not and has not been so fixed a notion in Asia as it became in Europe.
Some 150 years before the Han shu was written, Sima Tan 司馬談 and Sima Qian 司馬遷, who both held the position of Director of Astronomy or perhaps Archives (Taishi ling 太史令), had compiled their Shi ji. They took the tale of superhuman, or semi-divine, heroes, and of the men and women of the earliest times to approximately the end of the reign of Wudi (141–87 bce). Their work was not complete, lacking a chapter of imperial annals for that emperor from early times,Footnote 2 and, as received today, it includes interpolations by later writers such as Chu Shaosun 褚少孫 (c. 104–c. 30 bce). Certainly two copies of this great work existed at the time of Sima Qian's death, (c. 86 bce), but these were not publicly accessible. It was Sima Qian's grandson, Yang Yun 楊惲, before his execution in perhaps 54 bce, who sought to publicize the work, but permission to see it was not always granted, as was the case for Liu Yu 劉宇, nominated king of Dongping 東平 in 52 bce. The reasons given were that the work was not always consistent with the principles of the approved texts, that it included references to spirits and strange phenomena, and that its tale of intrigue and fighting in the Zhan guo years would not be fit reading for a king of the empire.Footnote 3 By the third or fourth century, we learn, ten chapters of the Shi ji had been lost.Footnote 4
The initiative for compiling a successor work was due first to Ban Biao (3–54 ce), nephew of Ban Jieyu 班婕妤 who was one of Chengdi's secondary consorts. It then passed to his son Ban Gu (32–92 ce), whose first attempt to write a history nearly cost him severe punishment. Those writings, completed perhaps during the reign of Han Mingdi (r. 57–75), aroused the charge that he had introduced his own private alterations to the history of the dynasty, and for this reason he was arrested. Thanks to the intervention of his famous brother Ban Chao 班超, whose successful military career began in 73, he was saved from punishment.Footnote 5 He probably wrote his second attempt during the reign of Zhangdi (acceded 75 ce) by which time he was certainly in favour. Author of several fu 賦,Footnote 6 he had also compiled the official account of the discussions of rituals and textual matters of 79 ce and he was even commissioned to write the history of the contemporary Eastern Han dynasty. Ban Gu's work, which covered the whole of Western Han and the Xin 新 dynasty of Wang Mang 王莽 (9–23 ce), was incomplete at his death. It fell to his sister, Ban Zhao, a woman of considerable literary merit, to complete the unfinished chapters of the Tables.Footnote 7
Such is the information that we may gather from the Han shu and Hou Han shu. Incomplete as the Han shu was at the time of Ban Gu's death in prison in 92 ce,Footnote 8 we may suppose that it was the preceding years that saw Ban Gu's final efforts to finish it, and we may ponder what sources had been available to Ban Gu, his father and sister, to tell them the facts or fiction of dynastic history and court life. Just a few parts of the Han shu, the imperial annals of Yuandi and Chengdi, are identified as the work of Ban Biao who at one point refers to his aunt as his source of knowledge.Footnote 9 Ban Gu would surely also have been able to gather information in the course of discussion with some of his elder contemporaries. Possibly he may even have talked with a few old men and women who had attended the court of Yuandi (r. 48–33) or, perhaps more probably, that of Chengdi (r. 33–7 bce).
The Han shu thus includes parts that were written by Ban Biao and Ban Zhao and as will be seen it incorporates the writings of other scholars and litterateurs. Where appropriate and possible, a distinction is made here between those known to be the authors of a particular part of the work. However, in the absence of definite and reliable information, the references to Ban Gu that follow carry the implication – but not proof – that it was he, as compiler, who may be taken as being primarily responsible for the contents and statements of the Han shu.
Possibly it was on the basis of what he knew of the reigns of Yuandi and Chengdi, and those of the succeeding emperors, Aidi (r. 7–1 bce) and Pingdi (r. 1 bce–6 ce), and the interlude of Wang Mang's dynasty, that Ban Gu's attitude to the past history of the dynasty was grounded. From his writings we gain the impression that Yuandi's reign was uneventful and undisturbed, while that of Chengdi laboured under a sense of crisis.Footnote 10
Yuandi had little taste for taking major decisions or indeed for participating actively in public life. The years of his reign were not marred by frantic military activity, for either defensive or offensive puposes; rather there are signs of a wish for retrenchment or even withdrawal from venturesome operations. Nor were the imperial palaces the scene of bitter antagonisms between the families of different consorts to the same degree as those of both earlier and later reigns. Officials expressed their views of how to manage economic problems – in response to criticism a number of measures were taken to reduce palace expenditure.
A very different atmosphere seems to have prevailed during the reign of Chengdi, for whom we have two contradictory opinions. Ban Biao wrote, on the basis of the hearsay of his aunt Ban Jieyu. He praised Chengdi for his personal deportment, adding that he “inspired awe like a god, and it might be said of his behaviour that it was as majestic as befitted the Son of Heaven”.Footnote 11 Ban Biao then remarked on the emperor's indulgent and depraved habits, and we read what is perhaps the most severe and damning criticism ever voiced against a reigning emperor of China in the protests of Gu Yong 谷永 in the years 29, 15 and 12 bce.Footnote 12 With a poor sense of responsibility, Chengdi was subject to the influence of some of the women in the palace, who practised abominable cruelties so as to secure their own ends. He deliberately chose the company of women of low grade, escaping from the palace and taking part in their drinking parties incognito. Gu Yong warned him that he was endangering the continuity of his dynasty.
Behind all this there lurked a realization of a highly critical situation. For all his frolics, Chengdi had failed to sire a successor to the throne and there were those who questioned whether the dynasty had in fact run its course and was due to come to an end.Footnote 13 Succession by two short-lived emperors (one an infant) who were chosen in response to the rival claims of two of the consorts’ families, could hardly have brought a sense of stability. Ban Gu would have been well aware of how members of the Wang family had secured commanding positions during Chengdi's reign and of how Wang Mang, at first a loyal supporter of the house of Liu, had shaken free of such ties and established a dynasty in his own name. The subsequent tale, and its warnings, of Wang Mang's end and the restoration of Han to stability, would have been even more fresh in the memory of Ban Gu's immediate elders.
Such were the dynastic and political changes that had marked the closing decades of Western Han and the years of the Xin dynasty. We may surmise that Ban Gu was well aware how far they could affect the conduct of public life, the choice of a government's decisions and the fortunes of those who took a prominent part in their adoption. We can only speculate how far they affected Ban Gu's view, as an historian, of China's past, or his assessment, as an official, of the strength, weaknesses and dangers of the contemporary situation. We may wonder about Ban Gu's reactions to a number of questions that would beset the mind of any observer of current affairs, such as the proper function of the emperor, the means of ensuring the continuity of the dynasty or respect for traditional values, as ascribed to the kings of Zhou and espoused by Wang Mang, against the more direct attitude to the needs of the present, of the time of Guangwudi (r. 25–57) and Mingdi (r. 57–75).
We may note a profound difference in some generally accepted concepts, as between the time when Sima Tan and Sima Qian were writing and when Liu Xin 劉歆 (46 bce–23 ce) and Wang Mang were formulating the principles on which they rested the authority of the Xin dynasty. By the time Ban Gu was reporting on the conclusions reached at the Bohu tong 白虎通 discussions, in Luoyang, the differences in the modes of thought were even more marked, as may be seen in three instances.
Until the close of Wudi's reign in 87 and for several decades thereafter three principal characteristics that lay beneath official thinking in later times had yet to be established and accepted. For many years, even centuries, it had been asserted that a ruler's authority depended on the charge he received from Heaven. That claim was always applied to the kings of the past, notably those of Xia, Shang-Yin and Zhou; it was only after Wudi's reign that the concept was adopted to apply to the Han dynasty.Footnote 14 Secondly, whatever beliefs or practices had been under Qin, it was not until Wang Mang that an emperor consciously and determinedly invoked the power of one of the Five Phases of existence (Wu xing 五行) to confer protection on dynastic rule.Footnote 15 Thirdly, in the time of Sima Qian very few officials cited the sayings of Kongzi to support their arguments; Dong Zhongshu 董仲舒 (c. 198–c. 107 bce), the one man who chose to do so, never held high office.Footnote 16
A wide difference may also be seen in the attention paid to learning, scholarship and ancient literature. Certainly positive steps to enhance their value in public life date from Wudi's reign, when the Taixue 太學 was established, certain texts were designated for study and some measures were introduced for testing individuals before appointment to official posts. However, it was not until some decades later that these decisions became operative in a significant way, as may be seen, for example, in the rise of the four different schools for interpretation of the Zhou yi 周易.Footnote 17 In addition texts of a new type were beginning to circulate, to be received with praise by some but denigrated somewhat later under the terms chen 讖 and wei 緯.
Between the accessions of Wudi in 141 bce and Yuandi in 48 bce a major change had been setting in, the meaning and direction of which may not have been perceived by all those who attended the imperial court or staffed the offices of state. A significant outlook of attitude would soon be affecting the view of empire and its place in the cosmos, the ideals that an emperor should strive to pursue and the character of official life, as compared with those that were acceptable when the Han empire was founded. It was a move away from the practices of Qin towards those that were believed to have characterized the kingdom of Western Zhou, and has been described elsewhere as a change from a modernist to a reformist point of view.Footnote 18 As historians, Ban Biao and Ban Gu may well have realized what had been taking place.
At its foundation the rulers of the Western Han had inherited and practised the way of life of Qin, much as they may have wished to avoid being seen to do so. The First Qin Emperor ruled his people by means of a carefully constructed set of institutions: he and his officials, such as Li Si 李斯 (?280–208 bce), administered the land with rigour and controlled the population by means of a severe and complex set of punishments. Despite the claims that Han simplified these laws and mitigated such punishments, it in fact made little change in the way in which the empire was governed. And the officials who served Jingdi (r. 157–141 bce) and Wudi (r. 141–87 bce) took steps to intensify the controls exercised over the working lives of the people in order to increase the resources on which they could call, and to divert riches away from private hands into those of the government. In this way Han had intensified the methods of its predecessor, as may be seen in the schemes to stabilize prices and to set up the government's monopolies for the production of salt and iron wares, and in the extension of Chinese authority in ever wider areas. Modernist policies were aimed at moving with the times so as to satisfy, or further, material gains.
By 48 bce a different attitude was gaining hold. This included an idealized view of the kings of Zhou and their administration, who were calling on ethical principles to bring maximum benefit to all those who lived below the skies. Some officials proposed to abandon some of the schemes adopted in order to bring system to bear on economic ventures; both in Yuandi's reign and later there were some who resisted the extension of Han authority among the non-Han peoples of the outlying regions. A call to reduce the extravagant expenditure of the palaces of the emperor and his consorts resulted in some economies, such as a reduction in the mounts kept in the imperial stables, or of the delicacies served at imperial banquets. Officials who oversaw these moves were attempting to reform the way of life by restoring practices that they ascribed to the ancient and revered kings of Zhou.
The nature and depth of these differences of approach are illustrated in a rich and fascinating book compiled shortly after Yuandi's accession and known by the title The Discourses on Salt and Iron (Yan tie lun 鹽鐵論). Officials set out their arguments, either defending the policies adopted in so far as they responded to the practical needs of the present, or else condemning them on the grounds of ethical principle and divergence from the hallowed past. Ban Gu certainly knew of the existence of this text.Footnote 19
Comparison of the statements issued by Wang Mang after the establishment of his dynasty with the memorials of men such as Chao Cuo 鼂錯 (executed 154 bce) illustrate this major difference of outlook. In asserting his claim to be ruling legitimately Wang Mang does not appear to have criticized Qin vigorously, but his determination to pose as a successor to Zhou runs in many of the documents that he issued. It is seen even during Han times in his claim that he had been acting as Zhou Gong 周公 in guiding an under-age stripling in his work on the throne. As emperor of Xin he adopted institutional terms, for example for some official posts, on the model of what he thought to have been the practice of the kings of Zhou.Footnote 20
Ban Biao had lived through some of these developments and we may perhaps dare to assume that Ban Gu was well aware of the intellectual and dynastic issues involved. By the time he compiled his history a strong and united dynasty had arisen and formed the immediate background to his writing. The earlier reigns, of Chengdi, Aidi and Pingdi, are usually regarded as a time of decline when the authority of an emperor and his government was weak and its leadership lay under question. But it was an age of notable intellectual activity that formed an equally significant element in the background in which Ban Gu was writing, though its importance is not always recognized. Yang Xiong 揚雄 (53 bce–18 ce) reaffirmed the values and teachings of the Lunyu, not yet designated as a text for instructing young aspirants to office. The younger of the two men named Jing Fang 京房 (executed 37 bce) had been explaining the meaning of the Zhou yi and seeking to show how abnormal and perhaps dangerous or disastrous acts of nature were to be related to the deeds of mankind. Liu Xiang 劉向 (79–8 bce) and Liu Xin 劉歆 (46 bce–23 ce), who also expressed their interpretation of such events, had collected copies of literature from all parts of the empire, creating approved versions of the texts, arranging them in categories for the imperial library and thereby pointing to the direction in which much of China's scholarly achievements would be facing. Liu Xin's name recurs in connection with technical advances, as seen in the production of the calendar known as the San tong li 三統歷.
Ban Gu and his father and sister would have been far more able to understand these writings and beliefs than we are today, and we may perhaps assume that he was better able to assess the nature of the religious controversies of the day. These had started, c. 30 bce, with the proposal that the cults of state should be addressed no longer to the Five Powers (Wu di 五帝) but to Heaven and Earth, and it was only by Eastern Han that this change was finally agreed. As noted, it was Wang Mang who first claimed that a dynasty existed under the blessing or patronage of one of the Wu xing 五行, in his case that of tu 土; and it would seem that Wang Mang was the first recorded person to have stated that Han had existed thanks to the power of another one, that of huo 火.Footnote 21
Such are the assumptions we may make regarding the extent of Ban Gu's understanding of past practices and his knowledge of the intellectual context in which he judged the decisions taken by emperors and their advisers. We are constrained by the absence of other persons' accounts with which to compare or test Ban Gu's own statements. We may believe that, thanks to his experiences, his choice of subjects to discuss or of which viewpoints to stress may have been inhibited by contemporary attitudes in high places. Did his relationship to one of Chengdi's secondary consorts affect his description of that emperor's reign? And what allowed him to set forth Gu Yong's stringent criticisms of Chengdi without fear of dangerous consequences? We cannot know in what ways the sources of information at his disposal were complete or defective, thereby prejudicing a balanced assessment of his subject.
As has been observed, Ban Gu compiled the official account of the conference called in 79 ce and would therefore have known which of the controversial problems of the day were of sufficient importance to require an official solution.Footnote 22 It was a time of uncertainty when different scholarly views were being advanced, some of which affected the compilation of the history of China's dynasties. The validity of certain texts, such as the Zuo zhuan 左傳, was subject to question.Footnote 23 There was a need to explain the meaning and application of terms such as San zheng 三正 and San tong 三統 which concerned the principles of dynastic succession.Footnote 24 Traditions regarding social distinctions differed, with five ranks being ascribed to Western Zhou times and three subsequently, and it was necessary to identify what was to be the exemplar.Footnote 25 A further controversy, which may date at least from the time of the Yan tie lun’s compilation but which has not drawn much attention, concerned the priority or emphasis to be given either to zhi 質, the substantial or realistic aspects of a situation, or to wen 文, the external covering or decor in which it was dressed.Footnote 26 Differences abounded over the interpretation of the wu jing 五經 (“Five Classics”); the need to regulate the calendar was settled by the introduction of the Si fen li 四分歷 in 85 ce.Footnote 27
We cannot know precisely what texts or documents were available in Luoyang for Ban Gu and Ban Zhao to consult. Their work arose in response to official orders, and we are told that Hedi (r. 88–106) gave orders that Ban Zhao should have access to the items held in the Dong guan 東觀.Footnote 28 It is possible that they were also able to consult records held in some of the lesser offices of government, but how far any of these collections would have covered the whole of the Western Han period is subject to question. Formidable difficulties would have faced those responsible for storing documents dated perhaps two centuries previously. In any case Chang'an had been severely damaged by fire in 23 ce and the destruction may have extended to the depositories of books and archive documents. While we are told that no fewer than 2,000 waggon loads of these were transported from Chang'an to the new capital, we are left to wonder just what type of material was preserved in that way, whether of official records or literary collections.Footnote 29 The present writer has suggested elsewhere that, for all these difficulties, at least some official records of Western Han, such as the lists of holders of nobilities, had been included and that Ban Zhao consulted some of them when completing the tables of the Han shu.Footnote 30 We can only ask what proportion of the literary works carefully collected by Liu Xiang and Liu Xin took their place in the waggons and ponder how the wooden rolls or silken spreads and scrolls, with their illustrations, survived the hazards of the journey.Footnote 31 It may also have been possible for Ban Gu and Ban Zhao, like Wang Chong 王充, to find copies of books that were for sale in bookshops.Footnote 32
However hard Yang Yun may have worked to see that copies of the Shi ji were readily available, it is likely that these were rare.Footnote 33 Copying a work of such length would have been no small undertaking, but although there is no direct evidence that Ban Gu and Ban Zhao had a copy on their table, we can hardly doubt that there was. However, several questions arise: we cannot know whether such a copy was identical with the Jingyou 景祐 print we have received, dating from 1035. In particular we may ask whether Ban Gu had access to the ten chapters that were missing from at least the third or fourth century, or to a copy which included the passages that Chu Shaosun had added; and we cannot know whether the sources on which Sima Qian had drawn were still extant. We need also to bear in mind the possibility that some chapters of the Shi ji as received today were not the original work of Sima Tan or Sima Qian, but rather chapters reconstituted on the basis of the Han shu and infiltrated into our present Shi ji.Footnote 34
Comparison of those duplicated parts of the two histories with few or no textual changes reveals features of an initiative and a critical approach with which Ban Gu handled the material for the first century of Western Han. These are seen, for example, in his treatment of economic issues. While, other than in chapter 129, the Shi ji devotes a single chapter to all the matters concerned (chapter 30), in the Han shu these are separated into two parts, one treating agriculture and the other coinage. There are also instances in which the Han shu preserves a fuller account of an incident than the Shi ji.Footnote 35 Ban Gu's critical sense is seen in the comments or appreciations attached to some of the chapters. The textual variants of some passages of the two histories may be revealing, the Shi ji tending to use particles more frequently than the Han shu. In somewhat loose terms, the difference may be described as that between a colloquial and a literary style.
We will draw attention to four features of the Han shu which are seen by comparing some of the treatises and tables of the two works and which may perhaps be termed characteristic. These are: the incorporation of tracts written by persons other than the compilers; the use of archive material; the creation of new material; and a critical view of the compiler's own times.
Ban gu the copyist
Ritual and music
The Shi ji treats these subjects in two chapters, no. 23 on li and no. 24 on music. The chapter on li includes long passages from the Li lun 禮論 of the Xunzi and the Yue shu 樂書 of the Li ji; the chapter on music draws very largely on the Yue ji 樂記 of the Li ji. The Han shu treats the two subjects in consecutive sections of the one chapter (Li yue zhi 禮樂志, no. 22). It traces their growth from pre-imperial times until Wang Mang and includes a highly critical view of the practices of the early reigns of Eastern Han.Footnote 36 Of considerable value, for which we must presumably thank Ban Gu, is the inclusion of the two sets of hymns sung at the religious cults, the seventeen collected as An shi fang zhong ge 安世房中歌 and the nineteen as Jiao si ge 郊祀歌.Footnote 37 Whether these hymns were in use in the time of Sima Qian may perhaps be subject to question.
Pitch-pipes and ritual
As against the two treatises of the Shi ji for lü 律 and li 歷 respectively (chapters 25 and 26), the Han shu handles the subject in two parts of the treatise named Lü li zhi 律歷志 (no. 21). The Lü shu 律書 of the Shi ji, which includes one section on military matters (bing 兵) and one on the pitch-pipes, is far from being free of problems, and different opinions have been expressed regarding its form and authenticity.Footnote 38Shi ji chapter 26, Li shu 歷書, includes a chronological account of dynastic history from mythological times until Qin and an account of the calendar up to the adoption of the Taichu li 太初歷 from 104 bce.Footnote 39 Much of the chapter consists of text that is now seen in the Da Dai li ji 大戴禮記, Zuo zhuan or Guo yu 國語, and we cannot necessarily assume that the chapter we have before us was available to Ban Gu.
Ban Gu tells us something about the authorship of the Lü li zhi. Footnote 40 He writes that Wang Mang had assembled some hundred men who were familiar with musical instrumentsFootnote 41 and ordered Liu Xin and others to submit proposals on the subject in considerable detail. Ban Gu himself had removed what was false in Liu Xin's writings and selected what was correct, and had written this up to form a fascicule (pian 篇). The treatise concerns a number of related subjects such as numbers and calculation, musical notes, and measures of length, capacity and weight. The second part of the chapter gives an historical account of the calendar, written by Ban Gu, to be followed by further text by Liu Xin. Liu Xin had been responsible for creating the San tong 三統 calendar introduced at the end of Western Han or early in Wang Mang's reign, and he writes about the three stages, or periods, of time known as the San tong.Footnote 42 The chapter also includes a separate piece of Liu Xin's writing entitled 世經 “Passage of the generations”. This is an historical account of China's dynasties which includes several pertinent features.Footnote 43 It also includes passages that cannot have been written by Liu Xin, such as its references up to 57 ce, over thirty years after Liu Xin's suicide.Footnote 44
Bibliography
Han shu chapter 30, Yi wen zhi 藝文志, is perhaps the most widely quoted of all the treatises of the Han shu, and perhaps the most significant in cultural terms in view of its influence on China's literary and scholarly developments. Ban Gu's preface runs through the sad tale of literary losses and the initiative of Wudi's time of trying to make good the deficiencies. He then moves swiftly to the orders given in Chengdi's reign (26 bce) that resulted in the collection of texts, the editorial work of Liu Xiang and Liu Xin's production of his ordered list of the items in seven categories, entitled Qi lue 七略. Ban Gu tells us that he had extracted the essential parts and presented them to his readers in the chapter.Footnote 45
Astronomy
There are major differences between the Tian guan shu 天官書 of the Shi ji (chapter 27) and the Tian wen zhi 天文志 of the Han shu (chapter 26), the work of Ma Xu 馬續 which was incorporated into the Han shu after Ban Gu's death. Unlike the other treatises of the Shi ji, there is no preamble or introduction in the Tian guan shu; there is, however, a long postface, which one commentator thinks may have been written as a preface, and to have been displaced.Footnote 46 The text starts immediately with an account of the Five Palaces (wu gong 五宮) of the heavens, each with its own stars, followed by treatment of the five planets. These are linked with the material elements that symbolize the Wu xing, given in the order of growth rather than conquest. The treatise then moves towards subjects that are more likely to be classified as astrological, such as relations between the heavenly realms and terrestrial areas, prognostication drawn from the behaviour of the sun and the moon, and matters such as the invisible energy pertaining to the clouds (yun qi 雲氣).
Ban Gu (d. 92 ce) can hardly have known of the studies of astronomy that had been completed by Ma Xu (c. 70–c. 140), brother of Ma Rong 馬融 (79–166), and inclusion of his chapter as the Tian wen zhi may have been attributable to the decision of Ban Zhao.Footnote 47 After a short preamble, the text reproduces the account of the wu gong from the Shi ji with little if any change.Footnote 48 In a final section it sets out dynastic changes from the end of Zhou until 2 bce, together with events that occurred in the heavens and very often with prognostications of what they foretold.Footnote 49 Unlike other chapters of the Han shu, such as that entitled Wu xing (chapter 27), the Tian wen zhi does not mention the named sources of these explanations of untoward events, but it does at times tell of the events that showed how the prognostication was fulfilled. Sima Qian had himself been engaged in observation of the heavens, and in recording what was reported, sometimes in competition with others. We do not know whether Ma Xu had had similar experiences.
For the sake of completeness attention is also due to those treatises of the Han shu where there is no immediate evidence that Ban Gu had deliberately included writings of others, except for parts of the Shi ji. The treatise of the Shi ji that recounts religious cults (chapter 28, Feng shan shu 封禪書) includes two parts. The first (pp. 1355–84) treats pre-imperial times, the Qin empire and the first eighty years of Han rule up to the death of the Empress Dowager Dou 竇 in 135 bce; the second (pp. 1384–1404) takes the account up to 104, the year of the highly significant adoption of the Tai chu 太初 calendar, whose very title suggested the idea of re-dedication and renewal. The second part is duplicated in the Shi ji as chapter 12, taking the place of what should have been a chapter of imperial annals for Wudi. It may be noted that chapter 28 uses the expression jin tian zi 今天子 or jin shang 今上 and may thus be deemed to have been composed during Wudi's lifetime. Chapter 12, which uses the posthumous title Wudi, was presumably compiled after that emperor's death.
Han shu chapter 25 Jiao si zhi 郊祀志 duplicates almost everything seen in Shi ji chapter 28, except for the preamble. As was correct, it uses the term Wudi. It fills out the Shi ji with additional text for the period that that work covers and takes the account up to the time of Wang Mang.
The writings of other authors
As it is given in the Xin shu 新書 of Jia Yi 賈誼 (c. 201–169 bce),Footnote 50 his essay Guo Qin lun 過秦論 (“Determining the mistakes made by Qin”) is divided into two parts. It is inserted in chapter 6 of the Shi ji (Qin Shi huang ben ji 秦始皇本紀) following a short preamble by the Taishi gong 太史公, but in a different order, whereby the second part, as in the Xin shu, precedes the first. The first part in fact consists of two sections.Footnote 51 That Chu Shaosun includes the first part elsewhereFootnote 52 raises the question of whether the essay had been included in chapter 6 of the Shi ji in Chu's own time. In addition, the historical sequence is curiously mishandled in this chapter of the Shi ji, where an account of the much earlier period (beginning with Xiang Gong 襄公 777–766) follows Jia Yi's essay. The Han shu includes no more than the first part of the essay.Footnote 53
The Han shu also carries a pronouncement of Jia Yi on matters of the economy, which is not included in the Shi ji.Footnote 54 It also includes the somewhat lengthy memorials submitted by Chao Cuo on military and other concerns, and his response to Wendi's rescript of 165 bce.Footnote 55 The Shi ji includes the records of medical cases treated by Chunyu Yi 淳于意 which are not seen in the Han shu.Footnote 56 By contrast Ban Gu includes the three imperial rescripts and responses of Dong Zhongshu 董仲舒 of c. 134 bce,Footnote 57 and compositions of Dongfang Shuo 東方朔 in dialogue form.Footnote 58
Some of the writings of Sima Xiangru 司馬相如 (c. 179 to 117 bce) are presented in identical or near-identical form in both histories.Footnote 59 In the Shi ji the chapter is placed, somewhat curiously, along with the biography of the kings of Huainan 淮南 (chapter 118), to follow the chapters on non-Han peoples and their leaders. In the Han shu the biography of Sima Xiangru is placed very reasonably among those of men who were prominent in Wudi's reign. The inclusion of some of Sima Xiangru's writings in the Han shu has not raised any doubts, as far as is known to the present writer; but it has been suggested that chapter 117 of the Shi ji, far from originating from the brush of Sima Qian, is a later addition composed on the basis of the biography in the Han shu.Footnote 60
A clear instance in which the compiler of the Han shu states that he drew on other writing is seen in reference to the famous debate of 81 bce. He recounts the circumstances in which the debate was held and records how Huan Kuan 桓寬, a specialist in Gongyang 公羊 teaching, drew up his account of it in Xuandi's reign, with some amplifications.Footnote 61 According to Ban Gu, Huan Kuan called on information supplied by Zhu Sheng 朱生 of Runan.Footnote 62 In fact the whole of this passage of the Han shu, with its following comments on the participants of the debate, is largely identical with text found in the final pian of the Yan tie lun as received.Footnote 63
Ban Gu's readiness to include the compositions of other authors is seen where he covers the last half of Western Han, e.g. in the Wang ming lun 王命論 of his father Ban Biao, Liu Xin's 劉歆 letter of bitter criticism to the academicians and the writings of Yang Xiong 揚雄.Footnote 64
The use of archive
Chapter 60 of the Shi ji, entitled San wang shi jia 三王世家, is perhaps the clearest case in these early histories of the compiler directly copying material that had been framed in one of the offices of the imperial government, apparently with little change. There is no preamble to the documents; there is a final comment from the Taishi gong and text which is appended in the name of Chu Shaosun. The chapter is in fact a file or dossier of fifteen decrees and memorials that concerned a controversial issue, namely the investiture of three of Wudi's sons as kings in 117 bce. That the dates mentioned are given according to the full official form suggests that these documents had not been heavily edited before their insertion in the history.Footnote 65 Chu Shaosun wroteFootnote 66 that he had been unable to find the text of the San wang shi jia in the biographies of the Taishi gong, despite the statement that the text was available for inspection. He had been able to acquire the documents from elderly persons who loved the history of the past, and he had included them in the Shi ji.Footnote 67 This material is not included in the Han shu.
Two considerations affect the extent to which archives, i.e. documents compiled in the course of transacting an official's duties, were available to the compilers of the Han shu. First, one must ask how it came about that documents which had originated from officials who had worked in Chang'an made their way to Luoyang; and secondly, the form of certain chapters of the Han shu merit attention.
The contents of the two-thousand wagon loads of documents that were said to have been transported from Chang'an to Luoyang are described as jing die bi shu 經牒祕書. We have no further information about their contents and we can only assume that these terms included such official documents.Footnote 68
The particular chapters of the Han shu that are in question are those that are titled and classified as tables (biao 表). Two forms of table are included in both the Shi ji and the Han shu. In one the entries in each of the columns are to be read vertically, each following the other in chronological sequence.Footnote 69 Such a form suits the accounts of events such as the inheritance of the nobilities; it provides for successive entries for the foundation and closure of the nobility, including the deaths of the nobles, at whatever point of time they occurred and irrespective of the length of the intervals that separated those events. The second form of the tables sets out consecutive columns that are read from right to left; each one identifies a month or a year for long sequences of time, with no month or year being omitted.Footnote 70 This second form of the table thus allows for entries to be made at any point or moment in time. It is read horizontally so that, for example, the fortunes of a particular king may be followed throughout the passage of time from month to month or from year to year. ElsewhereFootnote 71 the present writer has suggested that in inserting the vertical tables both in the Shi ji and the Han shu, compilers from Sima Qian to Ban Zhao were working directly from lists drawn up by officials in the course of their duties; whereas it was the compilers themselves who composed the horizontal tables, based as these were on a somewhat different concept of historical record.
A further instance in which the compilers may have drawn directly from archive may be seen in the entries for the many communities who lay athwart the trade routes of the Western Regions. Such entries form the main structure of Han shu chapters 96A and 96B. They often begin with a straight statement of information, perhaps of a statistical nature, such as the number of inhabitants and arms-carrying men in the community, the titles of its officials, the distance from Chang'an or from the seat of the Protector General (Duhu 都護)Footnote 72 with geographical or topographical notes and remarks about the lifestyle of the inhabitants. Where it is relevant, the Han shu includes a full account of the relations, peaceful or warlike, that a community enjoyed with the imperial government and the tale of how its own rulers rose or fell. It has been suggested elsewhere that for the statistical and factual information at least, set out as it is in a regular formal manner, Ban Gu was copying the reports that the Protector General submitted to his government.Footnote 73
The main body of chapter 28 of the Han shu (“Di li zhi” 地理志) consists of a list of the major administrative units of the empire (jun 郡 and guo 國) and their constituent counties (xian 縣), nobilities (hou 侯) and estates (yi 邑), for the year 2 ce, together with the figures for the households and individuals registered in the commanderies (jun) and kingdoms (guo).Footnote 74 It may be suggested that this chapter was copied from a list of the administrative units of the empire made for official purposes.Footnote 75 As received the chapter includes intercolumnar notes which are attached to the names of commanderies, kingdoms, counties or other units. They tell of: the circumstances and dates of foundation; the changes of name that had taken place, including those assigned by Wang Mang; geographical features, mainly mountains and rivers; famous sites or buildings, such as palaces or sites of worship; the situation of the agencies set up and administered by the central government, such as those responsible for the salt and iron industries; the presence of a controlled point or pass (guan 關); some account, occasionally, of the distances involved; and historical incidents, such as journeys undertaken by notable individuals in Chun qiu times.
Elsewhere in the Han shu, intercolumnar notes are preceded by the name of their authors, and the absence of such a notification here may suggest that these were written by none other than the compiler. Alternatively they may have come from a later hand. In so far as there is no chapter in the Shi ji that sets out administrative and geographical information in this way, Ban Gu deserves credit for adding to his history and for supplementing the bare lists of the units with both a preamble and further lengthy material at the end of the chapter.
Ban Gu the creator
Chapters 23, Xing fa zhi 刑法志 and 27, Wu xing zhi 五行志 of the Han shu show something of Ban Gu's initiative in creating essays of a type that, as far as we know, had not been seen previously.
Xing fa zhi
None of the traditional Chinese commentators or Western writers have suggested that this chapter was not composed by Ban Gu. There is no corresponding chapter in the Shi ji and no comparable type of writing in the materials that survive from Han times. The title, which may be translated as “Treatise on the system of punishments” in no way suggests the existence of a codified body of law that was based on accepted principles of justice, or that set out the function of the laws as a means of controlling the excessive actions of a government. Following an introduction, a section of the text concerns the army and military matters from ancient times until the start of Han. A third part surveys, perhaps in a general rather than a specialized manner, the punishments as known from early times until the end of Western Han. Little if any attention is paid to any others that might have been introduced during the Xin dynasty. The chapter includes a considerable amount of citation from texts such as the Xunzi, Zuo zhuan, Zhou li and Lun yu.
The combination of military matters and punishments for treatment in one chapter has drawn comments from scholars such as Wang Mingsheng 王鳴盛 (1722–98) onwards, with Hulsewé concluding that in choosing to do so Ban Gu was “following ancient precedents”.Footnote 76 There is some overlap with those small parts of the chapter of the Shi ji on Lü (ch. 25) that concern military matters; but it cannot be known whether Ban Gu had before him the long lost treatise of the Shi ji known as Bing shu 兵書.
The discoveries of manuscripts since 1972, including some that were not available to Hulsewé, may give us some idea of the type of material on which Ban Gu was able to call, which by his time this had grown extensively.Footnote 77 He may have had a copy of the corpus of decrees of Western Han, such as that of which fragments were found at Juyan.Footnote 78 He may also have seen examples of other type of document found at Shuihudi 睡虎地 and Zhangjiashan 張家山, such as texts of the Statutes (lü 律) and Ordinances (ling 令), and case histories. The latter show how the Statutes and Ordinances were to be applied, or they set out to show whether their prescriptions were relevant to the case in question. Ban Gu may also have seen lists of such problems with guidance for officials on how to solve them.Footnote 79
From the outset of the Han dynasty it was claimed that the new regime had simplified the provisions of the Qin emperors and modified the severe punishments they imposed. Such had been the beneficent action of the emperor who would be known as Han Gaozu, but the claim was more fictional than factual. The recently found documents make clear that Han's system was no less complex or detailed than that of Qin, and that its punishments were no less cruel. Ban Gu had once been charged with perverting the account of history and he might well have thought it wise to avoid presenting a detailed catalogue of all the practices that Han pronouncements laid down. Han shu chapter 23 should be evaluated not as an attempt to list the provisions of the Statutes and Ordinances that affected the punishment of criminals; it was written more as an historical review. Whatever his precise intention may have been, credit is due to Ban Gu for innovation and for setting a precedent that would be followed in the Wei shu, Jin shu and Sui shu.
Wu xing zhi
The Wu xing zhi (Han shu c. 27) is by far the longest of the treatises, being divided into five parts. While the chapter is entitled wu xing its contents are not organized on the basis of that theory or openly divided into those categories. There are indeed references to the Five Phases but these do not dominate the concepts or structure of the text.
The treatise sets out incidents of strange, abnormal types as recorded from the Chun qiu period until the end of Western Han, together with explanations of their occurrence, the warnings that they gave and the lessons to be learned. The preambles to some of the sections in which these events are recorded discuss the subject in general terms, often referring to the Yi 易 or the Hong fan 洪範. In general there is no regular or rigid assembly of these events in categories, except for the treatment of solar eclipses.
Ban Gu's sources for the information given in the treatise may well have included the Chun qiu and the Shi ji, and there are citations from the three zhuan attached to that record. We may ponder whether he had at his disposal any lists of strange incidents that had been kept in the office of the Taishi gong, or whether he may have seen documents such as those that concerned the movements of the planets, the record of comets or the operation of Yun qi雲氣, as found at Mawangdui 馬王堆. We cannot know on what sources he drew for the interpretations of the incidents that he quotes, whether on the part of Dong Zhongshu 董仲舒, Sui Hong (Meng) 眭弘(孟) (f l. 78 bce), Xiahou Sheng 夏侯勝 (fl. 70 bce), Liu Xiang 劉向 (79–8 bce), Liu Xin 劉歆 46 bce–23 ce, Jing Fang 京房, Li Xun 李尋 (fl. 5 bce) or Gu Yong 谷永 (died 9 bce), some cited frequently, some only rarely. It might well have been dangerous for a man such as Dong Zhongshu to have made or retained a written record of his views, as they might well have been cited as evidence if he was charged with disloyalty or conspiracy against the throne. Particular difficulties are seen in the citations from Jing Fang, as it is necessary to determine whether they were the utterances of the elder (c. 140–c. 80 bce) or the younger Jing Fang (executed 37 bce).Footnote 80
At the outset of the chapter Ban Gu writes about the scholarly background of those whose interpretations he quotes and the differences between them.Footnote 81 Inspection of the text of the chapter shows that by no means all the writers who are cited gave their explanations for all the events that are recorded. In general, the interpreters gave one of three reasons for the occurrence of a strange event of nature or of unknown origin; it may have resulted from earlier actions of human beings; an abnormality may have been due to human behaviour that was awry; or it may have been a mark of Heaven's beneficence in warning human beings of their errors. Varying characteristics pertain to the interpretations of these incidents that Han shu 27 includes.
a) There are seventy-four explanations attributed to Dong Zhongshu of which no more than one concerned an event of Han times. Thirty-five of these are coupled with those of Liu Xiang. Fourteen explain an event as a warning sent by Heaven (Tian jie ruo yue 天 戒若曰). These explanations cite events that had preceded the strange incident under discussion; there is no conspicuous trust in Yin Yang.
b) Liu Xiang frequently calls on the warnings of Heaven.
c) All except one of the events which Liu Xin seeks to interpret were of Chun qiu times.
d) The seventy citations from Jing Fang are short and precise, such as “First joy then grief; the prodigy for this is a rain of feathers from Heaven”.Footnote 82
e) Of the ten events on which Gu Yong comments, only one took place earlier than his own time.
Scholars such as W. Eberhard, H.H. Dubs and H. Bielenstein have discussed the arbitrary but indeed purposeful way whereby certain events were chosen for treatment or omitted from mention.Footnote 83 Ban Gu deserves credit for the care with which he chose both the incidents to be recorded in this chapter and the different interpretations that had been expressed. It is possible that personal considerations affected his choice, so as to avoid putting down statements or opinions that might have endangered him; he may also have omitted some incidents that might have embarrassed or angered the rulers of his own time.
Ban Gu the critic
There are indications that, despite these inhibitions, Ban Gu was able to express his criticism of the characters or activities of some of those whose history he recounts, either subtly or openly. He records a series of occasions from 147 bce to 2 ce when locusts afflicted the land.Footnote 84 For these events he does not call on the interpretations of any of the masters whom he cites to explain other acts of nature, and indeed those masters may well have refrained from commenting on events of their own times. To each event he adds his own observations which take the form of noting the military adventures the government had initiated previously. To the last instance, of 2 ce, he simply remarks that it was at this time that Wang Mang took hold of the government. Ban Gu's indication of cause and effect, while not stated, is implicit; it may be contrasted with his treatment of similar incidents that took place in pre-imperial times and drew comments from Liu Xin, Dong Zhongshu and Liu Xiang.Footnote 85
By including the texts of three memorials of Gu Yong 谷永 and a letter written by Liu Xin, Ban Gu also contrived to include some sharp criticism of the last decades of Western Han. Gu Yong never rose to hold a senior post in the central government. In three memorials, dated in 29, 15 and 12 bce,Footnote 86 he delivered what were probably the most courageous and forthright protests ever addressed to an emperor of the early dynasties. He enjoined Chengdi to play a responsible part as emperor rather than neglect his duties as such; he should appoint suitable persons to hold office and eliminate oppression. Time and again Gu Yong criticized his emperor for his personal conduct, such as his indulgence in entertainments and his extravagance. But it was in connection with his relations with women that Gu Yong wrote most sharply. For some years the emperor had failed to sire an heir to the throne; but he was in the habit of associating with women of the lowest class and joining them and their friends in drinking bouts and other frolics that were marked by degrading and disorderly behaviour. Gu Yong reminded Chengdi of the harm and dangers that certain women had brought to bear upon the throne and the dynasty; and he described in some detail the abominable cruelties practised by some of the women of his own court.
Gu Yong's strictures against the women of Chengdi’s palaces were probably addressed to the sisters Zhao Feiyan 趙飛燕 and Zhao Zhaoyi 趙昭儀, who were of low class origins.Footnote 87 Ban Gu would hardly have included these memorials had he thought that they had been addressed to Ban Jieyu 班婕妤, his own great aunt and also a secondary consort of Chengdi. By Ban Gu's time, Zhao Feiyan and Zhao Zhaoyi had been discredited; Ban Jieyu, a quick-witted woman of some literary ability, had indeed been accused of imprecation but it was probably due to Chengdi's personal intervention that she was saved from dire punishment. Surely neither Ban Biao, nor Ban Gu, nor Ban Zhao would have included material that could be interpreted as discrediting one of their own relatives.
Ban Gu may well have harboured his own views regarding the shortcomings of some of the scholars of his day. It was the refusal of the academicians to accept the Zuo zhuan as a serious subject of study that led Liu Xin to write his indignant complaint against their point of view. In including it in the Han shu,Footnote 88 Ban Gu may perhaps have had in mind that the scholars of his own day deserved precisely the same type of rebuke. Liu Xin recounted the whole tale of the growth and losses of Chinese literature from earliest times through the restrictions imposed under Qin and the revival of interest during Wudi's reign. But despite the efforts made then, scholars and teachers were men of small calibre, capable of doing no more than read one of the ancient texts aloud. Few of them could handle the whole of a text in a scholarly way, and they found it necessary to work in groups in order to do so. Liu Xin referred to the discoveries of texts and the way in which, to his distress, these had been stored away and concealed from open view, as had happened with the works that he and his father Liu Xiang had found during Chengdi's reign. Liu Xin continued:Footnote 89
This state of affairs was highly distressing for anyone capable of thought and it was a matter of deep concern to scholars and gentlemen shi junzi 士君子. Hitherto those who had been piling up scraps of learningFootnote 90 had never pondered on the losses that literature had sustained, whether these were due to deliberate suppression or to a break in transmission. They simply fastened on questions of the narrowest importance and concentrated on insignificant detail. They split up their texts and explained the meaning of the characters with interminable verbiage. They were worn out and old, those men of learning, quite incapable of studying even one of the texts right through to its end. They believed what they were told by word of mouth but turned a blind eye to written records; while accepting what the latter-day teachers said, they rejected anything that was really old. So, when a major event was to take place, such as the construction of the Biyong 辟雍,Footnote 91 or performance of the feng 封 or shan 禪 rites, or an imperial tour of inspection, they just looked blank – none of them understood how these affairs had been conducted originally.
Liu Xin complained of the scholars’ obdurate refusal to accept new ideas and their rejection of newly found texts such as the Zuo zhuan. He supported his case by referring to the ways in which Xuandi had attempted to expand the scope of scholarly activities.
Ban Gu did not stop at including criticism of what had been the practice in Western Han. At the close of his account of li 禮,Footnote 92 Ban Gu is directly critical of some aspects of the reign of Mingdi (57–75). He recounted some of that emperor's praiseworthy actions such as his own performance of the li, the services he paid to Guangwudi in the Ming tang 明堂, and his attention to the San lao 三老 and Wu geng 五更.Footnote 93 Nevertheless, for a number of reasons there had been no spread of culture. There had been no practice of li and music, no recitation of texts at the lower reaches of society, and no establishment of schools. The writings of Shusun Tong 叔孫通 (died c. 188 bce) on li and correct form (li yi 禮儀), and the Statutes and Ordinances lay concealed in store in the government's offices; there was no transmission of the work of the specialists in the correct models for activities. The Institutions of Han (Han dian 漢典) lay dormant, never mentioned by officials or others. After Shusun Tong's death, Liu De 劉德, king of Hejian (155–130/129), had collected some of the old items of li and music, increasing the volume of these writings to over 500 pian Footnote 94 but the scholars of the present day (jin xuezhe 今學者) as the compiler of the Han shu wrote, “are incapable of taking an intelligent view of them; their talk of what is right and just is muddled and different from what is acceptable”.
It is not certain whether Ban Gu was directing his criticism solely at the time that ended at Mingdi's death in 75 or whether he was also alluding to conditions in his own time. As Balazs has shown, he was by no means the only prominent person of Eastern Han to hold and express critical views, as is seen a little later in the cases of Wang Fu 王符 (c. 90–c. 165), Cui Shi 崔寔 (died c. 170) and Zhongchang Tong 仲長統 (180–220).Footnote 95
Only a few parts of the Han shu may be identified as stemming from the hands of Ban Biao and Ban Zhao and thus independent of Ban Gu's own authorship. Some chapters are known to have been composed by others such as Liu Xin and, at a subsequent date, Ma Xu. For the greater part of the Han shu it is to Ban Gu that we ascribe the work of collecting sources of information and determining those parts of this that were worthy of inclusion in his historical record. As compiler and editor of the Han shu we find Ban Gu ready to copy and incorporate writings such as the poems of Sima Xiangru or the hymns of state; to produce or create sections on certain topics that had not been treated previously, such as that of punishments, with his own pertinent introductions and comments; and to include an historian's criticism of the decisions and actions of the past, both in the form of views expressed at the time, e.g. by Gu Yong, and as his own conclusions, whether enunciated openly or implied tacitly. Such notes of criticism may have concerned conditions in his own time, such as the competence of some scholars; or they may have been directed, somewhat boldly, at the reputation of an earlier emperor of the Eastern Han dynasty. In all these respects the function of the copyist, creator and critic overlaps. For our part, we may credit Ban Gu with courage and initiative and be grateful for his introduction of new elements into the composition of China's historical writings.
