It was with genuine excitement that I opened Brian Bruya and Wenwen Li’s new translation of the Kongzi jiayu 孔子家語, the largest early anthology of sayings, stories, and dialogues attributed to Kongzi 孔子 (Confucius; trad. 551–479 BCE) and his students, making The Dialogues of Confucius: The Complete Text the largest English translation of Kongzi material ever published. As important as Kongzi obviously is, we still do not have good resources in English for studying Kongzi literature beyond the most canonical sources. A complete translation of the Kongzi jiayu goes a long way toward rectifying this problem for students and scholars alike.
The good news is that Bruya and Li have absolutely succeeded in creating a readable, user-friendly, richly annotated translation of this important work. The best compliment I can pay is that the experience of reading The Dialogues of Confucius reminded me why I was so drawn to Kongzi literature in the first place. What’s more, the authors have gone out of their way to make The Dialogues of Confucius a useful reference work and study guide for Kongzi and classical Chinese philosophy more broadly. Between their 84-page “philosophical lexicon” and 64-page glossary, Bruya and Li provide more than 3,000 references to specific Kongzi jiayu passages—a veritable treasure trove of information. As someone who has spent years mapping and reading early Kongzi literature, I wish I had had access to this resource a long time ago.
Unfortunately, for all its strengths as a translation and reference work, The Dialogues of Confucius is significantly undermined by the authors’ overall theory of the text. Minimizing the almost total lack of evidence for the anthology’s existence prior to its introduction by Wang Su 王肅 (195–256 CE), not to mention the nearly 1,000-year history of skeptical scholarship, Bruya and Li invite us to read the Kongzi jiayu as an authentic record of Kongzi’s life and teachings compiled no later than 350 BCE. In their eagerness to establish the text as a Kongzi canon on par with the Lunyu 論語 (Analects), whose authenticity they also take for granted, the authors somehow ignore the best reason to care about the Kongzi jiayu: as a superb introduction to the dynamic, diverse, and highly intertextual world of Kongzi literature from the Warring States through the Han, a literature whose interest and appeal did not necessarily depend on its connection to the historical Kongzi. In this critical respect, Bruya and Li do their translation a major disservice.
The argument
Bruya and Li summarize their approach early in their 141-page introduction:
It is impossible to know for certain the textual history of the Dialogues, but in what follows we lay out what we think is the most plausible scenario, a scenario that is supported by a growing body of evidence. We believe the Dialogues can provisionally be accepted as largely genuine and accurately portraying the activities and thoughts of Confucius. While some of it remains suspect, it can nevertheless be used as a resource for understanding Confucius, his interactions with his students, and his philosophy. (p. 2)
On the one hand, they acknowledge the impossibility of confirming the textual history of the Kongzi jiayu with any degree of certainty. On the other, they confidently affirm its value as a “largely genuine” and “accurate” record of the historical Kongzi.Footnote 1 (Unfortunately, the authors never specify which parts of the collection, if any, they find “suspect.”Footnote 2)
How do Bruya and Li square this circle? The keyword is “provisional”: they essentially adopt an innocent-until-proven-guilty approach to the dating of ancient texts. Setting up a straw-man appeal to “some scholars,” never cited, who “say that, where we lack absolute confidence, we should refrain from attributing authorship” (p. 18), Bruya and Li counter that we can never be 100% sure that any ancient source truly reflects the words and ideas of the person it is credited to. Were we to demand “100-percent confidence…then a vast range of texts…would be off-limits” (p. 17), from Plato’s dialogues to the Gospels, the Aristotelian corpus to Shakespeare’s plays. Moreover, we need named authors if we want “to position a thinker’s ideas not only in the history of the tradition but also in the current conversation” (p. 18). To that end, they argue, it is acceptable to proceed on the presumption of genuine authorship as a “hermeneutic shorthand” (p. 18), pending evidence to the contrary.
The problem with applying a provisional standard to any source of Kongzi is that it applies to lots of other sources, too. We could make the exact same argument about Kongzi material in the Da Dai Liji 大戴禮記 (Elder Dai’s Ritual Records), Han Feizi 韓非子, Hanshi waizhuan 韓詩外傳 (Outer Traditions of the Han Odes), Liji 禮記 (Records of Ritual), Lüshi chunqiu 呂氏春秋 (Annals of Lü Buwei), Mengzi 孟子, Shiji 史記 (Grand Scribe’s Records), Shuiyuan 說苑 (Garden of Persuasions), Xiaojing 孝經 (Classic of Filial Piety), Xunzi 荀子, Yanzi chunqiu 晏子春秋, Zhuangzi 莊子, and Zuozhuan 左傳 (Zuo Tradition), not to mention the Fuyang 阜陽, Dingxian 定縣, Shanghai Museum, Anhui University, and Wangjiazui 王家嘴 manuscript corpora, among other texts.Footnote 3 We have no definitive proof against these other versions of Kongzi, either, nor can we disprove the traditional attribution of the Six Classics to Kongzi. In fact, most of these other texts are far more securely attested than the Kongzi jiayu. That title appears just once in the early textual record: within the “Yiwen zhi” 藝文志 (Record of Arts and Letters) bibliography of the Hanshu 漢書 (History of the Han; ca. 100 CE), attached to a text in 27 chapters instead of the 44 chapters of Wang Su’s version.Footnote 4
Canonization is about exclusion as much as inclusion, so how do Bruya and Li justify elevating the Kongzi jiayu over these other sources? Their solution is to accept without reservation the testimony of a postface to the Wang Su edition ostensibly written by Kong Anguo 孔安國, a descendant of Kongzi active during the reign of Han Emperor Wu 漢武帝 (141–87 BCE). According to that postface, helpfully translated by Bruya and Li in an appendix, Kongzi’s students took the best records of the master’s teachings to create the Lunyu and then used the leftovers to compile the Kongzi jiayu. From there, the Kongzi jiayu was transmitted by Kongzi’s students through Mengzi and Xunzi down to the Qin 秦 dynasty (221–206 BCE), when it “was grouped among the books of the Masters and so escaped destruction” (p. 496)—a bizarre claim given the Shiji account of the Qin bibliocaust.Footnote 5 The Han imperium eventually received a copy of the Kongzi jiayu; however, it “became mixed up with events of the various states and the words of the seventy-two students, and it became impossible to tell [the texts] apart” (p. 496). This is where Kong Anguo comes to the rescue, claiming to have “made private solicitations to noblemen and to officials high and low for their assistance,” “acquired the entire collection,” and “set about putting it into order by topic” (pp. 496–7). A second postface to the Wang Su edition (pp. 499–501) apparently seeks to explain why Kong Anguo’s text was so neglected in the Han. The witchcraft scandal of 91 BCE prevented Kong Anguo from submitting the Kongzi jiayu to the Han court, and his grandson Kong Yan 孔衍 successfully promoted the text to Emperor Cheng 成帝 (33–7 BCE) only for the emperor to die before the recommendation could be implemented.
A careful reading of the relevant Han sources reveals a few reasons not to take this story at face value. Here as elsewhere, Bruya and Li’s insistent focus on the early Warring States leads them to neglect the intricacies of the Han context. First, Bruya and Li lean heavily on the first half of the postface but ignore the obvious tension with the second half, which presents the Kongzi jiayu as a collection that was (1) compiled from disparate sources at some point in the Han and (2) far too easily confused with other Kongzi texts. This half of the story much more aptly describes the Kongzi jiayu as we have it today, more than 90% of which has verbatim parallels in other early sources.Footnote 6 Is the text an authentic Warring States record or a product of Han-era compilation processes? The postface’s ambivalence on this point is telling.
Second, the postface’s account of the Kongzi jiayu’s origins smacks of a Han imperial context, thus undermining its value as a source of pre-Han textual realities. Whatever one’s view of the Lunyu’s origins, no extant source mentions a “Lunyu” title prior to ca. 100 BCE.Footnote 7 Likewise, the postface’s grouping of the Xiaojing alongside the Lunyu (p. 495) is a rhetorical habit first attested in the latter part of Emperor Wu’s reign, when the Lunyu and Xiaojing were adopted as a joint curriculum for the training of Han imperial princes.Footnote 8 Once emperors began broadcasting their mastery of the Lunyu and Xiaojing as a qualification for the throne, Kongzi bibliography became a tool for validating these symbols of Han dynastic legitimacy. The author of the Kong Anguo preface clearly drew from this tradition of imperial bibliography to promote the Kongzi jiayu.
A third reason not to trust the postface has to do with its relationship with the “Yiwen zhi” bibliography, the only extant early source to mention a “Kongzi jiayu” title. Bruya and Li invoke the “Yiwen zhi” as evidence that the Kongzi jiayu was known to Liu Xiang 劉向 (77–6 BCE) and thus could not have been forged by Wang Su (p. 6). However, they fail to notice that the postface appears to quote the opening lines of the “Yiwen zhi.” The passage in question—“Confucius passed away, and his subtle utterances came to an end. After his 72 students themselves passed away, his central ideas were distorted” 孔子旣沒而微言絶, 七十二弟子終而大義乖 (p. 495)—also appears in a memorial submitted to Emperor Ai 哀帝 (r. 7–1 BCE) by Liu Xiang’s son Liu Xin 劉歆 (50 BCE–23 CE), whose Qilüe 七略 (Seven Surveys) served as the basis for the “Yiwen zhi.”Footnote 9 The same lines get picked up and repeated in various contexts from the second century CE onward without any mention of Kong Anguo.Footnote 10 This single detail strongly suggests a post-Western Han date of composition.
Readers unconvinced by the testimony of the (pseudo-?)Kong Anguo postface are unlikely to be swayed by the “growing body of evidence” (2) cited by Bruya and Li in support of the Kongzi jiayu’s fifth-century origins. Here I will mention just three points. First, the authors argue that the Kongzi jiayu’s lexical profile—for example, the absence of words like “crossbow” (nu 弩) and “armored” (daijia 帶甲)—clearly reflects an early Warring States context, based on the earlier work of scholars (especially Yuri Pines) who have proposed various lexical criteria for dating early Chinese texts. In my view, the explicit circularity of this method renders any such results invalid.Footnote 11 Second, they argue that parallels between the Kongzi jiayu and early manuscript corpora “suggest the authenticity of the Dialogues” (p. 27), as seen in the substantial overlap between the Shanghai Museum *Min zhi fumu 民之父母 (Father and Mother to the People) manuscript and Kongzi jiayu 27.2. However, such parallels verify the pre-Han origins only of specific passages, not the anthology as a whole.
Third, Bruya and Li compare the Kongzi jiayu and Lunyu in an effort to show that “the ideas presented in the Dialogues are a reliable representation of the ideas of Confucius” (p. 39). In the process, they largely assume the Lunyu’s traditional chronology without giving readers a fair rendition of dissenting views, including my own.Footnote 12 They also tend to ignore connections with other sources of Kongzi, resulting in a greatly impoverished view of the Kongzi jiayu’s vast and complex intertextual profile. For instance, they observe that the Kongzi of Kongzi jiayu 41.7 attributes a saying to an “ancient record” 古也有志, whereas the Kongzi of Lunyu 12.1 claims the same saying for himself (“Ren is overcoming oneself and returning to li” 克己復禮為仁; p. 32). Observing that proverbs were increasingly attributed to Kongzi over time, they hypothesize that the Kongzi jiayu version “may even predate the similar passage in the Analects” (p. 33). However, they fail to note that the Kongzi jiayu version has a word-for-word parallel in the Zuozhuan, as do 18 of the 26 other sections of Kongzi jiayu chapter 41.Footnote 13 When Bruya and Li do compare the Kongzi jiayu with texts other than the Lunyu, their primary concern is to establish the Kongzi jiayu’s precedence. Thus, they say that the “Li yun” 禮運 (Li in Motion) chapter of the Kongzi jiayu “is the more authentic version of the essay” than its Liji counterpart due to its “internal consistency” (p. 37), and the Kongzi jiayu 11.1 version of a Kongzi-Laozi dialogue “makes much more sense” than the Shiji version and so “gives us some confidence that the rest of the story in chapter 11 may…track actual events” (p. 39). Nowhere do they justify their use of consistency and legibility as criteria of authenticity, nor do they acknowledge that later editors and compilers could also intervene to correct textual infelicities.
One of the most exciting aspects of studying the Kongzi jiayu is tracing its connections with other sources of Kongzi literature. Instead of giving these connections the attention they deserve and following the evidence of Kongzi jiayu intertextuality wherever it leads, Bruya and Li paint themselves into a corner with a far too rigid theory of the text.
The translation
Bruya and Li’s decision to treat the Kongzi jiayu as a Kongzi canon does have one major advantage, which is their decision to package the text as a primer to classical Chinese philosophy. They devote more than 80 pages of the introduction to an extensive “philosophical lexicon” (pp. 54–138) consisting of 28 “umbrella concepts” like “affection,” dao 道, junzi 君子, ren 仁, and “sage,” for which they provide copious page references sorted by use and meaning. Their entry for de 德, for example, splits references into four groups: (1) general statements on de, (2) de as “moral, virtue,” (3) as “merit, favor,” and (4) as “power, contribution, virtue.” I greatly appreciate their careful attention to the multivalence of these concepts, as well as their decision to retain the original Chinese terms throughout the translation. This is what makes The Dialogues a “philosophical translation”: “When we have to make the choice between technical accuracy and readability, we put our focus on the philosophical terminology… [W]e often simply transliterate the term, forcing the reader to slow down and think about what the term on the page means and how it relates to other terms in the same semantic web” (p. 47).
The greatest strength of Bruya and Li’s translation is its readability, which they manage to achieve without sacrificing too much in the way of accuracy. There are occasional errors, however—for example, when they nonsensically render the phrase cheng qi qin 成其親 as “fulfill your parents” (p. 161) instead of “bring about [a feeling of] kinship [among the populace]” (4.2/p. 161). Another example is when they introduce an anachronistic distinction between “legend” and “history” with the line “All we know about the Five Chiefs are legends, unlike the Three Kings, for which we have history” 五帝用説, 三王有度 (23.3/p. 321). The actual distinction, I think, is between earlier sage-kings known via “stories” or “explanations” (shuo 說) and later sage-kings whose “standards [of conduct]” (du 度) have survived in the form of rituals and institutional structures. Although these particular mistranslations are relatively inconsequential, a few others create unfortunate false impressions. At one point, the authors apparently confused xiaozi 小子 (little one, follower) with xiaoren 小人 (petty man) to have Kongzi call Zilu 子路 a “scoundrel” (8.13/p. 193). Kongzi is sometimes portrayed as a stern teacher, but “scoundrel” is a step too far, even for Zilu. At 12.22/p. 237 they also overtranslate the concluding line of a dialogue between Kongzi and Zigong as “Zigong bowed and said, ‘I’m going to go now and preserve [this teaching] for posterity’” 子貢跪曰: 請退而記之. The key word here is ji 記, which Bruya and Li accurately explain in a footnote: “This term is usually interpreted as to commit to writing, but it can also mean to commit to memory. The translation here preserves the ambiguity of the original” (p. 237, note 24). However, “preserve for posterity” not only leans too far in the direction of a written record, it also introduces an emphasis totally absent from the original. In fact, Kongzi frequently tells his students to “take note of” (ji 記, zhi 志) his teachings, often with a pithy saying or “motto” (also zhi 志) at the instruction’s conclusion. The point is to facilitate his students’ moral cultivation, not future readers.Footnote 14
Bruya and Li’s footnotes provide yet another rich layer of useful information and analysis. In general, they do an excellent job of explaining and contextualizing specific Kongzi jiayu passages, often noting relevant intertexts and variants. But here, too, their approach to the Kongzi jiayu’s intertextuality feels ad hoc. For many entries, readers would never know that the content is largely duplicated in other early sources. An appendix like the detailed “Table of Parallels” included by Robert Kramers in his 1950 study and partial translation would have gone a long way towards rectifying this problem.Footnote 15
Do I recommend The Dialogues of Confucius: The Complete Text? Yes, but with major reservations. The translation and accompanying resources definitely succeed in opening up the world of early Kongzi literature to an English readership. For that reason alone, I look forward to using The Dialogues of Confucius in my undergraduate teaching. But when I do assign it, I will caution my students to take the authors’ presentation of the text with a hefty grain of salt. As much as we would like to recover the life and teachings of ancient sages like Kongzi, overcoming the limitations of our sources is not nearly as simple as Bruya and Li make it out to be. The good news is that we do not need the historical Kongzi to appreciate a text like the Kongzi jiayu. As Robert Kramers recognized more than 75 years ago, it remains a fabulously rich source of early Kongzi material regardless of the date or circumstances of its compilation:
Judged by its general contents, the fact should again be stressed that it does not make a great deal of difference whether the Jiayu was entirely compiled in the third century A.D. or partly also in the second or first century B.C.: to us it mainly represents the Confucian lore existing in the third century B.C.Footnote 16
Personally, I would emphasize the Western Han context more than the late Warring States period. But Kramers’s larger point stands: its extensive intertextual profile proves beyond the shadow of a doubt that it represents a broad cross-section of Kongzi sayings, stories, and dialogues from the Warring States through the Han. Surely that is good enough!
