Whispering in Early China

Abstract

The word ouyu occurs with some frequency in early Chinese literature, signifying “to whisper” or “to murmur.” It is frequently associated with expressing dissatisfaction in a furtive manner and with behaving in an undisciplined or indecorous way. In this article, I cover all significant instances of this word that occur in Song and pre-Song sources and find them to be about whispering for different reasons, in different contexts and with different consequences. According to the Shiji, the First Emperor of the Qin banned ouyu about the Documents and the Odes (ouyu Shi Shu). This is the earliest attested use of the term and the passage is usually interpreted to mean that the First Emperor prohibited people from congregating to criticise his government by reference to past lore as narrated in the Documents and the Odes. This would make it the only instance in the available literature where ouyu is about anything. I argue that the commonly accepted interpretation rests on a misinterpretation of the early commentaries and suggest that the characters Shi Shu may be the result of a corruption in Shiji. What the Qin emperor outlawed was the activity of whispering as such – if people talked conspiratorially in a subdued voice, the regime saw this as proof that they harboured rebellious intentions and therefore deserved public execution, with no further evidence needed. In the last section, I discuss some literary senses of ou. I have previously argued that some of these are helpful to understanding the Qin Burning of the Books; I now dismiss them as irrelevant, since the widely attested use of ouyu as “to whisper” fully explains the Qin regime’s efforts to quell criticism of its rule, as well as all later uses of the word in similar contexts.

「偶語」這個詞經常出現在漢代至宋代的歷史記載或詩文當中。它的語義為「竊竊私語」，往往關係到謀反或擾亂隊形、朝班等事情。《史記》記載秦代「有敢偶語《詩》、《書》者弃市」。宫崎市定提出過要刪除「詩書」這兩個字；我贊成他的意見，并提供新的論證。秦始皇想取締的不是一般老百姓聚會討論經書；他反對的是有人暗地裡誹謗他。「偶語」有幾個同義辭：「聚語」、「族談」和「僔語」。我還探求「偶語」在文學領域裡面的幾個用法。

Keywords:

• whispering
• early Chinese history
• the Qin Burning of the Books

關鍵詞:

• 偶語
• 中國古代史
• 焚書坑儒

Notes

1 Le Grand Ricci, vol. IV, #8347 (偶語), #8361 (耦語).

2 Hanyu da cidian, vol. 1, p. 1548. Dai Kan-Wa jiten #899.15 defines the word with variations on “to talk tête-à-tête,” adducing the three Shiji 史記 occurrences discussed below.

3 There are many pre-Tang words meaning “to whisper.” They include eryu 耳語, siyu 私語, fu’er 附耳, nie 聶/囁, nieche 囁呫, nieru 囁嚅, ni’nan 呢喃, qi 咠, and miyu 密語. Here I will only discuss ouyu and a small number of synonyms occurring in the same context as ouyu.

4 Guxun huizuan, p. 142.

5 Wang Li 1982, pp. 185–186; Schuessler 2007, p. 407.

6 TLS, concept “speak,” https://hxwd.org/concept.html?uuid=uuid-07855030-3463-472d-bd01-1c7ceebbfb3d, (accessed 23 May 2022).

7 Hanshu 40.2031. Wangwang 往往 regularly collocates with ouyu. See examples in the following and Suishu 38.1141, Da Tang chuangye qiju zhu 1.10. In fact, the association is so strong that one can reliably posit that the occurrence of wanglai ouyu 往來偶語, Cefu yuangui 401.4555, should read wangwang ouyu 往往偶語.

8 Lu Zhi 陸贄 (754–805), when recounting this event, directly states that Liu Bang’s generals engaged in ouyu in order to raise a rebellion (ouyu moufan 偶語謀反); Quan Tangwen 471, 12a, p. 4812; see also Quan Tangwen 471, 9a, p. 4811.

9 Sima Guang 司馬光 (1019–1086) held that Zhang Liang must have known about the generals’ rebellious intentions and that he ought to have informed Liu Bang about them beforehand, not wait until Liu Bang saw the generals talking together; Zizhi tongjian 11.370. However, Zhang Liang could also have reasoned that they were plotting together.

10 Shiji 55.2042–3 has 往往相與坐沙中語. The sands are presumably by the Luo River, used as exercise grounds, but here the generals’ manner of speaking is not specified. See also Xinxu jiaoshi 10.8.1326, where the persons talking together are qunchen 羣臣. Chuxue ji 9.210 quotes the passage from Hanshu, but with the Shiji wording.

11 Liang Hanji 1.3.41.

12 尚書朱軌與中黃門嚴生不協，會大雨霖，道路陷滯不通，生因而譖軌不修道，又訕謗朝政，季龍遂殺之。於是立私論之條，偶語之律，聽吏告其君，奴告其主，威刑日濫，公卿已下，朝會以目，吉凶之問，自此而絕. Jinshu 106.2778.

13 石虎之法，雖周厲王之監謗，秦始皇之禁耦語，不如是之甚也. Zizhi tongjian 97.3070–1.

14 Li Zhan’s name (瞻) ideographically inspired Hou Jing to expose his kidney (dan 膽).

15 又禁人偶語，不許大酺，有犯則刑及外族. Nanshi 70.2009.

16 Quan Tangwen 171, 7a, p. 1745.

17 Suishu 36.1113, Beishi 13.536.

18 Xiong 2006, pp. 70–71. That Emperor Yang’s disaffected troops engaged in ouyu is also told in the Da Tang chuangye qiju zhu 1.10.

19 Guisso 1979, pp. 322, 326–327.

20 時京城恐懼，相傳將有革命之事，往往偶語，人情不安. Jiu Tangshu 51.2174.

21 自元忠下獄，臣見長安城內，街談巷議，皆以陛下委任奸宄，斥逐賢良，以元忠必無不順之 言，以易之必有交亂之意，相逢偶語，人心不安. Jiu Tangshu 187A.4882; cf. Xin Tangshu 112.4168.

22 親識宴聚及道路偶語，犯者有刑. Zizhi tongjian 241.7767.

23 民間有偶語者，官中罔不知. Wangshi jianwen lu 王氏見聞錄 quoted Taiping guangji 126.895.

24 兵士素憚東征，又怨刻薄，遂偶語竊忿，漸至喧呼. Cefu yuangui 104.1137.

25 The Zhengyi 正義 commentary says the word yue 約 here means “to reduce”; thus the line could also be parsed without the break to read: “I shall reduce the laws to three articles for you elders.” Hulsewé favours this reading (Hulsewé 1955, pp. 368–372). For an extended discussion, see Hori Toshikazu 1991.

26 召諸縣父老豪桀曰：「父老苦秦苛法久矣，誹謗者族，偶語者弃巿。吾與諸侯約，先入關者王之，吾當王關中。與父老約，法三章耳：殺人者死，傷人及盜抵罪。餘悉除去秦法。⋯」; Shiji 8.362; Hanshu 1.23 has 耦語, but this is probably an instance of archaisation. See also the Shiji quotes in Taiping yulan 87, 2b, p. 412b; 390, 4a, p. 1803b, and the Hanshu quotation in Beitang shuchao 45, 5b. Cf. translation in Nienhauser et al. 1994–, vol. 2, p. 38.

27 河朔故法，偶語軍中則死，弘宣使除之. Xin Tangshu 122.5632.

28 河北諸帥防其下相與聚謀以圖己，故嚴軍中偶語之法，以剛制之。盧弘宣至中山，乃除其法. Zizhi tongjian 248.8013. Hu Sanxing comments on 河北之法，軍中偶語者斬; cf. his commentary to 禁都下人毋得偶語 (Zizhi tongjian 283.9237): 以防姦人謀為變.

29 舊令：途無偶語，夜不燃燭，人或以酒食相過從者，以軍法論。度乃約法，唯盜賊、鬥殺外，餘盡除之，其往來者，不復以晝夜爲限。於是蔡之遺黎始知有生人之樂. Jiu Tangshu 120/4419; cf. Xin Tangshu 173/5212; Zizhi tongjian 240/7744.

30 爲政嚴酷，所在不敢偶語. Jiu Tangshu 124/3535; Xin Tangshu 213/5990. Li was inducted into the imperial clan in the following year, 778. Zizhi tongjian (225/7250) notes the beneficial effects of his administration: laws were uniform and taxes equitable and light.

31 Tonami Mamoru 1988, p. 123, translates ouyu as 肩ならべて話す (to talk shoulder to shoulder).

32 Cefu yuangui 944.10945.

33 Wenguo Wenzheng Sima gong wenji 32, 10b. Zhu Xi also mentions ouyu and zutan together, along with 街談巷議, Zhuzi quanshu, vol. 20, p. 629.

34 階蟻相逢如偶語、園蜂速去恐違程. Liu Yuxi ji jianzheng 1269.71; see also Liu Yuxi quanji biannian jiaozhu, pp. 721–722.

35 間數日，率一謁晉文公祠。既至祠下，必與神偶語，久之乃出，文公亦時時來謁景文。景文閉閭若與客語者，則神之至也. Zhou Zizhi 周紫芝 (1082–1155), Zhupo shihua 20; see also Yijian zhi 丙志 10.453.

36 先生坐睡甚酣，因諸生偶語而覺，即揖諸生. Zhuzi yulei 朱子語類 107 (丙辰後), Zhuzi quanshu 17.3500.

37 Bai Juyi shiji jiaozhu, p. 909, Han Wo shiji jianzhu, p. 152, Quan Tangshi 510.5795.

38 Luoyang jinshen jiuwen ji 4, 2a, p. 1036–159b.

39 Hong Kuo 洪适 (1117–1184), Panzhou wenji 74, 1b: 有小吏偶語喜甚。

40 Taiping guangji 205.1561: 有相偶語笑者。

41 Jiu Wudai shi 62.830: 部下五百人懼罪，將欲奔竄，聚泣偶語。

42 Fan Wenzheng gong ji 2, 12b: 范仲淹, 送河東提刑張太博。

43 For more instances where ouyu clearly has the meaning “to whisper,” see Suishu 43.1217 (避人偶語, dodging others to engage in ouyu); Quan Tangwen 795, 13a, p. 8337a (whispering while someone is asleep), 629, 12b, p. 6349a; Taiping guangji 484.3987.

44 Tangdai muzhi huibian xuji, Qianfu 乾符 (874–879), 017.1130. “Mumble” might be more apposite here.

45 Mozi chengshou ge bian jianzhu 15.121; Mozi jiangu 15.605. See Ian Johnston’s translation, Johnston 2010, p. 881; Johnston has “executed,” but states that the interpretation of duan 斷 is uncertain.

46 In the Sunzi bingfa 孫子兵法, we read that 諄諄翕翕，徐言入入者，失眾也; Sunzi jiaoshi 9.161. This is usually taken to mean that soldiers talking in subdued voices is a sign that their commander has lost their confidence: this would then be an additional case in which whispering signifies disaffection. However, I believe the general is the subject throughout this sentence; his meek manner of speaking is held to show that he seeks to curry his men’s favour, thus: “if the general is earnest and amiable and speaks slowly and gently, he has lost the support of his men” (ruru 入入 is here to be read as nana 訥訥).

47 Yan Buke 2000.

48 正班時，宰相宗楚客、楊再思常離班偶語，無競前曰：朝禮至敬，公等大臣，不宜輕易以慢恆典. Jiu Tangshu 140.5027. Cf. Xin Tangshu 48.1239 where the duties of the 殿中侍禦史 are stated to consist in “correcting those who leave their court order and speak frivolously” (糾離班、語不肅者). Cf. Tongdian 24.674.

49 Xin Tangshu 32.4078.

50 離坐離立，毋往參焉。離立者，不出中間. Liji zhengyi 2, 58b-59a; Couvreur 1913, vol 1, p. 28.

51 Cefu yuangui 515.5847; Xin Tangshu 122.4354, Taiping guangji 205.1561.

52 Zizhi tongjian 225.7240. Compare Jiu Wudai shi 88.1144: 既發喪，都人不得偶語.

53 Jiu Tangshu 128.3586; Xin Tangshu 153.4851.

54 Guxun huizuan, p. 1000. Zu belongs to the same word family as ju; Zhang Xifeng 1999, pp. 1–9.

55 帥其屬而以鞭呼趨且辟。禁慢朝、錯立族談者. Zhouli zhushu 35.1100; cf. Biot 1851, vol. 2, p. 349; Zhouli yizhu, p. 530. Zutan is also associated with “street talk” (xiangyu 巷語); Jiu Wudai shi 34.469. In the Song dynasty text Qidong yeyu (1.1), zutan is associated with neglecting one’s occupation (不修職業).

56 Jiu Tangshu 183.4762.

57 族談者苦邊疆之戍役，偶語者傷戶口之凋殘. Quan Tangwen 29, 8a, p. 329a.

58 Shuowen jiezi gulin #5156.8201.

59 Shuowen jiezi gulin #856.2158. 譐 and 噂 are probably variants. See the discussion Yuan Mei 2013, pp. 446–447, Ye Meng 1993, p. 295.

60 Shuowen jiezi gulin #602.1837. Cuoli 錯立 and zutan are mentioned together in Jiu Tangshu 183.4762.

61 Jiaoshi Yilin zhuyi 1.10.

62 Mao Shi zhengyi 12.852a; Karlgren 1950, p. 139.

63 噂噂沓沓，相對談語，背則相憎逐. Mao Shi zhengyi 12.852a.

64 Normally this was a four-year sentence; Zhang Jiaying 1997, pp. 125–126.

65 See also Bodde 1938, p. 83. Ren Guangyin 1997 holds 樹 to refer to building and construction.

66 臣請史官非秦記皆燒之。非博士官所職，天下敢有藏詩書、百家語者，悉詣守、尉雜燒之。有敢偶語詩書者弃市。以古非今者族。吏見知不舉者與同罪。令下三十日不燒，黥為城旦。所不去者，醫藥卜筮種樹之書。若欲有學法令，以吏為師. Shiji 6.255. Li Si’s Shiji memoir (Shiji 87.2546) contains a version of the memorial which leaves out the part on ouyu. See also the translation in Nienhauser et al. 1994–, vol. 1, pp. 147–148.

67 應劭曰：「禁民聚語，畏其謗己。」《正義》偶，對也. Shiji 6.255; see also Shiki kaichū kōshō fu kōhō 6, 37, p. 515.

68 應劭曰：「秦禁民聚語。偶，對也。」; Shiji 8.362; see also Shiki kaichū kōshō fu kōhō 8, 38, p. 640.

69 Guxun huizuan, p. 142.

70 Wang Fengyang 1993, pp. 485–486.

71 Baihua ershisi shi, 史记, p. 49. Quoted in simplified characters according to the source text.

72 Shiji jinzhu, p. 223.

73 Nienhauser et al. 1994–, vol. 2, p. 38. The translators here insert into Liu Bang’s speech the characters 詩書 from the edict of the First Emperor; I argue below that these should be expunged.

74 Shiji quanyi, p. 277.

75 Shiji xinzhu, p. 124. Ershisi shi quanyi, 史記, vol. 1, p. 81.

76 Shiji quanben quanzhu quanyi, p. 558.

77 Nienhauser et al. 1994–, vol. 1, pp. 146–148.

78 Watson 1971, vol. 1, p. 55: “Anyone who ventures to discuss the Odes or Documents.”

79 Bodde 1986, p. 69; cf. Bodde 1938, pp. 83.

80 Chavannes 1895–1905, vol. 2, p. 173: “Ceux qui oseront discuter entre eux sur le Che (king) et le Chou (king).”

81 Dawson 2007, p. 74. See also Lunheng quanyi 25.484: 两人私下说及《诗经》和《尚书》的.

82 Hu Shi 胡適 (1891–1962) wrote in 1923 that “he had always been a little dubious” (總有點不相信) about the Qin prohibition against 偶語詩書, but that he was surprised to learn that French-occupied Vietnam (Annan 安南) had a law stipulating that “if more than two persons discussed the actions [of the French], this was to be regarded as a conspiracy” (二人以上商議其行為謂之陰謀), as an explanatory footnote to the Qin law; “Tian hu di hu xu”《天乎帝乎》序, Hu Shi 1998, vol. 3, p. 327. This law was imposed based on (what I believe is) a misreading of the Shiji commentary.

83 Beijing daxue cang Xi-Han zhushu, vol. 3, pp. 147–194.

84 Xin Deyong 2019, pp. 144–145. Xin here introduces the zhusheng because he holds as historical the Burying Alive of the Scholars and sees it as a continuation of the Burning of the Books. Whereas the latter is historical, the former is a myth; see i.a. Neininger 1983, Li Kaiyuan 2010, Kern 2010, pp. 111–112.

85 I have discussed these terms in my 1995 article. I hope to reformulate my arguments on this topic in a separate article. Here I open up for an overlap between the two terms.

86 Shiji jiaozheng, p. 216.

87 Shiji 40.1700, 40.1708, 73.2334, 77.2383, 79.2417, 81.2449, 89.2584, 100.2733, 104.2776, 123.3175, 126.3200. There is a comparable number of 敢有.

88 Wuwei Xian Bowuguan 1984, pp. 36–37; Juyan xinjian shijiao, p. 907.

89 Miyazaki Ichisada 1977, pp. 230–266.

90 For excavated texts we have Zhang Feng 2016, Tan Shengli 2018, and Zheng Banghong 2019.

91 The following discussion builds on the CHANT corpus of Han and pre-Han texts at http://www.chant.org. For a preliminary organisation of this material, see Petersen 2020.

92 著, Mozi jiaozhu, p. 489, n. 22; 屠, Lüshi chunqiu jishi, p. 1116, n. 58.

93 嗜, Huainanzi jiaoshi, p. 1527, n. 6; 若, Lüshi chunqiu jishi, pp. 548–549, 576, Han Shi waizhuan jishi, p. 133; 居, Lüshi chunqiu xin jiaoshi, p. 842, n. 28; 皆, Guanzi jijiao, vol. 1, pp. 493–494; 百, Lüshi chunqiu xin jiaoshi, p. 1376, n. 29; 音, Yanzi chunqiu jishi, p. 323, n. 4; 自, Yantie lun jiaozhu, p. 85, n. 22; 有, Lüshi chunqiu jishi, p. 682, Huainanzi jiaoshi, p. 1898, n. 6, Lunheng jiaoshi, pp. 515, 1000; Mozi jiaozhu, p. 196, n. 124.; 日, Lunheng jiaoshi, p. 1046.

94 著, Xinshu jiaozhu, p. 106, n. 19; 吉, Baihu tongyi jiaokanji, 4, 12b; 若, Chunqiu Zuozhuan zhengyi, p. 1620, n. 4; Mozi jiaozhu, p. 502, n. 113; 君, Shuoyuan jiaozheng, p. 17; cf. Xinshu jiaozhu, p. 262; 耆, Huainanzi jiaoshi, p. 1004, n. 5; Mozi jiaozhu, p. 279, n. 89; 嗜, Yantie lun jiaozhu, p. 176, n. 16; 智, Huainanzi jiaoshi, p. 1607, n. 49; 有, Da Dai liji huijiao jijie, p. 581, n. 7; 曰, Zhouli zhushu, p. 1083, n. 3; Yanzi chunqiu jishi, p. 393, n. 5, Hanguan liu zhong, p. 169, n. 155; 其, Lunheng zhushi, p. 573, n. 2; Qianfu lun jian jiaozheng, p. 236, n. 5; 是, Xunzi huijiao huizhu, p. 575, n. 29.

95 Ji Xusheng 2014, pp. 219, 275.

96 Guodian Chumu zhujian 性自15, 性自16, 六德 24. Both nominal and verbal uses of 箸 are attested in pre-Han texts; He Linyi 2004, p. 520; cf. pp. 515, 521.

97 For Qin character forms, see Fang Yong 2012, pp. 84 (115), 102–103 (133); for Han forms, see Xu – Xiao 2016, pp. 441–443, 501–504. The evolution of the two graphs is traced in Li Xueqin 2012, pp. 231, 290.

98 Yuelu shuyuan cang Qinjian, vol. 4, pp. 9, 103 (1278 正).

99 Yuan Zhongyi et al. 1993, pp. 388, 496.

100 Chen Songchang 2001, 3.20, 4.38.

101 Zhang Shouzhong et al. 2000, pp. 65–66 – the Guodian corpus apparently does not contain the graph 書; Pian Yuqian 2001, pp. 124–132.

102 I believe that the second stage was immediately rationalised as the third stage and is thus nowhere attested.

103 Mizusawa Toshitada 水澤利忠 (Shiki kaichū kōshō fu kōhō 6, p. 37) notes 者 in the following editions: 瀧 (Shiki kaichū kōshō), 慶 (III1.6), 殿 (III4.1), 凌 (III3.6), 彭 (III2.2), 金陵 (III4.2), 韓 (III6), and 嵯 (III7.1). All other editions (ge ben 各本) lack it. Among the editions with 者, only the Qing 慶 edition (Huang Shanfu’s edition from 1195–1200) is a Song edition. However, this edition is the Patchwork edition and Mizusawa is probably mistaken in holding it to read 者. However, in his Zhaji 札記 (quoted by Mizusawa) Huang notes that he has supplied 者 from Wu jiao Songban 吳校宋版, indicating that it was not original with his base edition. Mizusawa presumably refers to this emendation. See Shiki kaichū kōshō fu kōhō 6, p. 37 for the sigla Mizusawa uses in his 校讎資料一覽. The numbers in parentheses refer to Mizusawa’s classification of the various editions.

104 Nienhauser et al. 1994–, vol. 1, pp. xiii-xiv. Neither Takigawa nor the editors of the Zhonghua shuju edition utilise these editions.

105 Lunheng jiaoshi 7.355. The 者 appears in the “cumulative” (遞修) edition (宋乾道三年原刻，元、明兩朝先後補刻三十卷本《論衡》，簡稱「遞修」本); Lunheng quanyi 25.483. Yet the Song edition from the reign of Xiaozong 孝宗 (1127–1194, r. 1162–1194), kept in the Kūnaichō shoryō 宮内庁書陵, contains no 者 (https://db2.sido.keio.ac.jp/kanseki/T_bib_frame.php?id=007407, accessed 3 June 2022). The textual transmission of Lunheng is known to be very shaky.

106 秦皇令，敢有挾持[>詩]書偶語者俱為城旦也. The corruption xia chi shu 挾持書 is obviously related to the fact that the Qin law against harbouring books was called xiashu lü 挾書律, at least in the Han period (Hanshu 2.90). Xiachi 挾持 is a common compound; presumably after 詩 had corrupted into 持, 挾 was supplied.

107 秦爲 … 畏人之議已，於是嚴誹謗訞言之法而偶語者且棄市. (Qin […] feared that people would criticise it and then toughened the laws against slander, and those who ouyu were all executed in the marketplace). The 且 is here a corruption of 俱 (具).

108 See McMullen 1988, pp. 203–205.

109 The headword has a 有, which must ultimately derive from the 敢有 of the Shiji account. The entry itself, however, does not have this character, so in this respect the headword is fuller than the entry, speaking to some degree against the possibility that abbreviation is at stake.

110 The Shang dynasty is said to have enforced a ban on “throwing ashes in the streets” (弃灰於街), and Shang Yang 商鞅 (390–338) seems to have introduced a similar law in the Qin period, though his motivations remain unclear; see Han Han 2013.

111 The ban on possessing books appears to have become divorced from the ban against voicing political criticism – Liu Bang must have repealed the latter law, but “the law against possessing books” was only repealed in 191; Hanshu 2.90.

112 Qianfu lun jian jiaozheng 5.40.

113 秦皇馭宇，吞滅諸侯，任用威力，事不師古，始下焚書之令，行偶語之刑，先王墳籍，掃地皆盡. Suishu 49.1298.

114 戮逮棄灰，誅及偶語. Tanglü shuyi, p. 577; cf. Tanglü shuyi yizhu, p. 5. “Throwing ashes in the streets” (see above) is mentioned as an example of an excessively punished petty transgression. Suishu 25.696 also mentions the two Qin laws together (棄灰偶語).

115 Jingjin Dongpo wenji shilüe 31, 6b.

116 及平六國，制夫人藏詩書及偶語棄市，以古非今者族. Tongdian 163.4195. 夫 occurs in the Siku quanshu edition as 曰; this is probably a rationalisation. 夫 is more likely a corruption of 天 in 天下, as Taiping yulan (I-4).

117 Shanghai Bowuguan cang Zhanguo Chu zhushu, vol. 6, p. 178.

118 Zhang Chongli 2007 (not seen), referred to in Liang Jing 2008, Ni Weichun 2009, pp. 171–174, and Nohara Masaki 2009.

119 Liang Jing 2008.

120 He Youzu 2007.

121 Liang Jing 2008.

122 Xin jiaoding liujia zhu Wenxuan 46.3039, 3053.

123 Huainanzi jiaoshi 8.874.

124 所計百許篇，率皆鄙質如偶語. Zhong Rong Shipin jianzheng gao B.214.

125 The modern commentator Cao Xu 曹旭 also has 相對私語; Shipin jizhu B.207.

126 Song Hong 2005. Song Hong points to the occurrence in Jiu Tangshu of 鄙辭俚語 and to the occurrence of bili 鄙俚 in Shiji and Sanguo zhi and moreover speculates that Hu Yinglin 胡應麟 (1551–1602) in his Shisou 詩藪 may be drawing from an edition of the Shipin which reads 俚語.

127 Kong Yingda’s commentary to Mao Shi ##27, 167, Mao Shi zhengyi 2.140b, 9.688b; Jiang Huai yiren lu 江淮異人錄, quoted Bai Kong liu tie 2.4, 14b; Cefu yuangui 480.5430.

128 This construction appears only from the Song on; see, e.g., Zhuzi quanshu 121.2845. Alfred Forke translates the passage from the Qin law, using this late sense, as “those who perchance should dare to discourse on poetry and history”; Forke 1907, vol. 1, p. 491. This is probably anachronistic.

129 Petersen 1995, pp. 12–16.

130 The former article makes a number of other misstatements, such as misrepresenting Ulrich Neininger’s views on the Burning of the Books and neglecting the evidence on baijia 百家 presented by Xunzi 荀子.

131 Wenxin diaolong jiaozhu tongyi 15.401–9; Shih 1959, pp. 170–175.

132 若夫事或孤立，莫與相偶，是夔之一足，趻踔而行也. Wenxin diaolong jiaozhu tongyi, 15.406; Shih 1959, p. 174.

133 其三讓公封，理周辭要，引義比事，必得其偶. Wenxin diaolong jiaozhu tongyi 22.266, Shih 1959, p. 178.

134 Huainanzi jiaoshi 21.2174.

135 Shi xiansheng aolun zhu, 前集 3, 雜論, 秦.

136 率皆立主客，使之相對語，故云「偶言」。又音寓，寓，寄也。故別錄云「作人姓名，使相與語，是寄辭於其人，故莊子有寓言篇」. Shiji 63.2144; cf. Qilüe yiwen, pp. 46–47.

137 寓言十九，藉外論之。親父不為其子媒。親父譽之，不若非其父者也. Zhuangzi jiaoquan 27.1089.

138 One is tempted to read 重言 as chongyan “doubled words,” that is, repeating the words of wise men of the past, but the commentary tradition holds 重 to have a qusheng 去聲 pronunciation.

139 Wang Niansun 2014, pp. 324–325; Shuowen jiezi gulin #5160.8207–8; Xin Deyong 2019, pp. 125–126. They evidently see “representation” (ji 寄) and “doubling” (ou 偶) as categorically different. I do not believe that this is necessarily the case: a statue could well be conceived as a double of what it represented.

140 Liang Hanji 1. 29.513; Zizhi tongjian 35.1122.

141 During the Han, streets were personified as a vox populi, directly pronouncing on current events; see Shiji 97.2703, 41.1755, Hanshu 64A.2782, 64B.2836, 77.3262.

142 Lunyu zhushu 19.4.292b; for a translation, see Lau 1992, p. 191.

143 This alludes to Mao Shi #254 (Mao Shi zhengyi 17.1347; Karlgren 1950, pp. 212–213): “The ancient people had a saying: ‘Consult with the grass- and fuel-gatherers’” (先民有言，詢于芻蕘).

144 小說家者流，蓋出於稗官。街談巷語，道聽塗說者之所造也。孔子曰：「雖小道，必有可觀者焉，致遠恐泥，是以君子弗為也。」然亦弗滅也。閭里小知者之所及，亦使綴而不忘。如或一言可采，此亦芻蕘狂夫之議也. Hanshu 30.1745. The last sentence alludes to Mao Shi #100 (Mao Shi zhengyi 5.395; Karlgren 1950, p. 65): “He breaks the willows of the fenced garden, the mad fellow in his flurry” (折柳樊圃，狂夫瞿瞿). See also Fölster 2016, pp. 375–376.

145 Loewe 2000, pp. 502–503.

146 For studies of the Qin and early Han baiguan, see i.a. Zhao Yan et al. 2009, Cao Lüning 2005, pp. 192–195, Wang Qizhou et al. 2015.

147 如淳曰：「稗音鍛家排。九章『細米為稗』。街談巷說，其細碎之言也。王者欲知閭巷風俗，故立稗官使稱說之。今世亦謂偶語為稗。」師古曰：「稗音稊稗之稗，不與鍛排同也。稗官，小官。漢名臣奏唐林請省置吏，公卿大夫至都官稗官各減什三，是也. Hanshu 31.1745. See also Fölster 2016, p. 375, n. 248. Fölster unfortunately follows my 1995 article and translates ouyu as “analogizing talk.”

148 Cf. Hou Hanshu 31.1094 and Shishuo xinyu jiaojian 24.578, translated Mather 1976, p. 393.

149 Huijiao Jiu zhang suanshu 2.69–71, Chemla – Guo 2004, pp. 222–227, where 粺米 occurs repeatedly.

150 Pan Jianguo 1999, p. 77. Donald Holzman uses the word “weed” in this connection; Holzman 2003, p. 77.

151 Durrant et al. 2016, Ding 10.02, p. 1797.

152 Zhuangzi jiaoquan 22.828.

153 Gassmann 2016, vol. 2, p. 195, n. 472.; Mengzi zhushu 11B.373.

154 See Zhang Xifeng 2004, pp. 302–309; Yin Jiming 2007, pp. 307–310.

155 Nan Qi shu 7.107 (written as 裨販); Liangshu 1.6.

156 Lu Guimeng 陸龜蒙 (d. 881), Youju fu xu 幽居賦序, Quan Tangwen 800, 17b, p. 8400b.

157 Yu Jiaxi 1963, p. 266. This however does not keep him from engaging in lengthy speculation about the baiguan.

158 Shiji 6.255.

159 In Zhang Heng’s 張衡 (78–139) Xijing fu 西京賦 (Western Capital Rhapsody), jietan xiangyi is associated with “itinerant persuaders” who “discoursed in the streets and debated in the alleyways,” and could make or unmake a man’s reputation; Xin jiaoding liujia zhu Wenxuan 2.88; Knechtges 1982–1996, vol. 1, p. 205.

160 Chen Hong 2007, quoting Chen Guanghong 2009.

161 For a similar text where 作者 means “writer,” see Wu Zhi 吳質 (177–230), Da Dong Ewang shu 答東阿王書, Xin jiaoding liujia zhu Wenxuan 42.2790.

162 夫書畫辭賦，才之小者，匡國理政，未有其能。陛下即位之初，先涉經術，聽政餘日，觀省篇章，聊以游意，當代博弈。非以教化取士之本。而諸生競利，作者鼎沸。其高者頗引經訓風喻之言；下則連偶俗語，有類俳優；或竊成文，虛冒名氏. Hou Hanshu 60B.1996. Translation based on Baccini 2011, pp. 169–170.

163 Hou Hanshu 60B.1992.

164 Baccini 2011, p. 170.

165 Xinshu jiaozhu 5.172–337.

166 Jao Tsung-I 1987.

167 Chunqiu Zuozhuan zhengyi Ding 10, p. 1829.

168 Shuowen jiezi gulin #4382.7165; cf. Guxun huizuan, p. 1629.

169 Xin Deyong 2019, p. 132.

170 Miyazaki Ichisada 1977, p. 46.

171 Zhu Xiaohai 2000.

Additional information

Notes on contributors

Jens Østergaard Petersen

Jens Østergaard Petersen is an independent scholar researching early Chinese historiography and philology. His publications include “The Zuozhuan Account of the Death of King Zhao of Chu and Its Sources,” Sino-Platonic Papers 159 (August 2005) and “The Zuozhuan Story about Qi Xi’s Recommendations and Its Sources,” Sino-Platonic Papers 255 (2015). The article “Which Books Did the First Emperor Burn? On the Meaning of the Expression Pai Chia 百家 in Early Chinese Sources,” Monumenta Serica 43 (1995), 1–52, contains his earlier attempt to grapple with the problems taken up in the present article.