Yóuxuān shǐzhě juédài yǔshì biéguó fāngyán 輶軒使者絕代語釋別國方言
Glosses of the Light-Carriage Envoys on the Speech of Past Ages and the Languages of Different States — generally cited as Fāngyán 方言 by 揚雄 (Yáng Xióng, 撰 — attributed) with annotations by 郭璞 (Guō Pú, 注)
About the work
The Fāngyán is the earliest extant comparative-dialect dictionary of Chinese. In its received thirteen-juàn shape it lists some 11,900+ characters’ worth of regional vocabulary, organized in nineteen sections by topic (kinship, body parts, tools, natural-history categories, etc.), each entry pairing a “common” word with regional equivalents from the lands of Qín, Jìn, Zhào, Wèi, Zhōu, Qí, Lǔ, Yān, Dài, Chǔ, Wú, Yuè, Sòng, Zhèng, Hán, Shǔ, Liáng, and so forth. The work is conventionally credited to the Western-Hàn polymath Yáng Xióng 揚雄 (53 BCE – 18 CE), the Fāngyán in his name being the chief Hàn philological monument apart from the Shuōwén. Guō Pú 郭璞 (276–324) supplied the standard commentary, in part on the basis of his own knowledge of fourth-century regional speech.
Tiyao
Your servants etc. report: Fāngyán in thirteen juàn. The old text titles the author as Yáng Xióng of the Hàn, with notes by Guō Pú of the Jìn. Investigating the Jìnshū’s biography of Guō Pú, there is a record that he annotated the Fāngyán; but the Hànshū’s biography of Yáng Xióng, while listing his works exhaustively, does not contain so much as a single mention of the Fāngyán; and the Yìwénzhì records under Xiǎoxué only Yáng’s Xùnzuǎn in one piān; under Rújiā only Yáng’s Xù in 38 piān (the Tàixuán in 19, Fǎyán in 13, Yuè in 4, Zhēn in 2); and under Záfù only Yáng’s fù in 12 piān — none of these is the Fāngyán. Nor in 190 years of the Eastern Hàn does anyone refer to Yáng as the author of a Fāngyán. It is only at the end of the Hàn that Yīng Shào 應劭’s preface to the Fēngsú tōngyì first asserts that the Zhōu and Qín used to dispatch yóuxuān shǐ in the eighth month of each year to gather the speech of distant regions, that this practice was lost at the fall of Qín, that Yán Jūnpíng 嚴君平 (Zhuāng Zūn) of Shǔ had a thousand and more such words and Línlǘ Wēngrú 林閭翁孺 had a rough method, and that Yáng Xióng, fond of these matters, “interrogated xiàolián and wèizú from across the empire, sequenced and supplemented [their reports] for twenty-seven years until at last he had it set in order, totalling nine thousand graphs.” Yīng Shào also cites in his commentary on the Hànshū one passage of the Fāngyán under Yáng Xióng’s name. So the attribution of the Fāngyán to Yáng begins with Yīng Shào. Wèi and Jìn scholars repeated this view without dissent; only Sòng’s Hóng Mài 洪邁, in Róngzhāi suíbǐ, began to cross-reference the Hànshū and conclude that this was certainly not Yáng’s work. Yet his selected proofs — that the included Liú Xīn—Yáng correspondence claims to be of Chéngdì’s reign and so should not call him “the Filial-Achieving Emperor”; that Eastern-Hàn Míngdì only began the taboo on zhuāng 莊, so the late-Western-Hàn cannot have called Zhuāng Zūn “Yán Jūnpíng” — do not actually go to the heart. The “Chéngdì” notation at the head of the book is a later editor’s tag run together with the original; Liú Xīn’s text in fact reads Yáng Zhuāng 揚莊, not Yán 嚴, and shows no anticipation of Míngdì’s taboo (the Yán Jūnpíng may be a copyist’s update). Neither point is decisive. The strongest evidence: Xǔ Shèn’s Shuōwén jiězì of the Eastern Hàn cites Yáng [Xióng]‘s sayings repeatedly, but none of these is in the present Fāngyán; and many of Xǔ’s glosses agree exactly with the Fāngyán without naming the source — i.e., in Xǔ’s day this book was not yet known as the Fāngyán and was not yet ascribed to Yáng. So Mǎ Róng, Zhèng Xuán 鄭玄 and the rest never refer to it. Yīng Shào is the source of the attribution; Sūn Yán 孫炎 (Wèi) glossing mòhé tángláng móu in the Ěryǎ and Dù Yù (Jìn) glossing shòu shī zǐ yān in the Zuǒzhuàn are the first to quote it transitively, until at last Guō Pú in the Eastern Jìn provided his commentary; and from then on “Yáng Xióng’s Fāngyán” became conventional. — Yīng Shào’s preface gives 9,000 graphs; the present text has more than 11,900, hence ~3,000 graphs in excess. Yáng’s exchange with Liú Xīn calls the Fāngyán fifteen juàn; Guō Pú’s preface speaks of “three [parts] and five [chapters]”; but the Suízhì and Tángzhì both record “Yáng Xióng Fāngyán thirteen juàn” — agreeing with the present text — i.e., the juàn-count is two short of the original. Both discrepancies present serious tensions. Examining Yáng’s reply to Liú Xīn, “the speech is sometimes in cross-grained inversion, I am still in the midst of pondering, gathering it in detail; if I can have an extension, I will not begrudge [the labor],” etc., we suspect that Yáng did indeed have such an unfinished compilation in hand; Liú Xīn requested a reading-copy and did not get one — that is why the Qīlüè does not list it and the Hànzhì does not record it. Later, perhaps Hóu Bā 侯芭 and his circle gathered up Yáng’s drafts, transmitting them privately, and over time accretions accrued — analogous to Xú Xuàn’s 徐鉉 additions to the Shuōwén, hence the swollen graph-count. Subsequently those who carried on the school, anxious that the Hànzhì lists no Fāngyán, found in the Xiǎoxué class an unattributed Biézì in thirteen piān; they could borrow that title to lend the Fāngyán a paper-trail back to Yáng — and amalgamated the two into thirteen juàn to fit the count, hence the loss of two juàn relative to the original. — In sum, neither side has decisive evidence, and one has had no choice but to follow the old practice and retain Yáng’s name on the title-page, in the spirit of “transmitting doubt with the doubt.” (Liú Xīn–Yáng correspondence is on Lǐ Shàn’s Wénxuǎn commentary’s authority, where the line “to be hung among the sun and moon as a book past expurgation” is already glossed as referring to the Fāngyán — i.e., the correspondence has been appended to the end of the Fāngyán since Suí–Táng, and we retain that placement.) — There exist printed editions, but the diction is archaic, the glosses obscure, and collators have generally despaired of detailed editing; hence corruption and lacunae have been such that the text was scarcely readable. Qián Zēng 錢曾, Dúshū mǐnqiú jì, attempted to correct it from a Sòng print, but his Sòng print is no longer extant; only the Yǒnglè dàdiǎn preserves a complete copy. The entry “Qín had a zhūé terrace” 秦有㯃娥之臺 etc. matches Qián Zēng’s citation, so the Yǒnglè dàdiǎn copy is from the same Sòng line. Comparing it with current reprints, we find Míng-period editors had wantonly altered, transposed, and dropped passages, well beyond what Qián Zēng noted: the book was extant only in name. We accordingly collated cross-checked, corrected 281 graphs, deleted 17 superfluous graphs, supplemented 27 missing graphs — the spiritual brilliance is restored to its original aspect. We also annotated each entry with citations from other sources, gloss by gloss, providing the editor’s ànyǔ 案語, so that the small-school philological tradition can be glimpsed in outline; and we have corrected the errors of the popular printings as a guide for posterity. The original title was Yóuxuān shǐzhě juédài yǔshì biéguó fāngyán; its diction is cumbersome, hence in citations and bibliographic catalogs it is uniformly abbreviated to Fāngyán; the JiùTángshū jīngjízhì gives it as Biéguó fāngyán. They are the same book. Hóng Mài’s Suíbǐ renders the title with juéyù 絕域 (“of distant regions”) in place of juédài 絕代, and has been alone in this; all witnesses read juédài, and the body of the book contains nothing about “distant regions / re-translation,” so Hóng Mài’s reading is a lapse and is not adopted. (Translated from Sìkù tíyào via Kyoto Zinbun digital tíyào at 0083501.html, which is the relevant fascicle.)
Abstract
The Fāngyán is the most important pre-modern Chinese source on the geographical distribution of regional vocabulary; it groups its lemmata both by common-word in modern lexicographic style and by the regions where each variant was spoken. The Sìkù compilers, surveying earlier debate, sit on the fence about authorship: they find Hóng Mài’s case against Yáng Xióng inconclusive but acknowledge there are real difficulties (notably Xǔ Shèn’s silence). Their working hypothesis — that Yáng Xióng left an unfinished collection that Hóu Bā and others rescued and transmitted, with later accretions — has remained the standard scholarly position; modern philological work (including Hú Yírǔ 胡以儒, Zhōu Zǔmó 周祖謨, and W. South Coblin) has continued to refine it. The juàn-count discrepancy (15 in Yáng’s lifetime, 13 in the Suízhì and beyond) is plausibly explained as conflation with an unattributed Xiǎoxué work Biézì shísān piān. Guō Pú’s annotation is a major Eastern-Jìn philological monument in its own right and frequently records fourth-century dialect terms (Shǔyǔ, JiāngDōngyǔ, Hézhōngyǔ) — those very glosses that Zhèng Qiáo KR1j0005 later attempted to remove on the assumption that the Ěryǎ (and by extension the Fāngyán) was a Jiāngnán product. The Sìkù compilers used the Yǒnglè dàdiǎn copy as the recension base, correcting some 281 graphs, deleting 17 redundancies and supplying 27 missing graphs against contemporary popular printings. Yáng’s Hàn-era dating frame here uses notBefore = 10 BCE (when Yáng’s correspondence with Liú Xīn fixes the project as already in progress) and notAfter = 18 CE (the year of his death). Wilkinson §6.3.4.1 stresses that the Fāngyán records dialects of Hàn-period Central Plain — its near-silence on Mǐn, Yuè and Cantonese reflects that those regions had not yet been Sinicized.
Translations and research
- Coblin, W. South. 1983. A Handbook of Eastern Hàn Sound Glosses. Hong Kong: Chinese University Press. — Treats the Fāngyán among the Hàn phonological sources.
- Serruys, Paul L.-M. 1959. The Chinese Dialects of Han Time According to Fang Yen. Berkeley: University of California Press. (UCal Studies in East Asian Linguistics 11.) — The standard Western monograph.
- Bridgman, R. F. 1955. “Le Fang-yen de Yang Hiong.” Bulletin de l’École Française d’Extrême-Orient 47: 1–94. — French digest with annotated tables.
- Zhōu Zǔmó 周祖謨. 1956. Fāngyán jiào-jiān fù tōng-jiǎn 方言校箋附通檢. Beijing: Kexue. — The standard textual edition.
- Hu Hua-shen 華學誠. 2007. Yáng Xióng Fāngyán jiào shì huì zhèng 揚雄方言校釋彙證. Beijing: Zhonghua. — Comprehensive commentary.
- Endymion Wilkinson. 2022. Chinese History: A New Manual, §6.3.4.1.
Other points of interest
The book ascribed to Hóu Bā as inheritor of Yáng’s “draft” is a recurring trope in Sìkù treatments of Western-Hàn philological texts; the same hypothesis is invoked for the Tàixuánjīng tradition. The discrepancy between the 9,000 graph-count of the Yīng Shào preface and the 11,900+ of the present text is the principal reason the Sìkù compilers note the Fāngyán as text-critically unsettled.