Guǎngyǎ 廣雅

The Expanded Yǎ by 張揖 (Zhāng Yī, 撰)

About the work

A Three-Kingdoms Wèi expansion of the Ěryǎ, modelled on its nineteen piān but enlarging the lexical inventory by drawing on the Sāncāng 三蒼, the Shuōwén jiězì KR1j0018, the Fāngyán KR1j0006, and the various Hàn glossators. Zhāng Yī 張揖 was an Imperial Academician under the Tàihé 太和 reign of Wèi Míngdì (227–232); the Guǎngyǎ is one of three lexicographic works of his explicitly recorded in the bibliographic record, the other two (Pícāng 埤倉 and Gǔjīn zìgǔ 古今字詁) being lost.

Tiyao

Your servants etc. respectfully report: Guǎngyǎ in ten juàn; composed by Zhāng Yī of the Wèi. Yī’s was Zhìràng 稚讓; he was a man of Qīnghé 清河 and during the Tàihé era held the office of Imperial Academician (博士). His given name 揖 is sometimes written 楫 [with the wood radical], but the parallel of his “Zhìràng” — the ràng of yīràng 揖讓 — confirms that the surname-graph is correctly 揖. Northern-Wèi Jiāng Shì’s 江式 Memorial on the Script says: “In the early Wèi, the academician Zhāng Yī of Qīnghé wrote Pícāng, Guǎngyǎ, and Gǔjīn zìgǔ — these inquire deeply into the Pí[-cāng] and Guǎng[-yǎ] and enlarge the lexical categories; they are an addition also to writing-craft. But his Zìgǔ, when set against Xǔ Shèn’s Shuōwén, is sometimes right and sometimes wrong.” That is, Jiāng Shì held that the Pícāng and Guǎngyǎ surpass the Zìgǔ. Today the Pícāng and Zìgǔ are long lost; only the Guǎngyǎ survives. The book follows the old Ěryǎ topical headings and gathers the annotations of various , the Sāncāng, the Shuōwén, and other works to enlarge it; nothing is omitted from Yáng Xióng’s Fāngyán. The Suí imperial library scholar Cáo Xiàn 曹憲 supplied a phonetic gloss; to avoid the taboo on Suí Yángdì’s [Yáng Guǎng’s] given name, he renamed it Bóyǎ 博雅; the two names co-exist down to today, but it is one book. The work has Zhāng Yī’s Memorial of Presentation prefixed, which reports a total of 18,150 graphs in three juàn (upper, middle, lower); the Suíshū jīngjízhì also gives three juàn, in agreement. But the Suízhì note records “in Liáng there were four juàn”; the Tángzhì also gives four juàn; the Guǎngé shūmù says: “today lost; only the three-juàn phonetic-gloss [version] survives.” Cáo Xiàn’s annotated edition the Suízhì gives as four juàn; the Tángzhì gives as ten juàn. The juàn counts are inconsistent. — Zhāng’s book was originally three juàn; the Qīlù records four because later copyists divided the piān. Cáo Xiàn’s annotated four-juàn version followed the Liáng configuration; later, with the gloss running long, it was further divided into ten; objection then arising that ten was over-divided, it was re-merged into three. The passages cited from the Guǎngyǎ by various sources are all present in the current text without loss, so the juàn-count differs but the book does not differ. — Hence what the Guǎngé shūmù calls “lost” means lost is the unannotated version; the surviving “phonetic-gloss three juànis Cáo Xiàn’s annotated version. Zhāng’s original was attached to the gloss to survive; it was never lost and never gapped. Only the present text now stands at ten juàn, this being a later subdivision to bring it into line with the Tángzhì. — Reading Táng Xuándù’s 唐玄度 preface to Jiǔjīng zìyàng KR1j0025: the conversion of pronunciation-glosses from fǎn 反 to qiè 切 in fact begins in the Kāichéng era of the Táng. Cáo Xiàn passed from the Suí into the Táng, was alive into the Zhēnguān era, but is far earlier than Kāichéng. Yet the present text often reads “such-and-such qiè” — a real difficulty: this is presumably a copyist’s emendation in subsequent editions, no longer Cáo Xiàn’s own. Respectfully edited and presented in the ninth month of Qiánlóng 46 (1781).

Abstract

The Guǎngyǎ is the first systematic expansion of the Ěryǎ — broadly preserving the latter’s nineteen-section topical structure but bringing in vocabulary from later Hàn glossatorial work. Bibliographically the work has had a tangled life: it survives essentially because Cáo Xiàn 曹憲 of the Suí–early Táng provided a phonetic gloss (eventually titled Bóyǎ 博雅 to dodge the Suí Yángdì taboo on guǎng 廣 / yáng 揚 puns; both titles refer to the same book). Internal juàn-counts vary across catalogs (3, 4, or 10) but reflect editorial division of the same body of material rather than textual loss. The Sìkù editors detected anachronistic qiè 切 readings in the surviving Guǎngyǎ yīn and infer Táng-Sòng-period emendation of the original fǎn 反 notation. The classical Qing study, Wáng Niànsūn’s 王念孫 Guǎngyǎ shūzhèng 廣雅疏證 (1796), made the Guǎngyǎ central to Old-Chinese reconstruction by treating its synonym pairs as a phonetic-cognate corpus. Wilkinson §6.2.1 and §6.3 lists it among the foundational pre-Táng lexicographical works.

Translations and research

  • Wáng Niànsūn 王念孫. 1796. Guǎngyǎ shū-zhèng 廣雅疏證. Modern reprint: Beijing: Zhonghua, 1983. — The classic Qing variorum, foundational for Old Chinese phonological reconstruction.
  • Schuessler, Axel. 2007. ABC Etymological Dictionary of Old Chinese. Honolulu: University of Hawaiʻi Press. — Cites Guǎngyǎ glosses extensively as Old-Chinese cognate evidence.
  • Endymion Wilkinson. 2022. Chinese History: A New Manual, §6.2.1.