¹ In this article the term ‘early Chinese’ is preferred to the more widespread form ‘Old Chinese’. In fact, by ‘Old Chinese’, a term which goes back at least to Joseph Edkins's writings, generally meant the ‘language’ from the Early Zhou (1046–771 bce) to the Western Han (206 bce–9 ce) period. This writer may be mistaken, but so far as he is aware, it seems that no scholar has ever produced concrete textual evidence from contemporary or unearthed documents about the existence of this ‘language’, which is not itself a language like, say, prisca Latinitas (Old Latin), but “a set of hypothetical sound classes”, as rightly pointed out in Coblin, W. S., ‘Review: E. G. Pulleyblank, Lexicon of Reconstructed Pronunciation’, Monumenta Serica 41 (1993) p. 305Google Scholar. On the other hand, terms such ‘early Chinese’ or ‘medieval Chinese’ are more advantageous, so long as they are supposed to indicate a Spracheinheit, intended as an abstraction of linguistic (in this case mainly phonemic) features that may have been common to a given group of speakers, at some time or other.

² The term ‘transcription’ in this article is intended as a synonym for ‘reconstruction’. This writer is aware of the existence of a dichotomy, in the field of historical Chinese phonology, between transcriptions and phonemic reconstructions that goes back at least to W. H. Baxter—see Baxter's, A Handbook of Chinese Phonology (Berlin, 1992), p. 27CrossRefGoogle Scholar. In the opinion of the present writer, this distinction is misleading, being the assessment of phonemic forms, de facto, an act of reconstruction. As rightly pointed out by Luschützky, H. C., in his “Typological Paradoxes in Phonological Reconstruction: The Case of Polabian Reduced Vowels”, Linguistic Reconstruction and Typology 96 (1997), p. 242Google Scholar, even the assessment of input forms from an underlying observable output has to be regarded as a synchronic reconstruction (i.e. the reconstruction of languages spoken hic et nunc). This fact seems to be clear to historical linguists but not to Chinese linguists.

³ See Marshman, J., Dissertation on the characters and sounds of the Chinese language: including tables of the elementary characters, and of the Chinese monosyllables (Serampore, 1809)Google Scholar.

⁴ The following discussion would similarly benefit from some attention dedicated to the subgrouping of Sinitic, but due to space limitations, it is not possible to provide this here, suffice to say that Chinese has been considered a primitive, isolated language for a long time. See Farrar, F. W., Language and Languages (New York, 1878), p. 376Google Scholar.

⁵ The citation is attributed to Golius (1596–1667), quoted in Bayer, T. S., Museum Sinicum (Petropoli, 1730), p. 103Google Scholar.

⁶ Literally “a mortal sermon”.

⁷ Attributed to Elias Grebniz (1682), see Harbsmeier, C. ‘Language and Logic in Traditional China’, Vol. VII of Needham, Joseph (ed.), Science and Civilization in China, 1 (Cambridge, 1998), p. 11Google Scholar.

⁸ See, e.g., Kuhn, E., Beiträge zur Sprachenkunde Hinterindiens (Sitzung, 2 März 1889)Google Scholar; Finck, F. N., Die Haupttypen des Sprachbaus (Teubner, 1909)Google Scholar; Przyluski, J., “Contribution à l'Histoire des Contes Indiens”, Journal Asiatique 205, (1924), pp. 101–137Google Scholar.

⁹ See Whitney, W. D., Language and the Study of Language. Twelve Lectures on the Principles of Linguistic Science (London, 1884), p. 336Google Scholar.

¹⁰ See, e.g., Barrow, J., A Voyage to Cochinchina, in the Years 1792 and 1793: To which is Annexed an Account of a Journey Made in the Years 1801 and 1802, to the Residence of the Chief of the Booshuana Nation (London, 1806)Google Scholar, and Leyden, J., On the languages and literature of the Indo-Chinese nations (Calcutta, 1808)Google Scholar.

¹¹ Ibid., p. 245.

¹² Marshman, J., Dissertation on the characters and sounds (Serampore, 1809), pp. xlii, xlvGoogle Scholar. See also Branner, D. P., “Notes on the beginnings of systematic dialect description and comparison in Chinese”, Historiographia linguistica 24, 3 (1997), p. 246CrossRefGoogle Scholar.

¹³ Ibid., pp. 248–249.

¹⁴ Marshman, Dissertation; Marshman, J., Elements of Chinese Grammar: With a Preliminary Dissertation on the Characters and the Colloquial Medium of the Chinese, and an Appendix Containing the Ta-hyoh of Confucius with a Translation (Serampore, 1814)Google Scholar.

¹⁵ See, e.g., Morrison, R., A Grammar of the Chinese Language, (Serampore, 1815–1823)Google Scholar, II, I (1819), pp. xv-xvii; consider also J. Klaproth, Asia Polyglotta (Paris, 1823), pp. 367–379.

¹⁶ Ibid. p. 247.

¹⁷ Information about Edkins's life comes from Bushell, S.W. ‘Obituary Notices, Reverend Joseph Edkins, D. D.’, The Journal of the Royal Asiatic Society of Great Britain and Ireland (Jan. 1906), pp. 269–271CrossRefGoogle Scholar.

¹⁸ See, e.g., The Chinese Recorder (1905), pp. 282–289, and the second volume of Lovett, Richard, The History of the London Missionary Society, 1795–1895 (London, 1899)Google Scholar. The present writer wishes to express his gratitude to an anonymous peer reviewer for having pointed him in the direction of this source.

¹⁹ Medhurst, W. H., A Dictionary of the Hok-këèn Dialect of the Chinese Language, According to the Reading and Colloquial Idioms: Containing about 12,000 Characters, the Sounds and Tones of Which are Accurately Marked: and Various Examples of Their Use, Taken Generally from Approved Chinese Authors: Accompanied by a Short Historical and Statistical Account of Hok-këèn: a Treatise on the Orthography of the Hok-këèn Dialect; the Necessary Indexes, &c. (Macao, 1832)Google Scholar; Douglas, C., Chinese-English dictionary of the vernacular or spoken language of Amoy, With the principal variations of the Chang-chew and Chin-chew dialects, (Shanghai, 1873)Google Scholar.

²⁰ J. Edkins, Transactions, p. 476.

²¹ Edkins, J., A Grammar of Colloquial Chinese: as Exhibited in the Shanghai Dialect (London, 1868), p. 56Google Scholar.

²² Edkins, J., “Early form of Chinese”, The Chinese Recorder (Shanghai, 1885), p. 251Google Scholar. Also quoted in The Chinese Rime Tables: Linguistic Philosophy and Historical-Comparative Phonology, (ed.) Branner, D. P. (Amsterdam, 2006), p. 216CrossRefGoogle Scholar.

²³ Ibid.

²⁴ In the present article, medieval Chinese forms are cited with preceding asterisks. This writer is inclined to question the practice, still widely diffused in modern scholarship, of omitting asterisks in medieval Chinese forms, on the grounds that they “are based in part on the so-called ‘rhyme tables’ of the Song dynasty, which list words in a grid pattern where one axis represents the initial consonant and the other the quality of the vowel” (see Handel, Zev, “Northern Min tone values and the reconstruction of ‘softened initials’,” Language and Linguistics 4, 1 [2003], p. 547Google Scholar). Thus, e.g., Zev Handel, in “Methodological considerations in the application of Tibeto-Burman comparison”, Studies on Sino-Tibetan Languages, papers in honour of Professor Hwang-cherng Gong on his seventieth birthday, (eds.) Ying-chin Lin et al. (Taipei, 2004), p. 609, believes that “[t]he traditional notational device of preceding an Old Chinese form with an asterisk is therefore misleading. Old Chinese forms, like Middle Chinese forms, are attested”, Notwithstanding this (and similar) opinion(s), the traditional comparative method, at least as explicated in Hoenigswald, H. M., Language Change and Linguistic Reconstruction (Chicago, 1965), p. 64Google Scholar, aims at producing as output a phonemic inventory, which is supposed to be a more or less close approximation to that of the language in question, by taking as input phonemically transcribed lexical items from related languages. As explained by Faber, A., “Interpretation of orthographic forms”, in Patterns of change, change of patterns: linguistic change and reconstruction methodology, (ed.) Baldi, Philip (Berlin and New York, 1990), p. 619Google Scholar, “[t]he approximate nature of the reconstructed system is a direct consequence of the data loss inherent both in phonemic analyses of the attested languages and in the process of reconstruction itself”. It follows that even in those cases like Proto-Romance, where independent evidence in the form of written Latin are available along with a bulk of lexical examples written with “glottographic” (cf. Sampson, G., Writing Systems [Stanford, 1985], p. 32Google Scholar) writing systems, asterisks cannot be omitted as a consequence of the fact that we cannot establish whether the divergences between reconstructed and real forms are due to structural deviations between the two languages (viz. Latin and Proto-Romance) or to the inadequacies of the comparative method. Given that we cannot even compare the sound classes of rime dictionaries with the phonemic inventory of real medieval Chinese (if any), and that, as rightly pointed out by Faber (“Interpretation of orthographic forms”), reconstructions based on epigraphic corpora are necessarily more distant approximations, it is the opinion of the present writer that medieval Chinese forms should be cited with a preceding asterisk.

²⁵ See, e.g., dì 地 /tei˨/ ‘earth’ and 帝 /tɐi˧/ ‘emperor’.

²⁶ See J. Edkins, “Introduction”, in Williams, S. W., A Syllabic Dictionary of the Chinese Language (Shànghǎi, 1874), p. xxixGoogle Scholar. Moreover, Edkins realised that Fújiàn dialects (it is not clear whether Southern, Northern or Eastern Mǐn) retained many features of the pre-medieval version of Chinese, such as k- for *h- (xiá initials 匣母), p- for *f- (fū initials 敷母), m- or b- for *w- (wēi initials 微母), t- for *ch- (zhī initials 知母), n- for *j- (rì initials 日母), etc. (Edkins, 1879–80:89). That, in his view, those dialectal pronunciations were somewhat older than their alleged value in medieval Chinese is clear from Edkins, J., The Evolution of the Chinese Language (London, 1888)Google Scholar. Medieval Chinese reconstructions are indicated in many works, see, e.g., Edkins, J., “Proofs of Ancient Chinese Sounds”, The China Review 21, 5 (1895), p. 414Google Scholar; Edkins, J., “Introduction”, in Williams, S.W., A Syllabic Dictionary of the Chinese Language (Shànghǎi, 1874), p. xxviiiGoogle Scholar; Edkins, A Grammar of Colloquial Chinese, p. 44; Edkins, J., A Grammar of the Chinese Colloquial Language Commonly Called the Mandarin Dialect (London, 1864), p. 84Google Scholar.

²⁷ Edkins, J., Introduction to the Study of Chinese Characters (London, 1876), p.186Google Scholar.

²⁸ Ibid. p. 189.

²⁹ Ibid. p. 206.

³⁰ Ibid.

³¹ Edkins, A Grammar of the Chinese Colloquial Language, p. 32.

³² Ibid. p. 85.

³³ Ibid. p. 95.

³⁴ The second set of tables was based on the tradition of the Sìshēng děngzi 四聲等子, which probably dates back to the Northern Sòng period (960–1126).

³⁵ Edkins's interest in the way that Chinese scripts were read is plainly reflected by his interest in Sinoxenic material, since they represented the most clear-cut example of a system of readings that were mostly unmatched by any vernacular language. To make a concrete example of how Sinoxenic readings were utilised, Edkins realised that Cochin Chinese agreed with many Southern dialects in treating the sibilant initial s- as a voiceless dentalveolar stop t-, e.g. sim ‘heart’ was tim in Cochin Chinese, see Edkins, Introduction to the Study of Chinese Characters, p. 198. Furthermore, although Edkins was aware of the fact that “[i]n the whole horizon of philology there is perhaps no greater chaos at first view to be found anywhere than in the Japanese transcriptions of Chinese sounds” (J. Edkins, “Influence of Chinese dialects on the Japanese pronunciation of the Chinese part of the Japanese language”, Transactions [1879–1880], p. 160), he relied heavily on Japanese transcriptions, and wrongly interpreted tō’on 唐音 as a sort of metropolitan pronunciation representing the variety of Chinese as spoken during the Táng dynasty (618–907), see Edkins, Introduction to the Study of Chinese Characters, p. 180. Nevertheless, this is probably a misconception, for tō’on readings, introduced by the monks of Ōbakusan 黄檗山 at the end of the seventeenth century, were supposed to be the ‘Chinese sound’ of the sinographs at the time of its introduction. Therefore, what it might reflect is probably the pronunciation of a Míng (1368–1644) variety of Mandarin, see Aston, W. G., A Grammar of the Japanese Written Language (London, 1904), p. ivGoogle Scholar; see also Satow, E. M., “Reply to Dr. Edkins on ‘chi’ and ‘tsu’”, Transactions (1879), p. 165Google Scholar. Unfortunately, misconceptions about Japanese pronunciations are still widely diffused even in modern scholarship. Go'on 吳音 readings (lit. ‘sounds from the Wú area’) are believed to represent a reading based on the classical pronunciation of sinographs of the eastern Jiànkāng 建康 dialect (see Pulleyblank, E. G., Middle Chinese [Vancouver, 1984], p. 62Google Scholar, Lexicon of Reconstructed Pronunciation [Vancouver, 1991] p. 2; Baxter, W. H., A Handbook of Old Chinese [Berlin and New York, 1992], p. 14CrossRefGoogle Scholar), while kan'on 漢音 (Chinese sounds) readings should indicate a more recent form of Sino-Japanese borrowed in Táng epoch (see Pulleyblank, Lexicon, pp. 2–3; and Miyake, M. H., Old Japanese: A Phonetic Reconstruction [London and New York, 2003), pp. 103–104Google Scholar). However, as pointed out by Miller, what the two terms seem to indicate is a distinction between ‘official sounds’ (guānhuà 官話) and ‘dialectal sounds’. This is also clear from the fact that go'on readings are often called wa'on 和音 ‘Japanese sounds’, while kan'on readings are also referred to as sei'on 正音 ‘correct sounds’. For further understanding, see Miller, R. A., The Japanese Language (Chicago, 1967), p. 108Google Scholar.

³⁶ Edkins, Introduction to the Study of Chinese Characters, p. 275.

³⁷ Edkins, A Grammar, p. 95.

³⁸ Zhāng Bǐnglín 章炳麟, “Yīn lǐlùn 音理論 [On the Theory of Sounds]”, in Zhāngshì Cóngshū 章氏叢書 [Mr. Zhāng's selected papers], I (reproduced Shanghai, 1917–1919), p. 18b.

³⁹ This is roughly what Pulleyblank did. He created a ‘language’ from the outset of the Yùnjìng 韻鏡. In his review of Pulleyblank's work, Coblin, commented that “to set up at the outset a fictitious Yunjing dialect and call it ‘Late Middle Chinese’, is to create a ‘language’ from whole cloth. And to then relentlessly cut and polish all available evidence to fit this ‘LMC’ scheme is Procrusteanism at its worst”, see Coblin, “Review: E. G. Pulleyblank”, p. 311.

⁴⁰ Parker, E. H., “More About the Old Language of China”, China Review (1884), p. 116Google Scholar.

⁴¹ Edkins, Introduction to the Study of Chinese Characters, p. 477.

⁴² Edkins, J., “The Old Language of China”, China Review (1896), p. 597Google Scholar.

⁴³ Edkins, J., “The Chinese Old Language”, China Review (1884), p. 1Google Scholar.

⁴⁴ Edkins, A Grammar of the Chinese Colloquial, p. 84.

⁴⁵ Ibid.

⁴⁶ See Edkins, “Early form of Chinese”, p. 337.

⁴⁷ In fact, there are many passages in the early literature which point toward the opposite. See, e.g., Zuǒzhuàn, Wéngōng 13^th year 左傳·文公十三年, Lǚshì chūnqiū, Zhīhuà 呂氏春秋·知化, Zhànguó cè, Qín cè III 戰國策·秦策三, Zhànguó cè, Yàn wáng 戰國策·燕王, Mèngzǐ, Téng Wén gōng, II, vi 孟子·滕文公下(六), etc.

⁴⁸ See, e.g., Edkins, J., “Recent Researches upon the Ancient Chinese Sounds”, China Review 22, 3 (1896), p. 568Google Scholar; Edkins, J., “The Old Initials”, China Review 15, 5 (1887), p. 311Google Scholar; Edkins, A Grammar of Colloquial Chinese, p. 166. Elsewhere, Edkins seems to imply that semantically-related words (allofams?) should be similar in pronunciation, see Edkins, J., “All Roots Labial”, China Review 16, 1 (1887), p. 48Google Scholar; and Edkins, J., “Defence of the Old Chinese Pronunciation”, China Review 22, 5 (1897), p. 731Google Scholar.

⁴⁹ Edkins, “Early form of Chinese”, p. 251.

⁵⁰ Marshman, Elements of Chinese Grammar, p. 90f.

⁵¹ See, e.g., Franke, O., “Chinese and Philology”, China Review 20, 5 (1893), pp. 310–318Google Scholar. See also Pezzi, D., Glottologia Aria Recentissima (Torino, 1877), p. 66Google Scholar.

⁵² See Edkins, J., China's Place in Philology (London, 1871)Google Scholar.

⁵³ Lepsius, K. R., Über die Arabischen Sprachlaute und Deren Umschrift: Nebst Einigen Erläuterungen Über den Harten i-Vocal in den Tartarischen, Slavischen und der Rumänischen Sprache (Berlin, 1861), p. 492Google Scholar.

⁵⁴ Ibid., p. 470.

⁵⁵ Ibid., p. 484.

⁵⁶ Grube, W., Die Sprachgeschichtliche Stellung des Chinesischen, (Leipzig, 1881), pp. 16–17Google Scholar.

⁵⁷ In ancient Written Burmese sources Asho Chin people and their relative language were also referred to as Thulung or Thulungya.

⁵⁸ Joseph Edkins is routinely cited in both Chinese and Western works on historical Chinese phonology as the first scholar to have posited initial consonant clusters for early Chinese, a trend that started with Lín Yǔtáng 林語堂 (see, e.g., Lín Yǔtáng 林語堂, “Gǔ yǒu fùfǔyīn shuō 古有複輔音說,” in Yǔyánxué lùncóng 語言學論叢 [Taipei, 1933], pp. 4–5; and Miller, R. A., “N. C. Bodman, A linguistic study of the Shih Ming, Initials and consonant clusters”, T'oung Pao 44 [1956], p. 266Google Scholar) and which has been somewhat reinforced by Professor Chu Chia-ning 竺家寧 (see, e.g., “Gǔ Hànyǔ fùshēngmǔ yánjiū tíyào 古漢語複聲母研究提要,” in Huáxué Yuèkān 華學月刊125 [1982], pp. 54–59). However, the picture that emerges from Edkins's writings is somewhat different, as he apparently rejected the hypothesis according to which consonant clusters may have existed in early Chinese. See, e.g., Edkins, J., Introduction to the Study of the Chinese characters (London, 1876), p. 190Google Scholar.

⁵⁹ Is it not a case that Ferdinand de Saussure also remarked that genetic classification of language without reconstruction is sterile in principle.

⁶⁰ This anecdote is also quoted in Andrews, S. P., Discoveries in Chinese (New York, 1854), p. 21Google Scholar.

⁶¹ For example, notions such as děng 等 ‘Grade’, ‘Division’, ‘Level’, etc., nèizhuǎn 內轉 ‘inner turn’, wàizhuǎn 外傳 ‘outer turn’, chóngniǔ 重紐 ‘repeated button’, etc., are still troublesome and far from having a clear definition. This is perhaps not the place to enter upon a discussion on Grades, but it would be useful to explain the reasons why their nature is still much debated. It has been known since the works of Chalmers, J. (“Kanghi's Dictionaries”, China Review 2 [1873], p. 338)Google Scholar that they are supposed to indicate a quality (front-back dimension) of the vowel and the presence or absence of medial glides; see Pān, W. Y. and Zhāng, H. M., “Middle Chinese Phonology and Qieyun”, in The Oxford Handbook of Chinese Linguistics, (eds.) Wang, William S-Y and Sun, Chaofen (Oxford, 2015), p. 86Google Scholar. If this is true, then these polysemic labels were mere descriptive tools and did not indicate a contrastive quality in the vowel, as they would seem to indicate, for no language in the world shows some variation in vowel quality that contrast in the front-back dimension; see P. Ladefoged and Maddieson, I., The Sounds of the World's Languages (Oxford, 1996), p. 286Google Scholar. Even those languages that exhibit a minimal vocalic system, such as Abkhaz or Bzehdukh (North West Caucasian), do not make any phonemic contrast in the front-back or rounded dimensions, see Catford, J. C., “Mountain of tongues: the languages of the Caucasus”, Annual Review of Anthropology 6, 1 (1977), pp. 293–295CrossRefGoogle Scholar. Volpicelli, Z., Old Chinese Phonology (Shanghai, 1896), pp. 8–14Google Scholar, was the only scholar, to the best of this author's knowledge, to make use of vowel height alone to distinguish the four grades. He also used a simplified version of the Bernoullian lois des grands nombres to test his theory and explain anomalies.

⁶² Chen Hsin-hsiung, Gǔyīn yánjiū 古音研究 [Research on Old Sounds] (Taipei, 1999), p. 9; Shēngyùnxué 聲韻學 [Historical Phonology] (Taipei, 2005), p. 490.

⁶³ Pān Wùyún 潘悟雲. Hanyu lishi yinyunxue 漢語歷史音韻學[Chinese historical phonology] (Shanghai, 2000), p. 80.

⁶⁴ See Karlgren, B., Études sur la Phonologie Chinoise (Uppsala, 1915–1926), p. 98Google Scholar; and also his Compendium of Phonetics in Ancient and Archaic Chinese (Stockholm, 1954), p. 5. However, this was also the view of many historians. See, e.g., Naito Konan 内藤湖南, Naito Konan Zenshū 内藤湖南全集, Vol. 10 Shina jōkoshi 支那上古史 [Chinese Ancient History], (Tokyo, 1997), or Itō Michiharu 伊藤道治, translated by Wú Mìchá 吳密察, Zhōngguó Tōngsh ǐ中國通史 [Comprehensive History of China], (Banqiao, 2010), pp. 170–171.

⁶⁵ See Baxter, W. H. and Sagart, L., Old Chinese: A New Reconstruction (Oxford, 2014), p. 4CrossRefGoogle Scholar. This view, however, has been harshly criticised by Harbsmeier, C. (“Irrefutable Conjectures. A Review of William H. Baxter and Laurent Sagart, Old Chinese. A New Reconstruction”, Monumenta Serica 64, 2 [2016], pp. 478–487)CrossRefGoogle Scholar and Schuessler, A. (“New Old Chinese”, Diachronica 32, 4 [2015], pp. 575–576)Google Scholar. Other scholars hold even more radical views, and think that the very notion of ‘Old Chinese’ is somewhat misleading. See, e.g., Orlandi, G., “Evaluating the Sino-Tibeto-Austronesian Hypothesis”, Journal of Language Relationship 36, 1 (2018), p. 3, fn. 6Google Scholar.

⁶⁶ See Edkins's work: China's Place; “All Roots Labial”, pp. 48–49; Evolution of the Chinese Language.

⁶⁷ See G. Orlandi, “The Linguistic Ideas of Joseph Edkins”, Journal of American Oriental Society, forthcoming.

⁶⁸ Branner, D. P., The Chinese Rime Tables (Amsterdam, 2006), p. 216CrossRefGoogle Scholar.

⁶⁹ Edkins, Introduction to the Study of Chinese Characters, p. 199.

⁷⁰ Ibid., pp. 199, 231, Appendix B.

⁷¹ See Karlgren, B., Compendium of phonetics in Ancient and Archaic Chinese (Stockholm, 1954), p. 366Google Scholar.

⁷² Starostin, S. A., Rekonstruktsiya Drevnekitaiskoi Fonologicheskoi Sistemy (Moscow, 1989)Google Scholar; Baxter, W. H., A Handbook of Old Chinese Phonology (Berlin and New York, 1992)CrossRefGoogle Scholar; Wú Ānqí 吳安其. “Hànzàngyǔ de Shǐdòng hé Wánchéngtǐ Qiánzhuì de Cáncún yǔ Tóngyuán de Dòngcí Cígēn 漢藏語的使動和完成體前綴的殘存與同源的動詞詞根” [Common Verbal Roots and hint of Transitive prefixation in Sino-Tibetan] Mínzú Yǔwén 民族語文, 6 (1997), pp. 21–32; Sagart, L., The Roots of Old Chinese (Amsterdam, 1999)CrossRefGoogle Scholar; Zhèngzhāng Shàngfāng 鄭張尚芳. Shànggǔyīnxì 上古音系 [Old Chinese Phonology], (Shanghai, 2003); Jīn Lǐxīn 金理新. Shànggǔ Hànyǔ Xíngtài Yánjiū 上古漢語形態研究[Studies on Old Chinese Morphology], (Hefei, 2006); Schuessler, A., Minimal Old Chinese and Later Han Chinese: a Companion to Grammata Serica Recensa (Honolulu, 2009)Google Scholar; Baxter and Sagart, Old Chinese.

⁷³ See, e.g., Edkins, J., “Chinese Roots”, China Review 13, 6 (1885), pp. 387–398Google Scholar; “Chinese Roots”, China Review 14, 2 (1885), pp. 67–80Google Scholar; “All Roots Labial”, pp. 48–49; The Evolution of the Chinese Language (London, 1888); “Sixteen Chinese Roots”, China Review 16, 4 (1888), pp. 241–242Google Scholar.

⁷⁴ See Karlgren, B., Grammata Serica Recensa (Stockholm, 1957)Google Scholar; Zhèngzhāng 鄭張, Shànggǔyīnxì 上古音系 [Old Chinese Phonology] (Shanghai, 2003); Schuessler, A., ABC Etymological Dictionary of Old Chinese (Honolulu, 2007)Google Scholar; Schuessler, Minimal Old Chinese and Later Han Chinese; Baxter and Sagart, Old Chinese.

⁷⁵ See, e.g., F. S. Hsue, Hànyǔ Yīnyùnshǐ Shí Jiǎng 漢語音韻史十講 [Ten lectures on the history of Chinese phonology] (Beijing, 1999), pp. 8, 22–23, or Xú Tōngqiāng and Yè Fēishēng, “Lìshǐ bǐjiàofǎ hé ‘Qièyùn’ yīnxì de yánjiū 歷史比較法 和<切韻>音系的研究 [The comparative method and studies on Qièyùn]”, Yǔwén Yánjiū 語文研究1 (1980), p. 29.

⁷⁶ Karlgren, Études, p. 45.

⁷⁷ Volpicelli, Old Chinese Phonology.

⁷⁸ Schaank, S. H., “Ancient Chinese Phonetics”. T’oung Pao 8, (1897), pp. 361–377CrossRefGoogle Scholar, 8, pp. 457–486; 9, pp. 28–57 (1898), and “Supplementary Note”, T’oung Pao Ser. 2, 3 (1902), pp. 106–108.

⁷⁹ This writer may be mistaken, but to his knowledge no historical Chinese phonologist has ever considered, on some inductive basis, the possibility of falsifying the hypothesis regarding the regularity of proposed sound changes. Rather, it seems that most linguists have generally reconstructed the prehistory or the early history of the Chinese language by hunting for cases that may verify theories and hypotheses. See, e.g., Zhèngzhāng, 鄭張, Shànggǔyīnxì 上古音系, p. 67, as perhaps the most explicative example. Quoting mutatis mutandis, S. J. Lieberman (in “The regularity of sound change: a semitistic perspective”, in Baldi (ed.), Patterns of change, pp. 299–324), “[a]ny acceptance of the absolute regularity of sound-change laws would require that we allow for the possibility of ‘falsifying’ the hypothesis, as Karl Popper would have insisted. Linguists have not, however, to my knowledge, considered the question on some inductive basis, and it is hard to see how any test of the regularity hypothesis could be made that would satisfy its advocates. They always insist on the possibility of other, undiscovered, factors modifying the basic rules, and concede the necessarily incomplete nature of our knowledge of all of the details”. It follows that every early Chinese reconstructing system is not scientific in Popperian terms, cum bona pace of Sagart, who has often claimed the contrary (cf. Sagart, The Roots of Old Chinese, p. 10; Baxter and Sagart, Old Chinese, p. 5; personal communication).

⁸⁰ Maspero, H., “Le Dialecte de Tch'ang-ngan sous les T'ang”, Bulletin de l'École Française d'Extreme-Orient 20, 2 (1920), pp. 1–124CrossRefGoogle Scholar.

⁸¹ See Gulya, J., “Some Eighteenth Century Antecedents of Nineteenth Century Linguistics: the Discovery of Finno-Ugrian”, in Studies in the History of Linguistics. Traditions and Paradigms, (ed.) Hymes, D. H. (Bloomington, 1974), p. 271Google Scholar.

⁸² Kuhn, T., The Structure of Scientific Revolutions (Chicago, 1962), p. 10Google Scholar.

⁸³ Mounin, G., Histoire de la Linguistique des Origines au XXe Siècle (Paris, 1967), p. 159Google Scholar.

⁸⁴ Gulya, “Some Eighteenth Century Antecedents”.

⁸⁵ As Schaank did in “Ancient Chinese Phonetics”, see Branner, The Chinese Rime Tables, p. 156.

⁸⁶ Chalmers, “Kanghi's Dictionaries”.

⁸⁷ Kühnert, F., “Zur Kenntniss der alteren Lautwerthe des Chinesischen, Stzungsberichte der Kaiserl”, Ak. d. Wissenschaften in Wien, phil-hist, (Vienna, 1890)Google Scholar.

⁸⁸ Schaank, “Ancient Chinese Phonetics”.

⁸⁹ Naoyoshi's, Ogawa Hidai Daijiten 日臺大辭典 (Taipei, 1907)Google Scholar.

⁹⁰ Igari Kōnosuke 猪狩幸之助. Kanbunten 漢文典, (Tokyo, 1898).

⁹¹ Edkins, “Defence”, p. 731.

⁹² Ibid., p. 732.

⁹³ See Volpicelli, Old Chinese Phonology; Schaank, S. H., “Ancient Chinese Phonetics”, T'oung Pao (1897)Google Scholar. Regarding early Japanese treatises on early and medieval Chinese, see, e.g., the section In-kyō geshaku 韻鏡解釋 of Kōnosuke's 猪狩幸之助 Kanbunten 漢文典, pp. 102–104, or Naoyoshi's Hidai Daijiten 日臺大辭典. For a detailed description of Ogawa's reconstructed medieval Chinese system, consult U. J. Ang, “Xiǎochuān Shàngyì yǔ Gāo Běnhàn Hànyǔ Yǔyīn Yánjiū zhī Bǐjiào - Jiānlùn Xiǎochuān Shàngyì zài Hànyǔ Yánjiūshǐ Shàng Yīngyǒu dì Dìwèi 小川尚義與高本漢漢語語音研究之比較-兼論小川尚義在漢語研究史上應有的地位 [Comparison of Ogawa Naoyoshi and Bernhard Karlgren's Chinese Etudes sur la phonologie chinoise—On the Role of Ogawa Naoyoshi's in the History of Chinese Studies]”, Táiwānshǐ Yánjiū 臺灣史研究 [Taiwan History Studies] 1, 2 (1994), pp. 25–84.

⁹⁴ Grammatical studies in Mesopotamia started around 1600 bce. In Europe, systematic studies on the grammar of Greek were begun by Krátēs of Mallos and Dionýsios Thrâx, though they were nowhere near as refined as Pāṇini's remarkable grammar of Vedic Sanskrit, composed around the beginning of the third century bce, which can be regarded as the culmination of Indian grammatical tradition. The first grammatical treatise written by a Chinese scholar is the Mǎshì wéntōng 馬氏文通, by Mǎ Jiànzhōng 馬建忠 (1845–1900), published in 1898.

⁹⁵ The term shēngyùnxué 聲韻學 ‘historical phonology’, literally ‘the study of the initials (shēng) and finals (yùn)’, is preferred over other terminologies, such as yīnyùnxué 音韻學, shēngyīnxué 聲音學, or yùnxué 韻學. It seems that Chinese scholars before the Republican period (1912–49) did not distinguish clearly shēng from yīn, as both could be used to indicate ‘sound’; see, Hsin-hsiung Chen, Gǔyīn Yánjiū 古音研究 [Research on Old Sounds], (Taipei, 1999), p. 742; and Hsin-hsiung Chen, Shēngyùnxué 聲韻學 [Historical Phonology], (Taipei, 2005), p. 14. However, following the distinction made by Huáng Kǎn 黃侃 (1886–1935) in his Shēngyùn tōnglì 聲韻通例 (1920), yīn ‘sound’ is made up of shēng ‘initial’ and yùn ‘rime’. Thus, terms such as yīnyùnxué, which is the common term for indicating ‘historical Chinese phonology’ in China, appear to be semantically redundant.

⁹⁶ The study of gǔyīn ‘old sounds’ in China can be periodised in six phases: (i) the first period goes from Zhōu (1046–256 bce) to Qín dynasty (221–206), when rime dictionaries did not exist. Duàn Yùcái's Liù shū yīnyùn biǎo 六書音韻表 and Yán Kějūn's Shuōwén shēng lèi 說文聲類 are among the best sources of information about this period; (ii) a period known in Chinese sources as liǎng Hàn 兩漢 ‘two Hàn’ (206 bce–220 ce), in which many sound changes, especially those involving initials, occurred, and the xiéshēng 諧聲 principle, developed in the previous period, was no more productive; (iii) the third phase occurs during the Wèi Jìn nánběicháo 魏晋南北朝 period (220–589), and marks the beginning of the compilation of first rime books, though the two most important works of this epoch—the Shēnglèi 聲類 and the Yùnjí 韻集—have been lost. All the phonetic principles, adopted in the earlier epochs, to indicate the reading of initial consonants were completely useless in this period; (iv) the fourth period starts with the Suí 隋 dynasty (581–618) and ends with the fall of the Sòng 宋 (960–1279). It may well be considered as the aetas aura of rime dictionaries, for works such as Qièyùn 切韻, Tángyùn 唐韻, Guǎngyùn 廣韻, and Jíyùn 集韻, also known as sì shū 四書 ‘four books’, were compiled in this period. Technical notions, in addition to the already known fǎnqiè 反切 spellings, were added in these works, e.g. píngzè 平仄 (tone pattern), qīngzhuó 清濁 ‘clear muddy’ (voiceless : voiced), yīnyáng 陰陽 ‘dark bright’ (referred to the presence of [+voice] [-voice] features in the initial of words in the level tone), hóngxì 洪細 ‘wide acute’ (probably referred to the presence of certain glides in medial position, kāikǒu 開口 [no medial] and hékǒu 合口 [with medial -u-] were glossed as hóng, while qíchǐ 齊齒 [with medial -i-] and cuōkǒu 撮口 [finals beginning with -y-] were considered xì); (v) the fifth period starts with the Yuán dynasty (1271–1368) and ends with the fall of the Qīng empire (1636–1912). As in the first phase of this period, the dominant poetry form was the qǔ 曲, much attention was given to Northern sounds (běiyīn 北音). In the second part of this phase, much attention was given to the scholarship of Táng and Sòng dynasties; (vi) the last period goes from the Republic of China (1912–49) to our present day. For further information about this periodisation, see Chen Hsin-hsiung, Gǔyīn yánjiū 古音研究 [Research on Old Sounds] (Taipei, 1999), pp. 6–9; Shēngyùnxué 聲韻學 [Historical Phonology] (Taipei, 2005), pp. 487–489.

⁹⁷ Watters, T., Essays on the Chinese Language (Shanghai, 1889), p. 50Google Scholar.

⁹⁸ Sometimes local variations were motive of dispute between scholars. For example, the Qīng philologists Zhōu Áng 周昂 harshly criticised the Lèi yīn 類音, written by Pān Lěi 潘耒 (1646–1708), because in his opinion it was a mere description of the so-called nányīn 南音 ‘southern sounds’.

⁹⁹ According to the Guó Cháo hànxué shīchéngjì 國朝漢學師承記 (1, 4), many of Gù’s ideas on phonology were criticised by other scholars.

¹⁰⁰ See, e.g., Edkins, J., “Twan Yu Ts'ai's Fifteenth Class”, China Review 15, 5 (1887), pp. 312–314Google Scholar.

¹⁰¹ According to classical terminology, the head of the syllable is called shēng 聲, niǔ 紐, or shēngniǔ 聲紐. Regarding the study of ‘old initials’, the interested reader may find useful information in two early Japanese essays on the subject: Kōno Rokurō 河野六郎. “Shinkan shōkai Takahata Hikojirō-cho [Shū Hata Kan san-dai no ko himo kenkyū] 新刊紹介高畑彦次郎著 [周・秦・漢三代の古紐研究], [Newly published by Takahata Hikojirō, Research in the old initials in Zhou, Qin, Han generations]”, Gengo Kenkyū 言語研究 [Linguist Research] 2 (1939), pp. 77–86; and Takahata Hikojirō 高畑彦次郎. Gengo-gaku-jō yori mitaru Shū shinkan Mitsuyo no ko himo kenkyū 言語学上より見たる周秦漢三代の古紐研究 [A view on the old initials of Zhou, Qin, Han generations from the perspective linguistic science] (PhD diss., Tōkyō Teikoku Daigaku 東京帝国大学 [Tōkyō Imperial University], 1942).

¹⁰² In traditional Chinese phonology, sounds are divided into five major groups (wǔyīn 五音 ‘five sounds’), viz. chúnyīn 唇音 ‘labial sounds’, shéyīn 舌音 ‘lingual sounds’, yáyīn 牙音 ‘velar sounds’, chǐyīn 齒音 ‘dental sounds’, hóuyīn 喉音 ‘guttural sounds’ (bàn shéyīn 半舌音 ‘half lingual sounds’ and bàn chǐyīn 半齒音 ‘half dental sounds’ were added later). Shéyīn ‘lingual sounds’ are further divided into shétouyīn 舌頭音 and shéshàngyīn 舌上音, the former class is described as shéjiānyīn 舌尖音, wrongly considered, in this author's opinion, ‘apical sounds’ (they are not at all apical but laminal), the latter class of sounds is described as shémiàn qiányīn 舌面前音 or shéguānyīn 舌冠音 ‘coronal sounds’.

¹⁰³ In traditional Chinese phonology the term shuāngshēng 雙聲 (lit. ‘double sounds’) means identical sounds (xiāngtóngyīn 相同音).

¹⁰⁴ ‘Medieval Chinese’ reconstructions are taken from Zhèngzhāng 鄭張, Shànggǔyīnxì 上古音系 [Old Chinese Phonology] (Shanghai, 2003) and Lǐ Róng 李榮, Qièyùn yīnxì 切韵音系 [The sound system of Qièyùn] (Beijing, 1956).

¹⁰⁵ Chen, (Taipei, 1999), p. 545.

¹⁰⁶ Lin Yin 林尹, Xùngǔxué Gàiyào 訓詁學概要 [Outline of ancient exegesis (xùngǔ)] (Taipei, 2007), p. 93; H. H. Chen, Gǔyīn yánjiū 古音研究 [Research on Old Sounds] (Taipei, 1999), p. 560. In the Yùmǔ gǔdú kǎo 喻母古讀考 of Zēng Yùngān 曾運乾 (1884–1945), yù initials are further divided into yù sān 喻三 and yù sì 喻四. The former, in early Chinese, were not distinguishable from xiá initials 匣母 [*ɦ], the latter were not distinguishable instead from dìng initials 定母 [*d]. See Zēng Yùngān 曾運乾. Zēng Yùngān Yùmǔ Gǔdúkǎo dì Sì 曾運乾喻母古讀考第四 [Zēng Yùngān's Research on the Old Reading of Yù Initials], Gǔ Shēngyùn Tǎolùnjí (Yángshùdá Jílù) 古聲韻討論集 (楊樹達輯錄), (Taipei, 1965).