Introduction
Minimal pairs involving tonal alternations (‘Derivation by tone change’) of roots in one of the three tones of Middle Chinese (even tone pīngshēng 平聲 A(1), rising tone shǎngshēng 上聲 B(2), entering tone rùshēng 入聲 D(4)) with the qùshēng 去聲 (Tone C(3)) in Middle Chinese are quite common in Archaic Chinese. Haudricourt (1954) was the first who proposed that the emergence of the qùshēng was caused by the loss of a suffix *-s (Haudricourt, 1954; Sagart, 1999), which changed into *-h and further into the qùshēng. The process probably took place in different stages during the Late Archaic (5th – 3rd c. BCE) and Early Middle Chinese periods (2nd c. BCE – 3rd c. CE). The reconstructed suffix *-s is the most generally accepted derivational affix of Archaic Chinese and it has been connected to different morphological processes in the subsequent literature. Numerous functions have been attributed to the suffix (e.g. Downer, 1959; Haudricourt, 1954; Jin, 2006; Sagart, 1999; Sun, 2015; Unger, ms; Wang, 2014; Wuyun, 2024) with eight major categories proposed in Downer (1959, see also Jacques, 2016), which have been reduced subsequently to two major functions, (a) a valency decreasing function, connected to passive (Schuessler, 2007; Xing & Schuessler, 2020), perfective (Jin, 2006), and resultative (telicity) values (Meisterernst, 2023), and (b) a valency increasing (causative) function (e.g. Schuessler, 2007; Wang, 2014; Wuyun, 2024; Xing & Schuessler, 2020). The investigation of these functions in verbal derivations are at issue in the present project.
The oldest and one of the most important sources for a reliable investigation in the tones of Archaic Chinese is the Jīngdiǎn Shìwén 經典釋文 (JDSW) by Lù Démíng 陸德明 (550–630 CE), completed before 589, which consists of a collection of semantic and phonological glosses from the earlier commentary literature (Branner, 2003). Different from the traditional commentaries on the Classical Chinese literature in which the running text is interspersed with different kinds of glosses and commentaries, Lù Démíng did not attempt to subordinate his thousands of glosses to a running text but instead only commented on singular words and phrases he cited from the Classical Literature. The texts he commented on overlap with those collected in the Thirteen Confucian Classics (Shísān jīng 十三經), leaving out the text by Mencius, but including two Daoist Classics, the Dàodé jīng 道德經 and the Zhuāngzǐ 莊子. Whereas earlier commentaries and sources, such as the character dictionary Shuōwén jiězì 說文解字 (1st to 2nd c. CE), mostly used homophonic glosses or the dúruò glosses (using the schema X 讀若《丙》) “read [this character] as X”), Lù Démíng used the so-called fǎnqiè 反切 glossing system, which developed from at least the 3rd c. CE on, based on Hàn period and later glosses to the Classics. In the fǎnqiè system a character is divided into an initial and a final (rime) part (including tones) which are each represented by a single character, as illustrated in Table 1 (see also Branner, 2000 and List, 2018).
Table 1. The four different tones represented by fǎnqiè spellings.
| 堂 táng[th] | EMC | daŋ: | 徒郎 | d [d] [ɔ l] | aŋ (MM: t(ú l)áng) |
| 蕩 dàng[t] | EMC | daŋ‘: | 徒朗 | d [d] [ɔ l] | aŋ‘(MM: t(ú l(ǎng) |
| 宕 dàng[t] | EMC | daŋh: | 徒浪 | d [d] [ɔ l] | aŋh (MM: t(ú l)àng) |
| 鐸 duó [t] | EMC | dak: | 待各 | d [d] [əj‘k] | ak ‘(MM: d(ài g)è) |
Scholars such as Downer (1959), Huang (1992), Wang (2014), Sun (2015), (Wuyun, 2024) based their research on the JDSW for their investigations of different aspects of the morphological functions of the suffix *-s, but frequently studies on the morphology of Chinese use different, including younger sources such as the Sòng period dictionary Guǎngyùn 廣韻, for their research. Additionally, the methods employed in research are not always transparent and the situation in Archaic Chinese is such that the establishment of cognate sets is faced with numerous challenges, since problems ranging ‘from allographs (yìtǐzì 異體字), phonetic loans (tōngjiǎzì 通假字), to lexical borrowings’ have to be considered (Pulini & List, 2024a). One example illustrating this problem involves different graphic representations of the lexical root shì 視, OC *gijʔ ‘to see’ and its derivation, shì 示 OC *s-gijʔ-s ‘to show’ (Baxter & Sagart, 2014: 56, cf. Pulini & List, 2024a: 2024).
In the present project all qùshēng roots involving a verbal reading, and all roots which alternate with a (verbal) cognate (a derivation) in the qùshēng are at issue. We include also those cases where a deverbalized noun appears in the qùshēng. The objective is to contribute to the research on the basic function(s) of the suffix *-s based on the Archaic Chinese instances glossed in the JDSW and the impact their loss had on the diachronic development of the Chinese syntax from a more synthetic to a more analytic system (Huang & Roberts, 2017). The JDSW data is the starting point of our investigation, since it provides the earliest data reliably representing differences in pronunciations, in order to clarify the existence of a clear morphological and derivational relation between cognates (Fellner & Hill, 2019) as far as possible. In relying on this very early source, we try to consolidate the considerable disagreement in the reconstruction of affixes in Chinese. For instance, our data show that the reconstructions of the Archaic Chinese affixes in the phonological literature do not always correspond to the pronunciation glosses in the JDSW (see also Jacques, 2022). This may be due to the use of different sources by the respective scholars for their reconstructions (for an overview on sources used in the reconstruction of Archaic Chinese see Baxter and Sagart, 2014: 2f). The collection of data so far also shows considerable differences in the fǎnqiè annotations of the JDSW and of the Guǎngyùn respectively, suggesting that the system of sound glosses underwent significant changes towards a certain degree of standardization from the Sui to the Song period. The differences between the JDSW and the Guǎngyùn systems provide some evidence for a genuine early system of sound glosses representing morphological differentiations contrary to the assumptions of the Chinese Purist School of the Kǎozhèng 考證 philologists of the Manchu period. The latter were opposed to the tradition of the variant readings collected in ‘rime-books’ and standard commentaries from the Han period on. The ideal of the purist school, whose scholars only accepted very few of the proposed tonal alternations, was the conception of the Chinese writing system in which one character stood for only one reading (Branner, 2003).
In a first step, the data collection and the resulting database is intended to enhance the transparency of work with the sound glosses in the JDSW, which are presented with their Middle Chinese pronunciations. Similar to the annotation of ancient Chinese documents written on Bamboo Slips (Pulini & List, 2025) a list of fǎnqiè with Middle Chinese pronunciations linked to the character and the respective occurrences in the JDSW via an ID could serve as a basis for a more consistent system of reference in the research on the historical morphology of Chinese.
Materials and methods
Materials
Initial data collection was done in tabular form, created as an Excel spreadsheet. The starting point for the materials collected manually so far was the digital version of Unger’s (1989) Glossar des Klassischen Chinesisch, a wordlist with 3500 entries. From this a list of 956 words with a qùshēng reading have been extracted into an excel file, i.e. ca. 27% of the entire lexicon assembled in the Glossar. According to a first survey, 562 of these 956 roots (58,79%) involve a verbal reading. This list of verbal roots has been checked against Baxter & Sagart (2014), Pulleyblank’s Lexicon of Reconstructed Pronunciations (1991), the monographs of Jin (2006) and Sun (2007), and an article by Wuyun (2024). All words which have a verbal meaning as one of their lexical entries in the respective sources were added to the spreadsheet. In a later step the reconstructions of Old Chinese by Schuessler (2007) and the fǎnqiè from the Guǎngyùn with their MC reconstructions (taken from Pulini & List, 2025) were included. From the two monographs and the article by Wuyun (2024) examples were additionally incorporated in the spreadsheet. The resulting sheet ‘Verbs_complete’ contains 1468 entries and has the structure indicated in Table 2.
Table 2. Structure of sheet “Verbs Complete” in the initial spreadsheet.
| Column | Name | Comment |
|---|
| A | Character | Chinese character, sometimes including variant forms |
| B | Reconstruction Schuessler | Schuessler’s (2007) Old Chinese reconstructions |
| C | Middle Chinese | Created by edictor or manually |
| D | Guǎngyùn fanqie | Fǎnqiè as taken from the Guǎngyùn |
| E | JDSW fanqie | Created by edictor or manually |
| F | Pīnyīn | Mandarin Chinese transliteration |
| G | Reconstruction MC Unger | Middle Chinese reconstruction by Unger (1989) |
| H | Reconstruction EMC Pulleyblank | Pulleyblank (1991) |
| I | Baxter & Sagart MC | Middle Chinese, Baxter & Sagart, 2014 |
| J | Baxter & Sagart OC | Old Chinese, Baxter & Sagart, 2014 |
| K | Reconstruction Jin (2006) | Reconstruction of Old Chinese |
| L | English gloss | Pulleyblank (1991); translation of Unger (1989) |
| M | German gloss | Unger (1989) |
| N | Tone alternation yes/no | Indication of the presence of tone alternations |
| O | Baxter & Sagart gloss | Baxter & Sagart (2014) |
| P / Q | alternative glosses of non-tone C variant | Baxter & Sagart (2014); Pulleyblank (1991); Unger (1989) |
| R | Tentative syntactic definition | Definition as state Verb, telic Verb, resultative Verb, Noun, etc. based on lexicon entries |
| S | Ref. Jin (2006) | Ref. to the discussed lexical item |
| T | Example Jin (2006) | Representative example for the discussed lexical item |
| U / V / W | comments | On problematic cases etc. |
| X | Reconstruction EMC Pulleyblank | Reconstructions of Early Middle Chinese of the non-tone C word |
| Y | Sun (2007) lexical root | Based on the distinction in Sun (2007) into the original (root) and derived form of a word |
| Z | Sun (2007): derived form | Based on the distinction in Sun (2007) into the original and derived form of a word |
| AA / AB / AC / AD | Wuyun (2024) examples and comments | Examples from the Wuyun (2024) article on the proposed causative functions of the suffix *-s and brief comment on their discussion |
All entries marked for tonal alternation were extracted to a separate table, which subsequently was checked against the JDSW. In a first step we specifically searched for those entries which appear with one or more fǎnqiè involving tonal alternation in the JDSW, leaving out the original (non-derived) readings marked by rú zì 如字 (‘like the character’) and the yīn 音 X (‘sounds like X’) glosses. All fǎnqiè were listed, including those that were homophonous or near homophonous. The total number of tone alternation roots from the ‘Verbs_complete’ sheet checked against JDSW is 547, of those 46 (8,41%) do not appear with a FQ in JDSW and 33 (6,03%) are not attested at all. Of the remaining roots, 496 were included in the Qù-Tone Alternations database.
Apart from the manually compiled spreadsheet, digital versions of additional databases were used as provided by their publishers or as they could be retrieved by us. These include fǎnqiè spellings and automatically generated Middle Chinese readings from the Guǎngyùn resource provided by Pulini and List (2024b), Baxter and Sagart’s Old Chinese readings in digital form as provided by the authors, a digital version of Schuessler’s (2007) Old Chinese reconstruction as provided in digital form by List et al. (2017), and a digital version of the Jīngdiǎn shìwén (Lu 583[1985]).
Methods
Creation of the basic spreadsheet. The data for the ‘initial sheet’ file, the basis for the first version of the Qù alternations database containing fǎnqiè from the JDSW was assembled and commented on manually based on the digitized version of Unger’s (1989) Glossar. In a second step Pulleyblank (1991) and the appendix of Baxter and Sagart (2014) were checked for additional lexical items which display possible connections with a verbal root. Those roots were entered manually into the extended spreadsheet together with the respective English glosses provided in Pulleyblank (1991) and Baxter and Sagart (2014), and with the German glosses in Unger’s (1989) Glossar. On the basis of the dictionary entries a preliminary classification of the parts of speech was added to the table for easier analysis of the data. This allows the automatic exclusion of proper names and of nouns, and other lexical items which do not display any possible connections to verbal roots. The relevant data from Jin (2006), Sun (2007), and Wuyun (2024) was added manually with the help of a research assistant. In a third step, the Old Chinese readings of Schuessler (2007), and the fǎnqiè from the Guǎngyùn with their Middle Chinese readings were added automatically, using the method by Pulini and List (2024a).
From the initial sheet file all entries marked for tone alternations were extracted automatically into a separate spreadsheet on tonal alternations in the JDSW which includes the following columns:
Character; Pīnyīn; possible alternations of initials, vowels, medials; fǎnqiè of tones A, B, C, D respectively, followed each by the Middle Chinese reconstruction generated by the Pulini and List's method. If the Middle Chinese reading was not available or incomplete, an attempt to reconstruct the reading following Baxter and
Sagart (2014) was undertaken manually; these readings are stored in a separate table. The JDSW qù alternation table also contains comments in the JDSW regarding double readings, and occasionally comments on scribal errors and other problems from later sources. For each tone, lists of the examples from the Archaic literature commented on in the JDSW were collected and checked against the Academia Sinica Corpus (https://ancientchinese.ling.sinica.edu.tw/ASACC_index/), the CTEXT Corpus (https://ctext.org), and the Scripta Sinica database of Academia Sinica (https://hanchi.ihp.sinica.edu.tw/ihp/hanji.htm).
Integration of data in cross-linguistic data formats. In order to make for a more detailed integration of the original spreadsheet with additional digital resources, we decided to model the data in CLDF. CLDF is a standard for tabular data, recommended by the Cross-Linguistic Data Formats initiative (Forkel et al., 2018). While CLDF originally targets cross-linguistic datasets, the basic format can also be adapted for monolingual datasets and has been used to model Old Chinese texts written on bamboo slips with additional annotations (Pulini & List, 2025). The advantage of using CLDF for data representation are two-fold. On the one hand, CLDF allows for a facilitated integration of individual datasets from different sources, thanks to CLDFBench (https://pypi.org/project/cldfbench, Forkel & List, 2020), a Python package that facilitates the creation of CLDF datasets by allowing users to mix custom Python code with standardized predefined functions that work across multiple platforms. On the other hand, datasets, once provided in CLDF, can be easily published online with the help of the CLLD framework (https://clld.org, Forkel, 2014), a framework that facilitates the exploration of tabular data in websites, offering a common look-and-feel specifically to linguists, who know the framework well due to the success of large databases like the World Atlas of Language Structures (https://wals.info, Dryer & Haspelmath, 2013) or the Database of Cross-Linguistic Colexifications (https://clics.clld.org, Rzymski et al., 2020). Furthermore, an extra bonus of creating CLDF datasets, consists in extra tests on data integrity that will be automatically carried out when aggregating data from different sources.
In order to create the integrated Database of Qù-Tone Alternations, we first designed a data model that links the different types of data to each other, and then created a Python script that would allow us to carry out the aggregation automatically with the help of the CLDFBench package. The data model consists of two tables, one entry table and one table for text examples. The entry table provides information on all the characters that were considered in the original spreadsheet, with additional information on fǎnqiè spellings, Middle Chinese and Old Chinese readings, and further information. The example table lists all automatically inferred occurrences of the characters in question in the JDSW, with automatically inferred fǎnqiè spellings. Rows in the entry table are linked to one or more IDs in the example table, accounting for the fact that several characters occur with multiple glosses and fǎnqiè spellings in the JDSW.
Creation of the CLLD web application. Once a CLDF data set has been created, the deployment of the data in the form of a CLLD web application is straightforward. The CLLD application allows to investigate the data from three perspectives through three separate tables. The Roots table lists individual characters along with all attested readings, mimicking the original structure of the spreadsheet, but with entries linked to detailed views of the data. Roots — representing root morphemes in Chinese that can surface in different forms, forming individual words — are linked to their Occurrences, which are in turn disambiguated through different Middle Chinese readings derived from different fǎnqiè spellers. Occurrences themselves link to the JDSW Reference. Here, individual occurrences of identical fǎnqiè spellers can in turn recur several times in the JDSW. The application allows for a convenient browsing of the three tables, given that they are interlinked and can additionally be filtered by all their individual fields.
Dataset validation
Detailed data inspection
In order to illustrate the benefits of the data curation and data integration steps that we have followed to create the Database of Qù-Tone Alternations, it is useful to inspect the database itself. Figure 1 shows the Roots table, the first table offered by the web application (https://qualternations.digling.org/parameters). Here, each character is linked to one or more (1) readings in Pīnyīn transliteration, (2) tonal values (using numbers to denote the four tones in Middle Chinese and Old Chinese), (3) alternation types (listing Chinese names in Pīnyīn for the four tones), (4) Middle Chinese readings (computed using the approach by Pulini & List, 2024b), and (5) the number of occurrences with different fǎnqiè spellers. The Roots table is filtered by the column Readings, listing only entries that start with bj.
Figure 1. Screenshot of the Roots table in the database.
When clicking on any of the identifiers in the column ID of the roots table, this will provide a filtered view of the entire Occurrences table of the database, listing only the occurrences of the character with the corresponding ID, as shown in Figure 2. Here, different fǎnqiè spellers with potentially resulting different readings in Middle Chinese can be contrasted with each other for individual characters.
Figure 2. Occurrences table of the web application, listing entries corresponding to the character fàn 氾 with ID 196.
Clicking on the ID in the ID column a further time will show the occurrences of a character with its particular fǎnqiè in the JDSW. This table view, illustrated in Figure 3, allows to check the original entry or the original entries and comments provided by the resource.
Figure 3. Occurrence of the character fàn 氾 with fǎnqiè 孚劒 in the JDSW.
General statistics on the data
The web interface with its filtering facilities also allows for a convenient — manual — inspection of major statistics on qù-tone alternations assembled from the JDSW. While this could also be directly computed from the CLDF representation of the data, with the help of Python scripts, the filtering functionalities allow for a more flexible, direct interaction with the data than a script run on the terminal of a computer, given that the results can be immediately inspected.
In Table 3, we have listed the absolute frequencies along with their proportion with respect to the total number of roots for all observed alternation types. Qù-tone alternations constitute the majority of observed alternations here. While this is not surprising — given that the objective of the Database of Qù-Tone Alternations are alternations involving the falling tone — we can see that the proportions of alternations with respect to other tones does not show any particular trend in the data. This coincides well with the hypothesis that alternations involving the falling tone were derivational in nature, and would thus apply to any root, independent of its original phonology. What can also be seen from the general statistics is that — with a total of 292 cases involving qù tone alternations — there is substantially more evidence for this alternation type than was previously discussed in the literature. This warrants a closer investigation in future work.
Table 3. Alternation types by frequency, inferred from the data presented in the web application.
Alternations constituting more than 10% of the data are highlighted.
| Alternation Type | Frequency | Proportion |
|---|
| 1 | 31 | 6.25 |
| 1 2 | 9 | 0.94 |
| 1 3 | 116 | 12.18 |
| 1 4 | 1 | 0.12 |
| 1 2 3 | 25 | 2.99 |
| 1 2 4 | 1 | 0.12 |
| 1 3 4 | 5 | 0.62 |
| 1 2 3 4 | 4 | 0.5 |
| 2 | 23 | 2.88 |
| 2 3 | 79 | 10.17 |
| 2 4 | 1 | 0.14 |
| 2 3 4 | 3 | 0.43 |
| 3 | 122 | 17.58 |
| 3 4 | 60 | 10.49 |
| 4 | 16 | 3.13 |
| Total | 496 | 100 |
Conclusion and outlook
With this brief study we have presented the digital version of a number of fǎnqiè from the earliest collection of pronunciation glosses, the JDSW, as the basis for a project on the diachronic development of the Chinese language. The specific objective of this database is to facilitate the research on the morphological functions of the reconstructed suffix *-s of Archaic Chinese, which developed into the qùshēng in Middle Chinese. Tonal alternations in Middle Chinese resulting from earlier affixation have been comprehensively discussed in the literature since Haudricourt (1954). The suffix *-s is the most generally accepted affix of Archaic Chinese and its morphological and syntactic analysis together with cross-linguistic studies on the derivational morphology in the Tibeto-Burman languages have gained considerable interest in recent years (e.g. Jacques, 2016; Zhang, 2022). The Qù alternations database presented in our study so far facilitates the analysis of the statistical distribution of the tonal alternations collected in the JDSW in providing the attested FQs for each character together with their Middle Chinese pronunciations.
In a next step, the reconstructions of Old Chinese by Schuessler (2007) and Baxter and Sagart (2014) and the English glosses from the literature (Baxter & Sagart, 2014; Pulleyblank, 1991) collected in the Verbs_complete excel table will be added to the database. Additionally, the database will be expanded to not only including those words which display tonal alternations, but all qùshēng words from the Verbs_complete table which have a verbal as one of their possible readings (around 950 more entries). This will result in a digital dictionary of Archaic Chinese lexical roots in the qùshēng (original or derived) with verbal (but not necessarily only verbal) readings with their Middle Chinese pronunciations based on the JDSW and the Guǎngyùn and with the original fǎnqiè from the JDSW. A number of these verbs display more than one alternation, i.e., alternations of the initial, the medial, and/or the vowel together with the qùshēng. A digitized version of the data allows for easy comparison of different annotation systems in the sound glosses and for philological and linguistic research on the historical morpho-phonology of Chinese.
Ultimately, in a last step, results of the research on the morpho-syntactic functions of the *-s suffix (the qùshēng in Middle Chinese) and the diachronic morphosyntax of Chinese will be linked to the database. The most representative examples from the Archaic Chinese literature for the proposed functions of the *s-suffix, e.g., as a marker of telicity or resultativity, causativity, nominalization etc., will be linked to the respective verbs. The examples will be presented with transcriptions in Pīnyīn, with interlinear glosses following the Leipzig glossing rules, and with translations. The glosses for the verbs discussed will include the Middle Chinese pronunciations based on the earliest FQ from the JDSW and the reconstructions of their Old Chinese forms following Baxter and Sagart (2014). The Digital Edition of the Cáo Mò Zhī Zhèn 曹沫之陳 (https://cmzz.digling.org/) will serve as a model for the linking and the representation of the examples. In this way, the final database will allow to query different aspects of Archaic Chinese such as the lexicon, the phonology, morphology, and syntax with corpus linguistic tools. Chinese characters and their fǎnqiè are transcribed in Pīnyīn, with their Middle Chinese pronunciations and Old Chinese reconstructions, and glossed in English; the examples for the syntactic structure of the verbs discussed will include English and linguistic glosses and translations. This way the digitally computational data will be made accessible also for linguists not familiar with (Archaic) Chinese, who are interested in the research on cross-linguistic diachronic morpho-syntax. An open access digitized version of the data and the ensuing research results as has been proposed here will enhance the transparency of the research on the historical morphology of Chinese and make it available to a greater linguistic community.
Ethics and consent
Ethical approval and consent were not required.
Acknowledgements
We thank Robert Forkel for providing tips with CLDF data handling and CLLD deployment.
Faculty Opinions recommendedReferences
- Baxter WH, Sagart L:
Old Chinese. A new reconstruction. Oxford University Press: Oxford, 2014. Reference Source
- Branner DP:
The rime-table system of formal Chinese phonology. In: Sylvain Auroux, E. F. K. Koerner, Hans-Josef Niederehe, and Kees Versteegh (eds): History of the language sciences. De Gruyter: Berlin and New York, 2000; 46–55. Publisher Full Text
- Branner DP:
On early Chinese morphology and its intellectual history.
J R Asiat Soc.
2003; 13(1): 45–76. Publisher Full Text
- Downer GB:
Derivation by tone-change in classical Chinese.
Bulletin of the School of Oriental and African Studies.
1959; 22(1): 258–290. Publisher Full Text
- Dryer MS, Haspelmath M:
World atlas of language structures online. Max Planck Institute for Evolutionary Anthropology: Leipzig, 2013. Reference Source
- Fellner HA, Hill NW:
Word families, allofams, and the comparative method.
Cahiers de Linguistique Asie Orientale.
2019; 48(2): 91–125. Publisher Full Text
- Forkel R:
The cross-linguistic linked data project. In: Proceedings of the 3rd Workshop on Linked Data in Linguistics: Multilingual Knowledge Resources and Natural Language Processing. 2014; 61–66. Reference Source
- Forkel R, List JM, Greenhill S, et al.:
Cross-linguistic data formats, advancing data sharing and re-use in comparative linguistics.
Sci Data.
2018; 5: 180205. Publisher Full Text
- Forkel R, List JM:
CLDFbench. Give your cross-linguistic data a lift. Proceedings of the Twelfth International Conference on Language Resources and Evaluation. 2020; 6997–7004. Reference Source
- Haudricourt AG:
Comment reconstruire le chinois archaïque [reconstructing ancient Chinese].
Word.
1954; 10(2–3): 351–364. Publisher Full Text
- Huang CTJ, Roberts I:
Principles and parameters of universal grammar. In: Roberts, Ian (ed.). The Oxford Handbook of Universal Grammar. Oxford University Press: Oxford, 2017. Publisher Full Text
- Huang黃 K坤堯:
Jīngdiǎn shìwén dòngcí yìdú xīntán 經典釋文動詞異讀新談. [New survey on the variant pronunciation of verbs in the Jingdan Shiwen]. Xuesheng shuju: Taibei, 1992.
- Jacques G:
How many *-s-suffixes in Old Chinese?
Bull Chin Linguist.
2016; 9: 205–217. Publisher Full Text
- Jacques G:
On the nature of morphological alternations in Archaic Chinese and their relevance to morphosyntax.
Bulletin of the School of Asian and Oriental Sciences.
2022; 85(3): 475–494. Publisher Full Text
- Jin L:
Shànggǔ Hànyǔ xíngtài yánjiū 上古漢語形態研究[Morphology in Archaic Chinese]. Hefei, 2006.
- List JM:
More on network approaches in historical Chinese phonology (音韵学). In: The 2nd Li Fang-Kuei Society Young Scholars Symposium. 2018; 157–174. Reference Source
- List JM, Pathmanathan J, Hill N, et al.:
Vowel purity and rhyme evidence in old Chinese reconstruction.
Lingua Sinica.
2017; 3(1): 5. Publisher Full Text
- Lu D, 德明 陸:
Jīngdiǎn Shiìwén. 經典釋文[Explanation of the Classics]. Guji Chubanshe: Shanghai, (583[1985]). Reference Source
- Meisterernst B, List JM:
Qù-tone alternations in the history of Chinese. Dataset. Version 0.1. Zenodo. 2025. http://www.doi.org/10.5281/zenodo.17244146
- Meisterernst B:
The loss of morphology and the emergence of analytic structures in Chinese.
J Hist Syntax.
2023; 7: 1–55. Publisher Full Text
- Pulleyblank EG:
Lexicon of reconstructed pronunciation in Early Middle Chinese, Late middle Chinese, and Early Mandarin. UBC Press: Vancouver, 1991. Reference Source
- Pulini M, List JM:
First steps towards the integration of resources on historical glossing traditions in the history of Chinese. In: Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation. 2024a; 7343–7348. Reference Source
- Pulini M, List JM:
Finding language-internal cognates in old Chinese.
Bull Chinese Linguist.
2024b; 17(1): 53–72. Publisher Full Text
- Pulini M, List JM:
Using cross-linguistic data Formats to enhance the annotation of Ancient Chinese documents written on bamboo slips. In: Proceedings of the Second Workshop on Ancient Language Processing. 2025; 31–37. Reference Source
- Rzymski C, Tresoldi T, Greenhill SJ, et al.:
The database of cross-linguistic colexifications, reproducible analysis of cross-linguistic polysemies.
Sci Data.
2020; 7(1): 13. PubMed Abstract
| Publisher Full Text
| Free Full Text
- Sagart L:
The roots of old Chinese. John Benjamins: Amsterdam, 1999. Reference Source
- Schuessler A:
ABC etymological dictionary of old Chinese. University of Hawai'i Press: Honolulu, 2007. Reference Source
- Sun 孫 Y玉文:
Hànyǔ biàndiào gòucí yánjiū 漢語變調構詞研究 [Investigation of the word-formation by tone-change in Archaic Chinese]. Shangwu Yinshuguan: Shanghai, 2007.
- Sun 孫 Y玉文:
Hànyǔ biàndiào kǎobiàn 漢語變調考辨 [Examination and distinction of tone-change in Archaic Chinese]. Shangwu Yinshuguan: Shanghai, 2015.
- Unger U, H好 K古:
Sinologische Rundbriefe. [Unpublished Manuscript]. (no date).
- Unger U:
Glossar des Klassischen Chinesisch [Glossary of Classical Chinese]. Harrassowitz: Wiesbaden, 1989.
- Wang 王 Y月婷:
王月婷 Jīngdiǎn shìwén yìdú yīnyì guīlǜ yánjiū “經典釋文”異讀音義規律研究 [Investigation of the patterns of different pronunciations in the Jingdian Shiwen]. Academy of Social Sciences: Beijing, 2014.
- Wuyun S:
A Distributed Morphology account of tonal change in Middle Chinese.
J East Asian Linguist.
2024; 33: 489–531. Publisher Full Text
- Xing JZ, Schuessler A:
Semantic extension in old Chinese: direction, transitivity, and voice. In: Janet Z. Xing (ed.): A typological approach to grammaticalization and lexicalization. De Gruyter Mouton: Berlin and New York, 2020; 165–190.
- Zhang S:
Rethinking the *-s suffix in old Chinese: with new evidence from Situ Rgyalrong.
Folia Linguist.
2022; 43(1): 129–167. Publisher Full Text
