Using 反切 to determine readings of characters in modern varieties of Chinese

Question

From my understanding, 反切 is presented in the format XY 切, where a pronunciation of a character is determined from the inital of X and the rime of Y. However, there are still some points that are unclear to me I hope others can help clarify.

1. It seems that sometimes, the tone from Y does not seem to match with the end product. For instance, 東 has the 反切 of 德紅切, but based on the above principal, I would expect the reading to be dong2. However, the actual reading is dong1. Why does this happen? I'm assuming it has something to do with what the tones of these characters were in Middle Chinese, and that they shifted in modern varieties of Chinese. If that is the case, is there straightforward or general mapping of tones in Middle Chinese to tones in Modern Chinese?
2. There are also some instances where the initial of X and the rime of Y don't create valid syllables. 浹, 子協切, would produce some sound like zie2, even though that isn't a valid syllable in Mandarin. What happens in cases like these? Wiktionary also states that the expected sound is jie (no tone listed), when the character is pronounced jia1 (PRC) or jia2 (ROC). Is there an explanation for cases like these where the expected sound is different from the current pronunciation?
3. Do the answers from the above two points also apply to other varieties of Chinese?

Answer 1

1. The tones system is actually one of the easiest parts of mapping Middle Chinese to any modern Chinese variety. This information has been assembled on the Wikipedia entry for Four tones (Middle Chinese). For modern standard Mandarin, this is rather simple and pretty consistent, apart from 阴入 (unvoiced initial consonant with final 'stop' consonant) which is almost completely irregular.

2. It is possible to map all the regular sound changes from Middle Chinese to any (well documented) modern Chinese variety; see page 30 of this paper. In the case of 浹:

• 浹 as a 阴入 character (子協切 , 子 is unvoiced and so 阴, 協 belongs to 入) falls into "tone unpredictable", and you also expect more problems on the way.
• The 子 resulting in j- and not z- is actually predictable and mostly regular in Mandarin. You mention *zie, which of course is not valid a pinyin syllable in modern standard Mandarin, but actually was valid in the Mandarin of previous centuries, and retained in Peking opera stage pronunciation. This sound change is called the loss of the "round-sharp distinction" (尖团音) from Beijing Mandarin in about the late 19th century (judging from the Western dictionaries of the time), which merged the /ts/ sound of 子 into the /tɕ/ sound of 即 if there was an /i/ sound after it (after applying other changes first though...)!. In the parlance of the Qieyun 切韵 dictionary, we would say "because 協 is in division IV". From a linguistic perspective, it is just the "palatalisation of alveolar sibilants before /j/".
• The -ia in 浹 is irregular: we would expect -ie from the fanqie. This irregularity is common across all the characters with 夾 as a phonetic, so at least it was (semi-)consistent!

3. Certainly. The details will differ. It must also be said that the topolectal differences contribute a lot to our understanding of fanqie. Additionally, it has often been the major source of training speakers to speak in topolect, especially when it comes e.g. to personal names, to formal loanwords.

Answer 2

Sound shift, although an important factor, is not the primary reason that Fanqie doesn't fully work for determining the sound of characters in modern varieties of Chinese. In actuality, Fanqie never fully worked for any variety of Chinese, modern or ancient. This is because the phonology recorded in rime dictionaries wasn't for any particular language, but at the very outset was a compromise between Northern and Southern varieties - that is, Middle Chinese was itself artificial, never being a natural language.

　　昔開皇初，有儀同劉臻等八人，同詣法言門宿。夜永酒闌，論及音韻，以今聲調既自有別，諸家取捨亦復不同。吳楚則時傷輕淺，燕趙則多傷重濁；秦隴則去聲為入，梁益則平聲似去。又支、脂、魚、虞共為一韻；先、仙、尤、侯俱論是切。欲廣文路，自可清濁皆通，若賞知音，即須輕重有異。

　　呂靜《韻集》、夏侯該《韻略》、陽休之《韻略》、周思言《音韻》、李季節《音譜》、杜臺卿《韻略》等各有乖互。江東取韻，與河北復殊，因論南北是非，古今通塞，欲更捃選精切，除削疏緩；蕭顏多所決定。魏著作謂法言曰：「向來論難，疑處悉盡，何不隨口記之。我輩數人，定則定矣。」法言即燭下握筆，略記綱紀，博問英辯，殆得精華。

　　於是更涉餘學，兼從薄宦，十數年閒，不遑修集。今返初服，私訓弟子，凡有文藻，即須明聲韻。屏居山野，交遊阻絕，疑惑之所，質問無從。亡者則生死路殊，空懷可作之歎。存者則貴賤禮隔，以報絕交之旨。遂取諸家音韻、古今字書，以前所記者，定為《切韻》五卷。剖析豪氂，分別黍累，何煩泣玉，未得縣金，藏之名山。昔怪馬遷之言大，持以蓋醬，今歎楊雄之口吃。非是小子專輒，乃述群賢遺意，寧敢施行人世，直欲不出戶庭。于時歲次辛酉，大隋仁壽元年。

—— 陸法言《切韻・序》

---

Some time ago, during the first year of the 開皇 era, I was host to 劉臻 and seven other ministers (蕭該, 顏之推, 盧思道, 李若, 辛德源, 薛道衡, and 魏彥淵). After an evening of joyous drinking and feasting, the discussion turned to the subject of phonology.

The sounds of previous eras and the other geographical regions are different, and the subject had different schools of discipline: the 吳楚 (southern) topolects were light and gentle; the 燕趙 (northern) topolects were heavy and thick; the 秦隴 topolects (now around Shaanxi and Gansu) read the departing tone as the checked tone; the 梁益 topolects (now around Sichuan) read the level tone as the departing tone. Some schools treat「支」and「脂」,「魚」and「虞」with the same rhyme; some treat「先」and「仙」,「尤」and「侯」with the same initial. If one wished to indulge in a wide variety of literature, then one can personally interchange light and heavy sounds; however, if one wished to understand phonology, then one must clearly differentiate them.

The subject matter given in 呂靜《韻集》, 夏侯該《韻略》, 陽休之《韻略》, 周思言《音韻》, 李季節《音譜》, 杜臺卿《韻略》 and others were all different to the point of being contradictory. The phonology of 江東 (Jiangnan) is completely different to that of 河北 (northern regions). Thus, we discussed differences between the northern and southern topolects and between past and present, leading on to the choice of a precise and refined phonology standard, freed from clumsy and inelegant representations of sound, with many of the choices made by 蕭該 and 顏之推. 魏彥淵 later wrote to me: "That was a challenging discussion, but the most difficult of problems have been solved. Why don't you record the results of our conference? We'll treat the points that we have agreed upon as the established rules [of phonology]." Holding a brush under the candlelight, I proceeded to summarise the phonological rules decided, the studious questions asked, and their incisive responses of that evening, attempting to capture their quintessence.

Over the next ten to twenty years, the knowledge established that evening underwent ceaseless additions from knowledgeable scholars and officials. Today, I've regressed to my roots, and have taken on students; as long as they compose works, they must understand the rules of phonology. I'm living among the mountains; ceasing to travel on far journeys for academic pursuits, the questions that I harbour have no place to be discussed. As for the old friends and colleagues of mine: those who have long since passed away, I impossibly wish for their resurrection; and those who are still alive have gone down different paths in their lives, and I wish to sever ties with them. From the phonological studies taken from the different schools and literature on the subject matter from the past to present, the body of knowledge amassed is composed into the five volumes of Qieyun. Minute differences are stressed and fine details are emphasised, whilst not being obstinate and conceited [in its opinions]. It is truly an illustrious composition, worthy as a treasured legacy for many generations. I previously criticised Sima Qian for the veneration of his own work, but now I'm fond of teasing those who gaze on, open-mouthed in awe [at such grand creations]. However, I don't dare to presume the work's propensity towards widespread popularity, and only wish for it to be used within the households.

Preface to Qieyun, by 陸法言
辛酉 year, 仁壽 era of the Sui Dynasty

Answer 3

The English page on Wikipedia for Fanqie explains this phenomenon:

Effects of sound change

The method described the pronunciations of characters in Middle Chinese, but the relationships have been obscured as the language evolved into the modern varieties over the last millennium and a half. Middle Chinese had four tones, and initial plosives and affricates could be voiced, aspirated or voiceless unaspirated. Syllables with voiced initials tended to be pronounced with a lower pitch, and by the late Tang dynasty, each of the tones had split into two registers (traditionally known as yīn 陰 and yáng 陽) conditioned by the initials. Voicing then disappeared in all dialects except the Wu group, with consonants becoming aspirated or unaspirated depending on the tone. The tones then underwent further mergers in various varieties of Chinese. Thus, both the changes in both the initial and the tone were conditioned each other, as represented by different characters in the fanqie pair.

For example, the characters of formula 東 [tuŋ] = 德 [tək] + 紅 [ɣuŋ] are pronounced dōng, dé and hóng in modern Standard Chinese; thus, the tones no longer match. That is because the voiceless initial [t] and the voiced initial [ɣ] condition different registers of the Middle Chinese level tone, yielding the first and the second tones of the modern language. (The pinyin letter d represents the voiceless and unaspirated stop [t].)

That effect sometimes led to a form of spelling pronunciation. Chao Yuen Ren cited the example of the character 强, which had two readings in Middle Chinese. It could be read as [ɡjɑnɡ] in the level tone, meaning "strong, powerful", which developed regularly into the modern reading qiáng. However, it could be read also as [ɡjɑnɡ] in the rising tone, meaning "stubborn" or "forced". The regular development would be for the voiced initial [ɡ] to condition the yang register of the rising tone, becoming the fourth tone of modern Chinese and for the rising tone to condition an unaspirated initial. Thus, jiàng would be expected, and this does occur in the sense "stubborn", but the character also has the unexpected pronunciation qiǎng for the sense "forced". Chao attributed that to the fanqie formula 强 = 其 [ɡi] (level tone) + 兩 [ljɑnɡ] (rising tone) given in dictionaries. Here, the first character is now pronounced qí because in the level tone, the voiced initial becomes aspirated, but the second character is now pronounced liǎng. That is because in the rising tone, sonorants like [l] conditioned the yin register, which led to the modern third tone.

Answer 4

Not probably a complete answer, but a fascinating glimpse of such a method is given by this problem from the 2008 ILO:

https://www.ioling.org/booklets/iol-2008-team-prob.en.pdf (question);

https://www.ioling.org/booklets/iol-2008-team-sol.en.pdf (answer).

They explaing how it happens that 反切 frequently works for Cantonese – and what kinds of shifts make Mandarin barely feasible.