Musical Settings (III): Diction

This entry is part 25 of 35 in the series Chinese Art Song

For thousands of years, Classical Chinese was the unique written language for literary and documentary works alike. On the other hand, due to the vastness of the land and the geographic attributes—especially those of major rivers and mountains, there was no common spoken language in China. Under the umbrella of Sinitic languages, spoken by the Han 漢 people, there were numerous branches of dialects including Mandarin 官話, Jin 晉, Wu 吳, Xiang 湘, Min 閩, and Hakka 客家, to name a few. Each of them was further divided into a great number of regional tongues.[1] More than accentual variants, these dialects share few commonalities and were mostly unintelligible to outsiders.[2]

Zhao Yuanren recounted his linguistic experiences as a young person: His family communicated in northern dialects. However, the old masters who taught him poetry and Classical literature spoke southern dialects. For a long time, he thought that northern tongues were for daily conversations, and only the southern dialects were to be used in reciting literature.[3]

Efforts were made in early 1920s to unify the spoken languages.[4]  A standardized system based on the Peking (Beijing) dialect was established to be the “national language” 國語.[5] The pronunciation principle of this language would form the foundation for proper diction in Chinese art songs.

In the following discussions, transliterations in pinyin (in italics), Romanization and Zhuyin will be quoted in brackets. For convenience and clarity, instead of diacritical marks, numbers will be used for tone indications in pinyin. When IPA symbols were used for clarification, they will be marked with slashes. The tone numbers in IPA—55, 35, 214, 51— are based on the tone letter system devised by Zhao Yuanren. Further details can be found in later sections of this article

When learning texts in western languages, singers often rely heavily on the spelling combinations and the pronunciation principles of each language. Due to some unique linguistic practices, when studying Chinese texts, in addition to knowing the sounds of consonants and vowels and learning to differentiate tones, several lexical semantics must be taken into consideration.

–Literary vs colloquial

Certain Chinese words have literary and colloquial pronunciations. While the choice of pronunciation will not alter the meaning of the text, it highlights the style. The language of refined literature should not be equated to colloquial Pekingese. While modern pronunciation is suitable for vernacular poems, modifications should be applied to traditional poetry. For example:

還, when meaning “still,” is pronounced [hai2] in daily usage. For poetic reading, it is pronounced [huan2].[6] The well-known verse “乍暖還寒時候” in Li Qingzhao’s “shēng-shēng-màn” 聲聲慢 should be read as: [zha4 nuan3 huan2 han2 shi2 hou4]. The final words in Su Shì’s “The Great River Flows Eastwards” 大江東去should be “一樽還酹江月” [yi4 zun1 huan2 lei4 jiang1 yue4]. Zhao Yuanren suggested a third pronunciation [han2], derived from southern mandarin, for 還 in his songs as it would be less casual but not too traditional.[7]

The conversational pronunciation of the adverb 了 is /lə/[8] in neutral tone—light and staccato. In literature, it is often pronounced as [liao3] in the third tone. The possessive particle 的 is spoken as /də/ in neutral tone. In lyrics, especially when legato e espressivo, it will be read as [di]. The conversational pronunciation of the conjunction 和 (meaning “and”) is [han4]. The literary version is /hə35/ in the second tone.

__Heteronyms 破音字

There are endless heteronyms in Chinese language. It would be necessary to understand the usages of words in context. Following are a few frequently used characters and their various pronunciations and meanings:

[9]
When meaning “obtain,” “receive,” “suitable,” or “content,” /də35/ in the second tone, [de2]
“Must,” [dei3] in the third tone
Used as an adverb, it is pronounced as /də/ in neutral tone.
In traditional literature, it would be read as /də35/, [de2]. Li’s “shēng-shēng-màn” would end with 了得 [liao3 de2].

[10]
“Target,” or “aim,” is pronounced [di4].
“Certain,” or “truly,” it will be in the second tone as [di2].

[11]
“Peace,” “ease,” or “smooth,” is pronounced /hə35/, [he2]
“To respond,” “to be in concert with,” “to echo,” /hə51/, [he4]
“Mix,” or “combine,” [huo4][12]
“Warm,” [huo5] in neutral tone.

[13]
Noun, a two-hour division of night, marked by gongs of sentry,” /kɤŋ⁵⁵/, [geng1].[14] Colloquially, [jing1]. “挨不明更漏” in Cáo Xuěqín’s “Verses of Red Beans” 曹雪芹, 紅豆詞should be read as [ai2 bu4 ming2 geng1 lou4].
Verb, “change,” “alternate,” /kɤŋ⁵⁵/, [geng1]
“More,” “further,” /kɤŋ51/, [geng4]

[15]
Noun, “tune,” “melody,” or a poetic form, [qu3]
Adjective, “bend,” or “curvy,” [qu1]

Once the syntax was understood, phonetic characteristics also require careful handling and adjustments in singing.

__Phonetic symbols and transliterations

Phonetic symbols were devised to classify and transcribe sounds. As supplementary tools, they can be helpful to beginners. However, with great limitations, phonetic symbols are no substitute to linguistic study.

Most Classical musicians are familiar with International Phonetic Alphabets. In European languages, the symbols largely resemble the corresponding letters, thus not too difficult to apply. The same cannot be said for symbols used to mark Chinese sounds. In some cases, these symbols can be used for differentiation and clarification. Otherwise, they should be reserved for linguists and phonologists.[16]

There are several systems of Chinese Romanizations—using Latin alphabets to transliterate Chinese characters.[17] The earliest system was created by the Jesuits in the 16th century. The Wade-Giles system developed in the late 19th century was the first to be used universally.[18] The Yale romanization of Mandarin was devised for teaching Chinese to westerners, especially American soldiers in mid-20th century.[19] Although both the Wade-Giles and the Yale systems have been largely replaced by Hanyu pinyin, the former is still used in transliteration of proper names in Taiwan; the latter, in western textbooks.

Among transliteration systems created by native Chinese, three are most noteworthy: Zhùyin fúhàu 注音符號, Gwoyeu Romatzyh 國語羅馬字, and Hànyǔ Pinyin 漢語拼音.

In the last decades of the Qing Dynasty, as part of the initiative to promote literacy, linguists and educators tried their hands on building a transliteration system. On June 10, 1908, Zhang Taiyan presented two sets of symbols in his article “駁中國用萬國新語説” (“Refuting the Discourse of Using Esperanto in China”) in 《民報》Min Bao (People’s Newspaper)—a revolutionary paper published in Tokyo.[20] Inspired by the application of Katakana, a Japanese syllabary system derived from radicals of Chinese characters, Zhang patterned his phonetic symbols on ancient seal script 小篆.[21] The first set of symbols are thirty-six niǒu-wén 紐文, representing initials/consonants; the second, twenty-two yùn-wén 韵文, rhymes/final/vowels.[22]

In 1912, the temporary government in Beijing 北洋政府 established The Commission on the Unification of Pronunciation 讀音統一會 led by Wu Zhihui 吳稚暉.[23] Adopting fifteen symbols created by Zhang Taiyan and, with the same approach, adding new ones, the Commission proposed a set of thirty-nine symbols, Zhùyin zìmǔ (phonetic alphabets, 注音字母). After dropping a few symbols used only in dialects, the Ministry of Education of the Republic government formally published of thirty-six symbols in 1918. The system went through several modifications and was rename Zhùyin fúhàu (phonetic symbols) in 1930.[24]

These symbols are the most direct references to the phonology of Mandarin Chinese. On the other hand, their nonuniversal features make its application challenging for westerners. Replaced by Hànyǔ pinyin in China, these symbols are still in use in Taiwan and have been adopted for use in a few dialects.

Rather than a phonological system, Gwoyeu Romatzyh 國語羅馬字 (GR) was designed to be a writing system of Chinese words. Conceptualized by Lin Yutang 林語堂 and developed by Zhao Yuanren between 1926 and 1928, GR was characterized by its “tonal spelling”—an intricate and strict spelling system as indicators for the four tones in Mandarin.[25]

Zhao explained the system in the “Lyric Diction” section of his New Poetry Songbook to the readers/musicians and provided a list of spelling for all 480 words appeared in the collection.[26] In most cases, tonal indication letters were added to the basic spelling of the words. These letters were not sound components and, therefore, would not alter the pronunciations of the original consonants and vowels. The voiced consonant m, n, l, r in the first tone would be written as mh, nh, lh, rh. The second tone would often be indicated by an added “r” after the vowel. Double vowels such as aa-, ee-, ii-, uu, are indicators of the given vowels in the third tone. Double consonants, e.g., nn- and ll-, signaling the fourth tone.

In other cases, alternating letters would be used: Initial i-, u-, and iu- only occurs in the first tone. They would be changed to y-, w-, y(u) in the second and the fourth tones, and to e-, o- in the third. Certain combination of vowels would require further modifications.[27]

The other significant feature of GR is the combination of words into meaningful units without separating them with spaces. The initial verse of “The Great River Flows Eastwards,” 大江東去, 浪淘盡, would be presented as: [dahjiang dongchiuh lanqtaurjinn]

GR was published by the Republic government as the standard spelling system in 1928. Zhao used the system in his Mandarin Primer (1948) and Grammar of Spoken Chinese (1968). Lin Yutang used it in his Chinese-English Dictionary of Modern Usage (1972).[28] Yet, the cumbersome spelling rules made its application challenging to most users.

The most widely known and universally adopted Chinese spelling system is Hànyǔ pinyin (literally, Han language spelling/spelt sounds). Its roots can be traced back to Beifangxua Latinxua Sin Wenz 北方話拉丁化新文字 (Latinized new script of the northern language) developed by Soviet Scientific Research Institute on China between 1928 and 1931. The system was endorsed by many Chinese intellectuals including Lu Xun 魯迅, Cai Yuanpei 蔡元培, Sun Ke 孫科, and Guo Moruo.郭沫若 and was used in over three hundred publications.

In the early 1940s, Sin Wenz was popularized in the communist-controlled Shaan-Gan-Ning Border Region 陝甘寧邊區.[29] On December 25, 1940, the administration of the SGNBR decreed that Sin Wenz should have the same legal status as the traditional characters.[30] However, it quickly went out of fashion.

After taking control of China in 1949, the Chinese Communist Party embarked on language reform and established State Language Committee 國家語言文字工作委員會immediately. In addition to promote Pǔtōnghua (普通話 standard Mandarin) and simplifying the traditional characters, the Committee also took on the task of developing a Romanized writing system. A group of linguists, led by Zhou Youguang, constructed Hànyǔ pinyin. Its spelling system was based on GR and Sin Wenz. The tone diacritics were taken from Zhuyin. Pinyin was approved 1958 by People’s Congress and adopted by International Organization for Standardization as an international standard writing system in 1982.

Transliterations of Chinese characters, be it in GR, Sin Wenz or Pinyin, can be helpful tools for non-native speakers, especially English or European language speakers. Nevertheless, when the letters are used to represent sounds that do not exist in western language, they can easily lead to mispronunciations. Similar to IPA, they should be used as supplementary tools and with great care.

__Syllabic consonants

A syllabic consonant is a consonant that forms an individual syllable, such as the “l” in “bottle,” the “m” in “rhythm,” and the “ng” in “morning.” There are seven syllabic consonants in standard Chinese. They are: zi-, ci-, si-, zhi-, chi-, shi-, ri in pinyin; and/orㄗㄘㄙㄓㄔㄕㄖin Zhuyin symbols.[31]; All of them are apical consonants, produced by sending air through a narrow passage between tongue, alveolar ridge, and teeth.

[Zi-/ㄗ] is similar to “-ds” in “beds;” [ci-/ㄘ], “-ts” in “cats;” and [si-/ㄙ], “s” in “Sam.” To pronounce these sounds perfectly, the tongue position must be flat and with the tip against the lower teeth.

The sound of [Zhi-/ㄓ] is closed to “-ge” in “judge;” [chi-/ㄔ], the initial sound of “church;” and [shi-/ㄕ], “shirt.” To distinguish these sounds from the western ones, the tip of the tongue should be rolled up, touching the hard palate right behind the teeth. This movement should be gentle and will carry the sides of tongue up, shaping the tongue like a cupped hand.

Ri/ ㄖsounds like the “r” in “round.” Nonetheless, the lips do not need to be rounded when pronouncing ri/ㄖ.

When combined with various vowels, these consonants are marked as /t͡s/, /t͡sʰ/, /s/, /ʈ͡ʂ/, /ʈ͡ʂʰ/, /ʂ/ and /ɻ̩/ in IPA. When standing alone, after the initial sound, they become voiced—/t͡sz̩/, /t͡sʰz̩/, /sz̩/, /ʈ͡ʂz̩/, /ʈ͡ʂʰz̩/, /ʂz̩/ and /ɻ̩-z̩/–and ending with a shallow vowel sound, which is unmarked in Zhuyin, marked as “i” in pinyin, “y” in GR, and /ɨ/ in IPA.

In singing, keeping the voiced /z̩/ vibrate consistently can help carrying the sound. Zhao Yuanren had suggested to open the shallow vowel to a schwa, or to an “i” sound for 日. These adjustments need to be done with gentle touch so not to lose the natural sounds of the words.

Since Chinese is a syllabic language, each of these sounds can represent many different words/characters. Some of them are extremely common:

子, [zi3, ㄗˇ], “son,” “child”
詞, [ci2, ㄘˊ], “word combination,” “lyrics,” “ci poetry”
思, [si1, ㄙˉ], “to think,” “to miss”
絲, [si1, ㄙˉ], “silk,” “thread”
知, [zhi1, ㄓˉ], “to know,” “knowledge”
只, [zhi3, ㄓˇ], “only”
吃, [chi1, ㄔˉ], “to eat”
師, [shi1, ㄕˉ], “teacher,” “master,” “to learn”
失, [shi1, ㄕˉ], “to lose,” “to fail,” “lost,” “failure”
十, [shi2, ㄕˊ], “ten”
石, [shi2, ㄕˊ], “stone,” “rock”
時, [shi2, ㄕˊ], “time,” “hour,” “season”
日, [ri4, ㄖˋ], “sun,” “day”

__Voiceless consonants

With the exceptions of /m/, /n/, /l/ and /r/, all Chinese consonances are voiceless.[32] In European languages letters b-/p-, d-/t-, and g-/k typically represent pairs of voiced/voiceless sounds. In GR and pinyin, they are used to differentiate unaspirated and aspirated sounds.
Letter [b-] in pinyin is marked /p/ in IPA—voiceless bilabial plosive, and ㄅin Zhuyin. It sounds like the “p” in “speak.”
[d-,] /t/—voiceless denti-alveolar plosives, and ㄉ, as “t” in “study.”
[g-,] /k/—voiceless velar plosive, and ㄐ, as “ch” in “school.”

__Initials and finals

In ancient Chinese lexicon, dúruò (讀若, sounds-as/reads-as [another word]) was used to indicate the pronunciation of words. By the third century, a new method fǎnqiē 反切 was developed to replace the direct-comparison method. It became the standard method used in Middle Chinese rhyme dictionaries.[33]

In fǎnqiē, the sound of each character would derive from two other characters: The initial/onset of the first character, and the final as well as the tone of the second. This dichotomic division (without considering the tone) is still adopted in the phonological study of modern Chinese. Nevertheless, each Chinese syllable can be further divided into smaller phonemes: C (consonant)-G (glide)-V (vowel)-X (coda).[34] The vowel, often considered the nucleus of the sound, and the coda are associated with the rime. Not all the units need to be present in the sound construction.

In singing, considerations should be given to all the sound units. The initial consonant (when present) must be clearly pronounced without interrupting the melodic line. Durations of glide, vowel, and coda affect not only the intelligibility but also the expressiveness of the lyrics.

__Glides, diphthongs and triphthongs

There are three glides in standard Chinese: /j/ ([y-] in pinyin and [ㄧ] in Zhuyin; sounding as “y” in “yellow”), /ɥ/ ([yu-, ㄩ], “u” in French “suis”), and /w/ ([w-, ㄨ], “w” in “word”).[35] They can be in the initial position when an initial consonant is not present; the medial position between the initial consonant and the main vowel, and the final positions.

In traditional Chinese singing, especially in Kunqǔ theater, the sound of each word is divided into head, belly, and tail.[36] After enunciating the head sound clearly, the voice should move smoothly, yet with fluctuations, through the belly and gradually round up the tail. The prenuclear glides, handled as the belly, are often elongated or sung with melodic embellishments. This is in total contrast of the western practice of deemphasizing the glides. While the traditional method can cause difficulties in understanding the texts, the western approach overstresses the main vowel and misses the characteristics of the language. Zhao Yuanren suggested a middle-of the road handling, allowing sufficient time and sound to establish the glide before moving on to the vowel.[37]

Common [GV] words include:

夜 [ye4, ㄧㄝˋ], “night”
亞 [ya3/ya4, ㄧㄚˇ / ㄧㄚˋ], “Asia”
血 [xue3, ㄒㄩㄝˇ], literary pronunciations of “blood.”[38]
月 [yue4,ㄩㄝˋ], “moon”
雪 [xue3, ㄒㄩㄝˇ], “snow”
花 [hua1, ㄏㄨㄚ], “flower(s)”
我 [wo3, ㄨㄛˇ], “Personal pronoun, I”

Similar considerations must also be given to closing diphthongs [VG] [-ai, -ei, -ao, -ou, ㄞ ㄟ ㄠ ㄡ] and triphthongs [GVG] [-uai, -uei, -iao, -iou, ㄨㄞ ㄨㄟ ㄧㄠ ㄧㄡ]. Frequently used words in these groups include:

[VG]

愛 [ai4, ㄞˋ], “love”
淚 [lei4, ㄌㄟˋ] : “tear(s)”
道 [dao4, ㄉㄠˋ], verb: “to say,” “to tell;” noun: “way,” “path”
偶 [ou3, ㄡˇ]: “couple/pair,” “even numbers,” or “by chance,” “accidentally”
豆 [dou4, ㄉㄡˋ]: “bean(s)”
頭 [tou2, ㄊㄡˊ]: “head”

[GVG]

外 [wai4, ㄨㄞˋ]: “outside,” “foreign”
微 [wei2, ㄨㄟˊ]: “gentle”
有[iou3, ㄧㄡˇ]: “to have”
友 [iou3, ㄧㄡˇ]: “friend(s)”
遙 [iao2, ㄧㄠˊ]: “distant,” “far”
教 [jiao4,ㄐㄧㄠˋ]: “to cause,” “to tell;” [jiao1,ㄐㄧㄠ]: “to teach”

__/n/ and /ŋ/ finals

The two nasal consonants /n/ and /ŋ/, marked as “N” in phonological studies, should always be pitched in singing. While /n/ can occur in the initial and final positions, in standard phonology the /ŋ/ only occurs in the ending position.

They can be free-standing, proceeded by neutral/shallow vowels. The only character with the pronunciation /ˀɤŋ55/ [eng1, ㄥ]— 鞥, meaning “horse rein,”—is a rare word. On the contrary, words with /ən55/ [en1, ㄣ] pronunciation, such as 恩 (“favor,” “grace,” “kindness,” etc.) or 嗯 (an interjectional utterance), frequently appear in lyrics.

When the nasal consonants take place immediately after an initial consonant, e. g. 門 [men2, ㄇㄣˊ], “door;” 崩 [beng1, ㄅㄥ], “burst” or “eruption,” or 風 [fen1, ㄈㄥ], “wind,” the same shallow vowels, marked “e” in pinyin and unmarked in Zhuyin, serve as linkages. The final/nasal consonants become the hierarchical phoneme.

The same hierarchy will be given to /n/ and /ŋ/ in [VN] or [CVN] words with /i, y, u/ vowels, such as:

因 [in1, ㄧㄣ]: “cause,” “reason”
應 [ing1, ㄧㄥ]: “should;” [ing4, ㄧㄥˋ]: “to respond”
雲 [yun2, ㄩㄣˊ]: “cloud(s)”
尋 [xun2, ㄒㄩㄣˊ]: “search”
文 /wən³⁵/ , [wen2, ㄨㄣˊ]: “script,” “language,” “literature”[39]
東 /tʊŋ55/, [dong1, ㄉㄨㄥ]: “east”[40]
桐 /tʰʊŋ35/, [tong2, ㄊㄨㄥˊ]: “paulownia”

On the other hand, in words with /-ɑn/ or /-ɑŋ/ finals, the main vowel /ɑ/ shares the stresses and the duration with the nasal consonants:

安 [an1, ㄢ]: “safe,” “peaceful”
晚 [wan3, ㄨㄢˇ]: “evening,” “late”
談 [tan2, ㄊㄢˊ]: “to discuss,” “to talk,” “to converse”
難 [nan2, ㄋㄢˊ]: “difficult,” [nan4, ㄋㄢˋ): “disaster”
山 [shan1, ㄕㄢ]: “mountain(s)”
千 [qian1,ㄑㄧㄢ]: “thousand(s)”
點 [dian3, ㄉㄧㄢˇ]: “point(s),” “dot(s),” “drops”
洋 [yang2, ㄧㄤˊ]: “ocean”
想 [xiang3,ㄒㄧㄤˇ]: “to think,” “to miss [someone/something]”
江 [jiang1,ㄐㄧㄤ]: “river”
黃 [huang2, ㄏㄨㄤˊ]: “yellow”
窗 [chuang1,ㄔㄨㄤ]: “window(s)”

Zhao Yuanren quoted the phrase “月光戀愛著海洋” from 教我如何不想她 “How Can I Help [but] Thinking of You” to demonstrate proper handling of all the phonemes in each syllable/word.[41] In the following illustrations, the first line of transliteration is the original GR notation; the second one, pinyin. The dotted lines underneath the transliterations indicate the portion of the notes during which the words were not identifiable.

1. Following the western approach, quickly moving from glides to the main vowels and delaying the arrival of the final consonants.

2. Using the traditional practice, lingering on the glides and ending quickly with the vowel/finals.

3. A third method: evening out the duration of glides and vowel and adjusting the vowel sounds leading into the finals.

Judging by the lengths of the solid lines—the durations when the words are intelligible, the last execution would be the most appropriate one. The key fact is the clarification of the text—in context.

__Tones and tone sandhi

For non-native speakers, differentiating and pronouncing the Chinese tones are the most challenging part in learning the language. Even with the words set on fixed musical pitches, tonal inflections are still crucial in singing Chinese lyrics. From time to time, an unwritten slight might help to clarify the texts; from time to time, by lightening the sound, the interpreter can totally change the expression.

A “tone-letter” system devised by Zhao Yuanren and adopted by IPA can be helpful for musicians. As Zhao explained:

The total [speech] range is divided into four equal parts, thus making five points, numbered 1, 2, 3, 4, 5, corresponding to the low, half-low, medium, half-high, high, respectively. . . .

As the intervals of speech-tones are only relative intervals, the range 1–5 is taken to represent only ordinary range of speech intonation, to include cases of moderate variation of logical expression, but not to include cases of extreme emotional expression. For purposes of tone drills, each step may be taken to be a whole tone, thus making the total range equal to an augmented fifth. This would make the successive pronunciation of a number of tones sound rather unmusical, which however is rather an advantage for phonetic purposes.[42]

In standard IPA, instead of tone letters, the tonemes are represented by numbers as shown in the following diagram. The first number for each tone indicates the onset pitch; the last number, the ending. For the third tone, the pitch quickly dips from 2 to 1 before rising to 4. Thus, 媽[ma1, ㄇㄚ], “mother,” will be marked as /ma55/; 麻[ma2, ㄇㄚˊ], “hemp” or “numb,” as /ma35/; 馬 [ma3, ㄇㄚˇ], “horse,” /ma²¹⁴/; and 罵 [ma4, ㄇㄚˋ], “to scold,” “to curse,”, /ma51/. The falling character of the fourth tone is comparable to grave accents in Italian: e. g. the Italian word (“was”) is a homonym of 父 [fu4, ㄈㄨˋ], “father.” The diacritical signs used in Zhuyin and pinyin, ˉ, ˊ, ˇ, ˋ, are derived from these graphic lines.

*The five lines here represent five whole-tone steps, not intervals of thirds as in a regular musical staff.

The tone of an individual word is often modified when combined with other word(s). Phonologically, this phenomenon is called “tone sandhi.” Unmarked in neither pinyin nor Zhuyin, tone sandhi must be learned through practical applications. It occurs mostly with the third-tone words.

Full third-tone pronunciations only happen when speaking individual words. When combined with first-tone, second tone and fourth tone words, the rising segment of the tone is often omitted, creating a “half-third tone.”

When two third-tone words are linked together, the first one will be pronounced in the second tone:

你好, “How are you” [ni3-hao3, ㄋㄧˇ ㄏㄠˇ], will be pronounced [ni2-hao3, ㄋㄧˊ ㄏㄠˇ]. 手指, “finger(s)” [shou3-zhi3, ㄕㄡˇ ㄓˇ], becomes [shou2zhi3, ㄕㄡˊ ㄓˇ]

In phrases with three consecutive third-tone words, the word grouping will determine the modification procedure:

處理好 [chu3-li3-hao3, ㄔㄨˇ ㄌㄧˇㄏㄠˇ] is a three-word phrase meaning “manage well.” The first two words form a verb-phrase, 處理 “manage,” modified by the adverb 好 “well.” This results in the first two words changing into the second tone. The natural (native) pronunciation of the phrase thus becomes [chuli2 hao3, ㄔㄨˊ ㄌㄧˊ ㄏㄠˇ].

In a phrase like 小老鼠 [xiao3-lao3-shu3, ㄒㄧㄠˇ ㄌㄠˇ ㄕㄨˇ], the first character, meaning “little,” is modifying the two-word noun “mouse.” Thus, the middle character will be pronounced in the second tone. The phrase will sound like [xiao3 lao2shu3, ㄒㄧㄠˇ ㄌㄠˊ ㄕㄨˇ].

Although there are endless phrasal combinations, the contents and the stresses are always determining factors of tone modification, which take place naturally to help with the flow of the language.

Two extremely common words 不 [bu4, ㄅㄨˋ], “no,” and 一 [yi1, ㄧ], “one,” follow special tone-change procedures:

不 changes to the second tone when followed by another fourth-tone word:
不見 [bu2jiang4, ㄅㄨˊ ㄐㄧㄢˋ]: “not seeing,” “disappear”
不負 [bu2fu4, ㄅㄨˊ ㄈㄨˋ]: “not to betray”

一 as a number or as the final character of a phrase will be pronounced in the first tone:
一零一 [yi1-ling2-yi1, ㄧ ㄌㄧㄥˊ ㄧ], “101”
唯一 [wei2-yi1, ㄨㄟˊ ㄧ], “only,” “unique”

一 is pronounced in the fourth tone after words in the first-, second- and third tones:
一天 [yi4tian1, ㄧˋ ㄊㄧㄢ]: “one day”
一直 [yi4zhi2, ㄧˋㄓˊ]: “straight ahead,” or “always,” “continuous”
一點 [yi4dian3, ㄧˋㄉㄧㄢˇ]: “a little”

It changes into the second tone when proceeding a fourth-tone word:
一定 [yi2ding4, ㄧˊ ㄉㄧㄥˋ]: “for sure,” “definitely”

In spoken Chinese, there is a neutral/fifth, and often neglected, tone. Its Chinese name 輕聲 [qing1sheng1, ㄑㄧㄥ ㄕㄥ], meaning “light sound,” reflects its phonological character perfectly. Unmarked in pinyin, it is marked with a staccato sign in Zhuyin “.”[43] With no fixed pitch level, neutral sounds take place at the end of phrases.

It is used in the character 們 [men, ˙ㄇㄣ], a suffix indicating plural form; 著 [zhe, ˙ㄓㄜ], when used as an aspect particle indicating the continuity of an action; 的 [de, ˙ㄉㄜ], possessive particle; and 得. [de, ˙ㄉㄜ], adverb.

Neutral tone is used in the second character of duplicated words, such as 謝謝 [xie4xie, ㄒㄧㄝˋ ˙ㄒㄧㄝ], “thank you”, and kinship terms, such as 爸爸 [ba4ba, ㄅㄚˋ ˙ㄅㄚ], “daddy,” 媽媽 [ma1ma, ㄇㄚ ˙ㄇㄚ], “mommy,” 爺爺 [ye2ye, ㄧㄝˊ ˙ㄧㄝ], “grandpa,” 奶奶 [nai3nai, ㄋㄞˇ ˙ㄋㄞ], “grandma,” etc.

In conversations, 不and 一, when occurring in the middle of three-character phrases, are often lightened, e. g. 是不是 [shi4bushi4, ㄕˋ ˙ㄅㄨ ㄕˋ], “Isn’t it?”, 想一想 [xiang3yixiang3, ㄒㄧㄤˇ ˙ㄧ ㄒㄧㄤˇ], “think about it.”

While the application of the neutral tone is not as crucial in phonological procedure, the appreciation of such words/phrases can have greatly enhance musical interpretation.

__Érhuà

In northern dialects, especially colloquial Beijing dialect, a vowelized “r” /ɚ/ is often attached to the end of words as a diminutive suffix. This procedure is name by its sound as 兒化 /ˀɤɻ³⁵ xwä⁵¹/ [er2hua4, ㄦˊㄏㄨㄚˋ]. The rhotacism can result in: 1) elimination or nasalization of final consonants, or 2) modification of nuclear vowel sounds. Similar sounds can be found in American English such as “-er” in “butter,” “-ir-” in “shirt,” and “-ear” in “tear.

Commonly used in daily conversations, erhua also appears in literary works as indications of intimacy. Liu Bannong 劉半農 included two erhua phrases in “How Can I Help [but] Thinking of You:”

. . . . 水底魚兒慢慢游。

[Little] fish swim leisurely down below.

. . . . 啊! 西天還有些兒殘霞,

Ah! [Little bit of] twilight glows are still lingering on the western sky.

Since these “兒” characters are part of the verse structure, they are most often set to fixed pitches and note values in musical settings. It will be appropriate to release the sound gently to reflect the phonological character as well as the expressions.

So, how should Li Qingzhoa’s “守著窗兒,獨自怎生得黑!” be handled?


[1] In addition to the Han languages, there were hundreds of minority languages.
[2] Sinitic_languages_Wiki, Varieties_of_Chinese_Wiki
[3] Chao Yuen Ren, Xin shi ge ji新詩歌集, revised edition, (Taipei, Taiwan shang wu yin shu gua 臺灣商務印書館, 1960), 9.
[4] The “Preparatory Commission for the Unification of the National Language” was created in 1919. It was restructured and renamed as the Preparatory Committee in 1928.
National_Languages_Committee_Wiki
[5]Early in the development of language unification, a system combining both northern and southern mandarins (Wu) was propositioned. This “old national pronunciation” was closer to Middle Chinese, influenced by the rhyme dictionary Zhongyuan Yinyun 中原音韻. It retained the checked/entering tone.
Old_National_Pronunciation_Wiki
A pronunciation dictionary Guoyu Cidian國音字典, using Zhuying transliterations 注音符號, was published in 1920.
[6] The character, when used as a verb, meaning “return,” is also pronounced [huan2].
[7] Xin shi ge ji, 54.
[8] In pinyin or Romanization, the pronunciation is marked as [le]. The schwa IPA is used here for clarification.
[9] 得_en.wiktionary.org
[10] 的_en.wiktionary.org
[11] 和_en.wiktionary.org
[12] Colloquially, it could be pronounced in the second tone, [huó2].
[13] 更_en.wiktionary.org
[14]Although the pinyin for 更 is spelled “geng,” the IPA symbol /k/ reflects the voiceless consonant more appropriately. The vowel /ɤ/ is further back from /ə/. See further explanations below.
[15] 曲_en.wiktionary.org. The sound of [u] in pinyin is similar to “ü” in German.
[16] Standard_Chinese_phonology_Wiki, Help:IPA/Mandarin_Wiki
[17] Romanization_of_Chinese_Wiki
[18] Wade-Giles_Wiki
[19] Yale_romanization_of_Mandarin_Wiki. The Yale system is widely used in teaching Cantonese. It is also the standard romanization of Korean.
[20] 駁中國用萬國新語説_zh.wikisource.org
[21] Katakana_Wiki, Seal_script_Wiki
[22] Phonetic_symbols_by_Zhang_Taiyan_commons.wikimedia.org
[23] Aka Wu Jingheng 吳敬恆.
[24] 注音符號_Wiki, Bopomofo_Wiki
[25] Gwoyeu_Romatzyh_Wiki, 國語羅馬字_zh.m.wikipedia.org, Spelling_in_Gwoyeu_Romatzyh_Wiki
[26] Xin shi ge ji, 56-59.
[27] Further modifications will apply to certain letter/vowel combinations.
[28] Mandarin Primer and Grammar of Spoken Chinese
[29] Shaan-Gan-Ning_Border_Region_Wiki
[30] John DeFrancis, The Chinese Language: Fact and Fantasy, paperback. (Honolulu, University of Hawaii Press, 1986), 254.
[31] The basic spelling of these consonants in GR are: [jy-, chy-, shy-,ry-, tzy, tsy, sy.] In GR, the [j-, ch-, and sh- ] initials, when combined with “I,” become the equivalences of [ji-, qi-, xi-] in pinyin, and ㄐ ㄑ ㄒin Zhuyin.
[32] One of the major phonological changes between the Late Middle and modern Chinese was the devoicing of consonances. goldfishodyssey.com/2021/10/24/chinese-poetry-xiii-turning-point/
[33] Fanqie_Wiki
[34] The X (code) can be a non-vocalic consonant or a glide.
[35] The Zhuyin symbols are used for the allophonic vowels: /i/, /y/, and /u/. The /ɥ/ glide can be considered as a lighter and shorter /y/, the German “ü.”
[36] 字頭, 字腹, and 字尾.
[37] Xin shi ge ji, 55.
[38] The colloquial/spoken pronunciation of 血 is [xie3, ㄒㄧㄝˇ]
[39] In ㄨㄣ(/u/+/n/) combination, as the tongue moves slightly inward, the glide turns into a schwa, which is spelt as an “e” in pinyin and unmarked in Zhuyin.
[40] The phonological procedure of theㄨㄥ (/u/+/ŋ/) combinations begins with the /w/ glide, which opens to an /ʊ/ vowel before reaching the final. Thus, it is spelt with as [-ong] in pinyin
[41] The word 光 was original set to e-g-sharp. Zhao omitted the second note for the convenience of discussion. In the GR transliteration, the “h” in [yueh] and the second “n” in [liann] are tone indicators—fourth tone here. [Ay] was the fourth tone modified spelling of the basic [ai] sound; [hae,] the third of [hai.]
[42] Tone_letter_Wiki
Y. R. Chao, “ə sistim əv ‘toun-letəz’” [“A System of ‘Tone-Letters”], Le Maître Phonétique Vol. 8 (45), No. 30 (avril-juin, 1930), 25. https://www.jstor.org/stable/44704341
[43] The staccato sign appears in front of the symbols in Zhuyin.

Musical Settings (II): Words, Tones and Music

This entry is part 24 of 35 in the series Chinese Art Song

Although Zhao Yuanren (1892-1982) received formal training in mathematics and physics at Cornell, and studied philosophy at Harvard, it was his innate linguistic ability that led to his unique and crucial role in the literary reform movement.[1] In 1921, he recorded the phonographic “Standard Chinese” as part of the national effort in unifying the pronunciation of spoken Chinese. His Mandarin Primer—An Intensive Course in Spoken Chinese (1948) and A Grammar of Spoken Chinese (1968) were among the most important references of standardized modern Chinese.

His love for music and language resulted in song compositions. His New Poetry Songbook (Xin shi ge ji 新詩歌集), a collection of fourteen settings of modern vernacular poems,was published in Shanghai in 1928 and reprinted in Taiwan in 1960.[2] In the preface and introduction, Zhao described to the cultural environment surrounding the creation of these songs. He also explicitly explained—to a new audience—the structures and styles of the works.

He traced the lineage of the songs in the collection to the “art songs” of Schubert and/or Schumann. Thus, solo voice with piano accompaniment would be the norm. As in Lieder, the piano parts had their own musical value. Zhao stipulated that the accompaniments of a few works could be played alone. The vocal parts were for all voice types and gender neutral.[3]

It is a common practice to transpose Lieder to the key most suitable for the voice of an individual performer. For convenience, most Lieder collections are published in three versions: for high, medium, and low voices. Zhao pointed out the impracticality of doing so for a limited market. He chose instead to set the music within a manageable range for most people. He also selected keys which could be easily transposed down a half step by the pianist. A quick lesson on this maneuver was given.

The question arose: Are Chinese art songs merely imitations of Lieder and melodies? In the introductory essays, Zhao carefully explored the evolution of the genre by first tracing the traditional relationship between words and music. Following is an analytical reading of his arguments:

__Recitation vs Singing

Melodious recitation of poetry had long been a tradition in China. Combinations of level and oblique tones—a crucial element of versification—shaped the melodic lines. Poems of the same tonal structure would be recited on the same tune. Tone patterns in jintishi 近體詩 (new style poetry) of the Tang Dynasty and ci of the Song Dynasty were highly regulated, following the standard classifications in rhyme dictionaries such as Qieyun 切韻 and Guangyun 廣韻.[4] Therefore, regional linguistic differences had limited impact on the recitation tunes. On the other hand, regional discrepancies were more apparent in recitations of gushi 古體詩 (old style poetry) and classical prose, as they were less regulated by standardized tone patterns.

Zhao commented that, even though there were basic melodic lines, due to each reciter’s inflections and his/her spontaneous modifications, the actual readings were rarely the same. While delineating the tone patterns, these recitative melodies did not convey the contents and expressions of individual poems.[5]

Preexisting melodies also played important roles in Chinese theater. Two melodic prototypes, namely xipi 西皮and erhuang 二黃—known together as pi-huang 皮黃, were the backbones of traditional theater, such as Peking opera, Guangdong opera and many other regional theaters.[6] Although each opera was notated in gongche notation 工尺譜,[7] without the texts, the melodic similarities would have made it impossible for the audiences to tell the operas apart. Zhao implied that, if not for the rhythmic patterns, the actors would have been “reciting the operas” instead of “singing the operas.”

Richer melodic materials could be found in folk songs. However, the same tune would be used for multiple verses. Good tunes were often “borrowed” to fit new texts. While the melodies were distinguishable from each other, they still did not fully support the poetic details. Contrary to the practice of fitting words into a generic tune, art songs were written specifically to reflect the linguistic and emotional features of each poem.

To clarify, Zhao compared two music settings of “Die Loreley” by Heinrich Heine (1797-1856): The strophic song by Friedrich Silcher (1789-1860) and the durchkomponiert (through-composed) setting by Franz Liszt (1811-1886).[8] The former had been popularized into a folk song in which the melodic theme was repeated several times in each of the three stanzas. In the latter, each section of the poem had different melodic lines supported by piano accompaniments, depicting the scenery, and shaping the dramatic details. Zhao elucidated that, when appropriate, musical repetitions were still found in through-composed songs.

__Poetry vs Songs

Without fixed tone patterns, modern poems could not be recited in the traditional manner. Art song settings, especially the through-composed approach, seemed to be a more suitable way to deliver their structures and contents. Would these poems benefit from the music?

Zhao believed that vernacular poems were best read aloud with proper inflections and expressions. Tonal fluctuations in the spoken words were subtle. No matter how the composers tried to match the pitches (gong-che 工尺) with tone levels (píng-zè 平仄), the suppleness of the inflections could be lost.[9] Sometimes, this could make certain words difficult to understand or it could affect the clarity of the texts. Musical rhythms, however rubato, were more regulated than speech rhythms. Unavoidably, some words would be elongated artificially.

Despite the possibility of being detrimental to the clarity and the expressiveness of poems, musical settings offered additional interest to the presentation. Good singing could enhance the overall aesthetic—this should have been the raison d’être of art songs.

Unless the poet intended for the poem to be used as lyrics, he did not have to consider whether the work would be suitable for singing. On the other hand, it would be necessary for the composer to select texts for musical settings. Zhao suggested the following selection criteria:

Words with sonorous tones
Organized phrasal structure
Easy to comprehend
Rhyming clauses
Repetitions
Declamatory clauses

Literary works and musical compositions derived from them were independent of each other. For example: Shakespeare’s Midsummer Night’s Dream and Mendelssohn’s incidental music for the play, or Goethe’s Faust and Gounod’s opera were separate entities with their own artistic values. However, matching captivating music with enthralling poetry, while reducing some effects from either side, could still achieve exciting results. Thus, the saying of “words [marrying] music.”

__Chinese Music vs Western Music

Traditional Chinese music shares many similarities with its western counterparts:

Solfege: The five tones in the pentatonic scale, gong, shang, jue, zhi, yu (宮商角徵羽) or, in gongche notation, shang che gong liu wu (上尺工六五) can be equated with do, re, mi, sol, la in western scales.[10]
Rhythm: In Chinese music patterns are organized by bǎn (板)—strong beats, played on the clappers, and yǎn (眼)—weak beats, on the small drums. Yì-bǎn-san-yǎn (一板三眼 one ban-three yans) would be similar to 4/4 time in the western music.[11]
Forms: In Chinese music, structural divisions can be marked as jia-yi-jia (甲乙甲), jia-yi-jia-bing (甲乙甲丙), etc., parallel to western forms, “ABA,” and “ABAC.”

Discrepancies are more noticeable in other areas, such as ornaments and portamenti. Chinese ornaments—huayin (花音, flowering tones) should be sung quickly but smoothly, while the western practice emphasizes the precision and clarity of these notes. Portamenti are frequently used in Chinese music, both instrumental and vocal. In western practice, sliding tones are used discreetly.

Since ancient time, Chinese music has been governed by the shí-èr-lǜ (十二律), a twelve-tone temperament similar to the Pythagorean system, using the 3:2 frequency-ratio to create intervals of pure fifths.[12] Instead of forming a chromatic scale—as in the western practice, the twelve tones were used as the foundation (tonic) of individual scales. While the western practice favors heptatonic scales, the Chinese melodies are mostly built upon pentatonic scales.

Some instruments are uniquely Chinese. For example, qin 琴 or guqin 古琴 (ancient qin): a seven-string fretless zither, known as the instrument of scholars and intellectual elites. In addition to the seven open-string sounds, there are ninety-one harmonics and one hundred forty-seven stopped sounds within its four-octave range. Its characteristic gliding stopped sounds are created by shifting the left hand. Extended melodic passages can be played on harmonic overtones.[13]

Zhao argued that, although Chinese music shared many fundamental elements with western music, it was lagging behind in its development:

Western music had more rhythmic variations.
There was greater diversity in western scale patterns, ranging from pentatonic scales, common in Scottish folk music, to twelve-tone series.
Instead of remaining in one key throughout, most western pieces included frequent tonal changes.
In addition to solo wind and string instruments, there were orchestral ensembles, and keyboard instruments, e.g., pipe organs with four manuals and pedals.
There were small pieces in ABA form as well as works in extended format, such as sonata form.
Other than monophonic melodies, there were homophonic and polyphonic textures in western music.

Zhao believed that allowing the musical commonality to supersede national characteristics enabled the advancement of the western tradition. He encouraged Chinese readers/audiences to be open minded and through appreciation of world music to search for the path of Chinese music development.

Using Russian music as an example, he suggested that it would be possible to combine universal musical elements with national characteristics as well as the personal approach of the composers. Huayin and portamenti could be used as much as desired, seven-string qin could be further explored as long as Chinese music making could reach the level the world standard.

__Works in New Poetry Songbook

Rather than describing individual works in the collection, Zhao explained his general approach to art song composition. He considered the implication of harmony a key issue in the development of Chinese music. Even though there could be infinite variations of monophonic melodies, the results were horizontal and one-dimensional. Harmonies would create a two-dimensional infinity. Harmonies would enable key changes, essential to the tonal structure of western music.

Since harmonic progressions did not exist in traditional Chinese music, it would be necessary to borrow the western rules. Nevertheless, these rules could be modified to enhance the “Chinese sounds.” Brief passages of parallel fifth and pentatonic voice leading were among the possible options. Composers such as Debussy and Borodin had previously experimented with these techniques.

Melodically, there were more opportunities to bring in Chinese characters. Zhao, in addition to using ornaments to imitate the traditional singing styles, also adopted old-fashioned recitatives and theatrical passages. He explained to the readers/musicians how his melodic writings evolved gradually from pure western style to almost entirely Chinese.

Finally, and most crucially, Zhao delved into matching the linguistic tones with melody. Poetic meter, combinations of stress and non-stressed syllables, played an important part in setting western texts. While in Chinese songs, the word-tones must be handled properly.

Zhao suggested that there could be two different approaches: One would be to fix the four tones to one or a few given pitches. [14] The other one would be to allow a wider range of suitable options. The former would limit melodic creativities. Hence, the latter would be a desirable choice.

The four tones in modern Chinese were yīnpíng (陰平, the level tone), yángpíng (陽平, the rising tone), shàng (上, the falling-rising tone), and (去, the falling tone). On the other hand, the Middle Chinese tones, used in rhyme dictionaries, were level tone píng (平) and oblique tones, shǎng (上, rising), (去, departing or falling), and (入, entering or checked).[15] Different readings of the same words could result in different melodic handlings. Zhao preferred to observe the traditional level and oblique divisions. Following were his approach in general:

Set level-tone words syllabically—preferably on do, mi, sol.[16] When setting a word/syllable to multiple notes, start with higher pitches. Ornaments did not have to follow this practice.
Use multiple notes for oblique-tone words. It would be possible, but not necessary, to use re, fa, la,and ti.
When level and oblique tones were linked together, use lower pitch for level-tone words.
Only important words, especially rhymed words, were restricted by the above-mentioned rules. Other words could be set freely.
The four above-mentioned rules were suitable for regular song [texts]. For songs of realistic or comic characters, it would be possible to observe the modern tone divisions or to combine the two approaches.

New Poetry Songbook was published at a time when music engraving equipment was not readily available in China. In addition to staff notation, the vocal lines of all the pieces were also presented in numbered notation, commonly used for school songs.[17] Zhao Yuanren self-deprecatingly used the word scherzando to describe his song collections. While being thorough and clear with his explanations, his wording was casual and down-to-earth. However, his eagerness to broaden the horizon of the general public was apparent.

Alarmingly, he rebuked the idea that “real Chinese music” must use solely peculiar things such as parallel fourths or fifths, natural minor[18] or pentatonic scales. To him, this misconception was rooted in “Museum Chinese”—something “picturesque” or “quaint” that would stir up curiosity but not true love or care:

. . .We Chinese must live normal human lives in China. We, the entire country, cannot be dressed in costumes from anthropology museums, readying especially for you to visit. China is not Chinatown in San Francisco. . ..

. . .我們中國的人得要在中國過人生常態的日子, 我們不能全國人一生一世穿了人種學博物院的服裝, 專預備著你們來參觀. 中國不是舊金山的中國市. . ..

Decades later, preparing for the reprint, Zhao held firmly to his convictions.


[1] Born and raised in southeastern China, Zhou was fluent in many Wu dialects. He spoke Japanese, English, German, French and Spanish. He was also knowledgeable in Greek, Latin and Russian.
[2] The first section of “Bouquets in the Vase” 瓶花, the last song in the collection, was set to a poem of the same title by Fan Chengda 范成大 of the Song Dynasty.
[3] Zhao suggested that 《勞動歌》 (”Laboring Song,”1922)would sound more convivial sung by a choir. 《海韻》(“Melody of the Sea,” 1927), the last piece in the collection, was written for mix-voice choir with soprano solo.
[4] Chinese Poetry (VII) Tang Poetry/Goldfish Odyssey
Chinese Poetry (IX): ci/Goldfish Odyssey
Regulated_verse_Wiki, ci_poetry_Wiki,
Qieyun_Wiki, Guangyun_Wiki
[5] 李白_兩首_趙元任教授吟誦_YouTube: Recording of Zhao reciting two poems by Li Bai in Changzhou dialect.
[6] Characteristics of xipi include leaps of large intervals, higher register, and vivid rhythms. Erhuang, on the other hand, has smooth melodic lines, lower range, and calmer rhythmic gestures. For each tune, there was a basic version in moderate tempo called yuanban 原板 (“original rhythm”). Variants with different speed and melodic modifications were used to enhance the dramatic intensities. Actors would often modify the music to deliver emotional details.
[7] Gongche_notation_Wiki
[8]The poem was written in 1824. Silcher’s set it to music in 1837. There were two versions by Liszt. The first one was written in 1841; the second between 1854 and 1856. Zhao did not specify which version was used in his study.
[9] “. . . 唱歌總是定音多, 滑音少, 不能像天然語調那麼定音少滑音多. . ..”
“. . . There are always more definite pitch(es) and fewer sliding pitch(es) in singing, unlike natural linguistic inflections with fewer definite pitches but more sliding ones. . ..”
[10] In practice symbols 凡 乙 were used for fa and ti.
[11] Yì-bǎn-yì-yǎn 一板一眼—one strong beat and one weak beat—is the equivalence of duple meter in western music. As an idiom, the term means following the regulation without wavering or extremely organized.
[12]shí-èr-lǜ_Wiki. The names of the twelve tones are:
Huáng Zhōng 黃鐘, Dà Lǚ 大呂, Tài Cù 太簇, Jiá Zhōng 夾鐘, Gū Xiǎn 姑洗, Zhòng Lǚ 仲呂, Ruí Bīn 蕤賓, Lín Zhōng 林鐘, Yí Zé 夷則, Nán Lǚ 南呂, Wú Yì 無射, Yìng Zhōng 應鐘. Explanations of the system were found in documents from the Waring States period (475–221 BC). The exact pitches changed throughout the history.
[13] Guqin_Wiki
[14] Zhao devised a five-tone diagram—based on the five-line western musical staff—to distinguish the four tones in Mandarin Chinese. Tone_letter#Chao_tone_letters_(IPA)_Wiki
[15] Four_tones_(Middle_Chinese)_Wiki
[16] Zhao seemed to use the words 平音 “leveled pitch(es)” to mean “single [unchanged] note.”
[17] Numbered_musical_notation_Wiki
[18] Without standard translations for western music terminology, Zhao often devised his own Chinese versions. Here, he used the words 平七度小調. Word for word translation would have been “flat-seven-degree minor scale(s).” Since the seventh were usually raised in harmonic or melodic (ascending) minors, I thus translated the phrase as “natural minor scales.”