Checked tone

A checked tone, commonly known by the Chinese calque entering tone, is one of the four syllable types in the phonology of Middle Chinese. Although usually translated as "tone", a checked tone is not a tone in the phonetic sense but rather a type of syllable that ends in a stop consonant or a glottal stop. Separating the checked tone allows -p, -t, and -k to be treated as allophones of -m, -n, and -ng, respectively, since they are in complementary distribution. Stops appear only in the checked tone, and nasals appear only in the other tones. Because of the origin of tone in Chinese, the number of tones found in such syllables is smaller than the number of tones in other syllables. Chinese phonetics have traditionally counted them separately.

Final voiceless stops and therefore the checked "tones" have disappeared from most Mandarin dialects, spoken in northern and southwestern China, but have been preserved in southeastern Chinese branches like Yue, Min, and Hakka.

Tones are an indispensable part of Chinese literature, as characters in poetry and prose were chosen according to tones and rhymes for their euphony. This use of language helps reconstructing Old Chinese and Middle Chinese pronunciations since Chinese writing system is logographic, rather than phonetic.

Phonetics
From a phonetic perspective, the prototypical entering tone is simply a syllable ending with a voiceless stop that has no audible release:, and/or a glottal stop  depending on the language variety. Middle Chinese has only the first three.

It is customarily called a tone regardless of whether a tonal distinction is possible in such syllables. In languages such as Early Middle Chinese and most varieties of Wu, such syllables do not have contrastive tones (i.e. the tone or pitch of the syllable is entirely predictable) and are therefore phonologically toneless. In languages such as Cantonese or Hakka, a small number of tonal distinctions exist (typically 2), which historically developed as a substitute for the lost Middle Chinese initial voicing.

Some Chinese varieties have innovated new final consonants from such historical syllables. A few dialects of Gan have (from historical ). In some dialects of Cantonese and Gan, the final stop is voiced.

History
The voiceless stops that typify the entering tone date back to the Proto-Sino-Tibetan, the parent language of Chinese as well as the Tibeto-Burman languages. In addition, Old Chinese is commonly thought to have syllables ending in clusters, , and  (sometimes called the "long entering tone" while syllables ending in ,  and  are the "short entering tone"). Such clusters were later reduced to /s/, which, in turn, became and ultimately "departing tone" in Middle Chinese.

The first Chinese philologists began to describe the phonology of Chinese during the Early Middle Chinese period (specifically, during the Northern and Southern Dynasties, between 400 and 600 AD), under the influence of Buddhism and the Sanskrit language that arrived along with it. There were several unsuccessful attempts to classify the tones of Chinese before the establishment of the traditional four-tone description between 483 and 493. It is based on the Vedic theory of three intonations. The middle intonation, ', maps to the "level tone" ; the upwards intonation, ', to the "rising tone" ; the downward intonation, , to the "departing tone". The distinctive sound of syllables ending with a stop did not fit the three intonations and was categorised as the "entering tone", thus forming the four-tone system. The use of this system flourished in the Sui and Tang dynasties (7th–10th centuries), during which the Qieyun rime dictionary was written.

Note that modern linguistic descriptions of Middle Chinese often refer to the level, rising and departing tones as tones 1, 2 and 3, respectively.

By the time of the Mongol invasion (the Yuan dynasty, 1279–1368), the former final stops had been reduced to a glottal stop in Old Mandarin. The Zhongyuan Yinyun, a rime book of 1324, already shows signs of glottal stop disappearing and the modern Mandarin tone system emerging in its place. The precise time at which the loss occurred is unknown though it was likely gone by the time of the Qing Dynasty, in the 17th century.

Example
{{Blockquote|text=怒髮衝冠，憑欄處，瀟瀟雨{{big|歇}}}{{IPA|[xĭɐt]}}； 抬望眼，仰天長嘯，壯懷激烈. 三十功名塵與土，八千里路雲和月； 莫等閒，白了少年頭，空悲切.

靖康恥，猶未雪； 臣子恨，何時滅. 駕長車，踏破賀蘭山缺； 壯志飢餐胡虜肉，笑談渴飲匈奴血. 待從頭，收拾舊山河，朝天闕！|author=Yue Fei|title=滿江紅 (Full River of Red)}}

Mandarin
The entering tone is extant in Jianghuai Mandarin and Minjiang Sichuanese. Other dialects have lost the entering tone, and syllables that had the tone have been distributed into the four modern tonal categories, depending on their initial consonants.

The Beijing dialect that forms the basis of Standard Mandarin redistributed syllables beginning with originally unvoiced consonants across the four tones in a completely random pattern. For example, the three characters, all pronounced in Middle Chinese (William Baxter's reconstruction), are now pronounced jī jǐ jì, with tones 1, 3 and 4 respectively. The two characters 割／葛, both pronounced, are now pronounced gē and gé/gě respectively, with the character 葛 splitting on semantic grounds (tone 3 when it is used as a component of a name, mostly tone 2 otherwise).

Similarly, the three characters 胳阁各 (MC ) are now pronounced gē gé gè. The four characters 鸽蛤颌合 (MC ) are now pronounced gē gé gé gě.

In those cases, the two sets of characters are significant in that each member of the same set has the same phonetic component, suggesting that the phonetic component of a character has little to do with the tone class that the character is assigned to.

In other situations, however, the opposite appears to be the case. For example, the group of six homophones, all  in Middle Chinese and divided into a group of four with one phonetic and a group of two with a different phonetic, splits so that the first group of four is all pronounced fú and the second group of two is pronounced fù. Situations like this may result from the fact that only one of the characters in each group normally occurs in speech with an identifiable tone, and as a result, a "literary pronunciation" of the other characters was constructed based on the phonetic element of that character.

The chart below summarizes the distribution in the different dialects.

Identifying checked tones in Modern Standard Mandarin
There are several conditions that can be used to determine if a character historically had a checked tone in Middle Chinese based on its current reading in Modern Standard Mandarin. However, there are many characters, such as, , , and  which do not satisfy any of these conditions at all.


 * A character with a nasal final in Modern Standard Mandarin will  have the checked tone in Middle Chinese. (The only exception is .)
 * A character with the sibilant final // in Standard Chinese, i.e. those with initials, , and final , will  have the checked tone in Middle Chinese.
 * A character with the final -uai or -uei in Modern Standard Mandarin will not have the checked tone in Middle Chinese. (Exceptions:, and others)
 * A character with a tenuis obstruent initial (pinyin:, , , , , , ）in Standard Mandarin and the third tone will not have the checked tone in Middle Chinese. (Exceptions: ,  when used as a surname, , , , among others)
 * Characters that begin with an unaspirated obstruent and end in a nasal final ( or ) in Mandarin almost never have light level tone (or second tone in Modern Standard Mandarin, marked in pinyin with an acute accent). This is a corollary of the first condition in the table above, where characters that begin with an unaspirated obstruent (pinyin, , , , , ), end in a vowel, and have a light level tone in Mandarin (corresponding to a rising tone in Standard Mandarin) almost always derive from an entering tone (e.g. , , , and  all come from entering tones). As such,  and  are not recognised syllables in Standard Chinese.
 * If a character has a phonetic component that is known to have an entering tone, other characters that have that phonetic component probably have an entering tone. For example, if one already knows that has entering tone, one can conjecture (correctly) that, , ,  also have entering tone. However, there are plenty of exceptions, such as  and , which lack the entering tone.

Wu
Most varieties of Wu Chinese preserve the entering tone. However, no contemporary Wu varieties preserve the,  or  distinction, but instead merges them all into a glottal stop. For example, in Shanghainese, the three lexemes, , , historically ending in , and , all end in a glottal stop, and are pronounced.

In some modern Wu varieties such as Wenzhounese, even the glottal stop has disappeared, and the entering tone is preserved as separate tone, with a falling-rising contour, making it unequivocally a phonemic tone in modern linguistics.

The pitch of the entering tones are divided into two registers, depending on the initials:


 * "dark entering", a high-pitched checked tone, with a voiceless initial.
 * "light entering", a low-pitched checked tone, with a voiced initial.

Many terms with grammatical functions also undergo sporadic evolution and gain a checked tone. This process can be considered a form of lenition, and is sometimes considered a form of glottalization.

Romanization used is Wugniu. This phenomenon can also be seen in many pronouns, such as Shanghainese aq-la (, "we") and Yuyaonese ⁸geq-laq₈ (, "they").

Cantonese
In general, Cantonese preserves the Middle Chinese finals intact, including the differentiation between -p, -t and -k final consonants. Standard Cantonese does not use any glottal stops as final consonants.

There are a few isolated cases where the final consonant has changed as a result of final dissimilation, but they remain in the checked tone.

Like most other Chinese variants, Cantonese has changed initial voiced stops, affricates and fricatives of Middle Chinese to their voiceless counterparts. To compensate for losing that difference, Cantonese has split each Middle Chinese tones into two, one for Middle Chinese voiced initial consonants (light) and one for Middle Chinese voiceless initial consonants (dark). In addition, Cantonese has split the dark-entering tone into two, with a higher tone for short vowels and a lower tone for long vowels. As a result, Cantonese now has three entering tones:


 * Upper dark entering / short dark entering (/)
 * Lower dark entering / long middle entering (/)
 * Light entering

Some variants of Yue Chinese, notably including that of Bobai County in Guangxi and Yangjiang  in Guangdong, have four entering tones: the lower light tone is also differentiated according to vowel length, short vowels for upper light and long vowels for lower light. Thus in such varieties:
 * Upper dark entering / short dark entering (/)
 * Lower dark entering / long middle entering (/)
 * Upper light entering / short light entering (//)
 * Lower light entering / long light entering (/)

Hakka
Hakka preserves all Middle Chinese entering tones and is split into two registers. Meixian Hakka dialect often taken as the paradigm gives the following:


 * "dark entering", a low-pitched checked tone
 * "light entering", a high-pitched checked tone

Middle Chinese entering tone syllables ending in whose vowel clusters have become front high vowels like  and  shifts to syllables with  finals in some of the modern Hakka, as seen in the following table.

Min
Southern Min (Minnan, including Taiwanese) has two entering tones:


 * Upper (dark, ), also numbered tone 4
 * Lower (light, ), tone 8

A word may switch from one tone to the other by tone sandhi. Words with entering tones end with a glottal stop ([-ʔ]), [-p], [-t] or [-k] (all unaspirated). There are many words that have different finals in their literary and colloquial forms.

Eastern Min, as exemplified by Fuzhounese, also has two entering tones:


 * Upper/dark entering,, which in Fuzhounese has the tonal value and ends in the glottal stop /ʔ/. This tone contour is not shared with any other tone category.
 * Lower/light entering,, which in Fuzhounese has the tonal value and also ends in the glottal stop /ʔ/.

Within its complex tone sandhi laws, Fuzhounese has a split in sandhi behavior between two separate upper/dark entering tones. This is believed to be a reflex of an earlier stage in its development, where final /k/ was distinguished from final /ʔ/.

In the related Fuqing dialect, a proportion of entering tone lexemes have lost their glottal stop and have merged into the phonetically equivalent tones:
 * Upper/dark entering,, with value , is merging into upper/dark departing, , with value.
 * Lower/light entering,, with value , is merging into upper/dark level, , with value.

This merger can also affect sandhi environments, but there is the option to use the sandhi pattern of the former checked tone while still eliminating the final glottal stop.

Additionally in Fuqingnese, sandhi environments where the light entering tone is non-final cause the glottal stop to weaken and in some tones lost, and where the tone changes to a low sandhi tone, the glottal stop is completely lost. The dark entering tone on the other hand retains its glottal stop in sandhi environments.

Entering tone in Sino-Xenic
Many Chinese words were borrowed into Japanese, Korean and Vietnamese during the Middle Chinese period so they preserve the entering tone to varying degrees.

Japanese
Because Japanese does not allow a syllable to end with a consonant except ん n, the endings -k, -p, -t were rendered as separate syllables -ku or -ki, -pu, and -ti (Modern -chi) or -tu (Modern -tsu) respectively. Later phonological changes further altered some of the endings:


 * In some cases in which the ending is immediately followed by an unvoiced consonant in a compound, the inserted vowel ending is lost, and the consonant becomes geminate.
 * Examples: 学 + 校 (> Modern Japanese ) becomes 学校, and 突 + 破 (> Modern Japanese when standing alone) becomes 突破 (breakthrough)
 * The ending changes into . . That process can be followed by  and.
 * Example: 十 becomes (phonologically )

Recovering the original ending is possible by examining the historical kana used in spelling a word, which has also aided scholars in reconstructing historical Chinese pronunciation.

Korean
Korean keeps the and  endings while the  ending is represented as  (tapped, , if intervocalic) as Sino-Korean derives from a northern variety of Late Middle Chinese where final  had weakened to.

Vietnamese
Vietnamese preserves all endings, and  (spelt -c). Additionally, after the vowels ê or i, the ending -c changes to -ch, giving rise to -ich and -êch, and ach (pronounced ) also occurs for some words ending with -k.

Only the sắc and nặng tones are allowed on checked tones. In Sino-Vietnamese vocabulary, those tones were split from the Middle Chinese "entering" tone in a similar fashion to Cantonese. Whether the syllable tone should be sắc or nặng depends on the original Middle Chinese syllable's initial consonant voicing.