Old Chinese

History

The earliest known written records of the Chinese language were found at the Yinxu site near modern Anyang identified as the last capital of the Shang dynasty, and date from about 1250 BC. These are the oracle bones, short inscriptions carved on tortoise plastrons and ox scapulae for divinatory purposes, as well as a few brief bronze inscriptions.

The language written is undoubtedly an early form of Chinese, but is difficult to interpret due to the limited subject matter and high proportion of proper names. Only half of the 4,000 characters used have been identified with certainty. Little is known about the grammar of this language, but it seems much less reliant on grammatical particles than Classical Chinese.

From early in the Western Zhou period, around 1000 BC, the most important recovered texts are bronze inscriptions, many of considerable length. Even longer pre-Classical texts on a wide range of subjects have also been transmitted through the literary tradition. The oldest parts of the Book of Documents, the Classic of Poetry and the I Ching also date from the early Zhou period, and closely resemble the bronze inscriptions in vocabulary, syntax and style. A greater proportion of this more varied vocabulary has been identified than for the oracular period.

The four centuries preceding the unification of China in 221 BC (the later Spring and Autumn period and the Warring States period) constitute the Chinese classical period in the strict sense. There are many bronze inscriptions from this period, but they are vastly outweighed by a rich literature written in ink on bamboo and wooden slips and (toward the end of the period) silk. Although these are perishable materials, and many books were destroyed in the burning of books and burying of scholars in the Qin dynasty, other texts have been transmitted as copies. Such works from this period as the Analects, the Classic of Filial Piety, the Mencius and the Zuo zhuan have been admired as models of prose style since the Han dynasty. The Classical Chinese of such works formed the basis of Literary Chinese, which remained the written standard until the early twentieth century.

Script

Each character of the script represented a single Old Chinese word. Most scholars believe that these words were monosyllabic, though some have recently suggested that a minority of them had minor presyllables. The development of these characters follows the same three stages that characterized Egyptian hieroglyphs, Mesopotamian cuneiform script and the Maya script.

Some words could be represented by pictures (later stylized) such as 日 rì "sun", 人 rén "person" and 木 mù "tree", by abstract symbols such as 三 sān "three" and 上 shàng "up", or by composite symbols such as 林 lín "grove" (two trees). About 1,000 of the oracle bone characters, nearly a quarter of the total, are of this type, though 300 of them have not yet been deciphered. Though the pictographic origins of these characters are apparent, they have already undergone extensive simplification and conventionalization. Evolved forms of most of these characters are still in common use today.

Next, words that could not be represented pictorially, such as abstract terms and grammatical particles, were signified by borrowing characters of pictorial origin representing similar-sounding words (the rebus strategy). An example of such a phonetic loan is 來 lái "come", written with the character for a similar-sounding word meaning "wheat". Sometimes the borrowed character would be modified slightly to distinguish it from the original, as with 毋 wú "don't", a borrowing of 母 mǔ "mother".

The final stage was disambiguation of phonetic loans by the addition of semantic indicators, yielding phono-semantic compound characters. For example, the character 其 originally representing jī "winnowing basket" was also used to write the pronoun and modal particle qí. Later the less common original word was written with the compound 箕, obtained by adding the symbol 竹 zhú "bamboo" to the character. This type was already used extensively on the oracle bones, and has been the main source of new characters since then. In the Shuowen Jiezi, a dictionary compiled in the 2nd century, 82% of the 9,353 characters are classified as phono-semantic compounds. In the light of the modern understanding of Old Chinese phonology, researchers now believe that most of the characters originally classified as semantic compounds also have a phonetic nature.

These developments were already present in the oracle bone script. The characters had been extensively simplified and linearized, implying a significant period of development prior to the extant inscriptions. This may have involved writing on perishable materials, as suggested by the appearance on oracle bones of the character 册 cè "records". The character is thought to depict bamboo or wooden strips tied together with leather thongs, a writing material known from later archaeological finds.

Development and simplification of the script continued during the pre-Classical and Classical periods, with characters becoming less pictorial and more linear and regular, with rounded strokes being replaced by sharp angles. The language developed compound words, so that characters came to represent morphemes, though almost all morphemes could be used as independent words. Hundreds of morphemes of two or more syllables also entered the language, and were written with one phono-semantic compound character per syllable. During the Warring States period, writing became more widespread, with further simplification and variation, particularly in the eastern states. The most conservative script prevailed in the western state of Qin, which would later impose its standard on the whole of China.

Phonology

Old Chinese phonology has been reconstructed using a variety of evidence, including the phonetic components of Chinese characters, rhyming practice in the Classic of Poetry and descriptions of later stages of the language, especially the Qieyun, a rhyme dictionary published in 601. Although many details are still disputed, recent formulations are in substantial agreement on the core issues. For example, the Old Chinese initial consonants recognized by Li Fang-Kuei and William Baxter are given below, with Baxter's (mostly tentative) additions given in parentheses:

Various initial clusters have been proposed, especially clusters of *s- with other consonants, but this area remains unsettled.

Most scholars posit optional medials *-r-, *-j- and the combination *-rj- as the origin of the retroflex and palatal obstruents of Middle Chinese, as well as many of its vowel contrasts. However the palatal medial *-j- has been challenged on a number of grounds, and a variety of different realizations for this distinction have been used in recent constructions.

Reconstructions since the 1980s usually propose six vowels:

Vowels could optionally be followed by the same codas as in Middle Chinese: a glide *-j or *-w, a nasal *-m, *-n or *-ŋ, or a stop *-p, *-t or *-k. Some scholars also allow for a labiovelar coda *-kʷ.

Most scholars now believe that Old Chinese lacked the tones found in later stages of the language, but had optional post-codas *-ʔ and *-s, which developed into the Middle Chinese rising and departing tones respectively.

Vocabulary

The improved understanding of Old Chinese phonology has enabled the study of the origins of Chinese words (rather than the characters with which they are written). Most researchers trace the core vocabulary to a Sino-Tibetan ancestor language, with much early borrowing from other neighbouring languages. The traditional view was that Old Chinese was an isolating language, lacking both inflection and derivation, but it has become clear that words could be formed by derivational affixation, reduplication and compounding. Most authors consider only monosyllabic roots, but Baxter and Laurent Sagart also propose disyllabic roots in which the first syllable is reduced, as in modern Khmer.

Sino-Tibetan

Middle Chinese and its southern neighbours Tai–Kadai, Hmong–Mien and the Vietic branch of Austroasiatic have similar tone systems, syllable structure, grammatical features and lack of inflection, but these are believed to be areal features spread by diffusion rather than indicating common descent. The most widely accepted hypothesis is that Chinese belongs to the Sino-Tibetan language family, usually as a primary branch. The evidence consists of some hundreds of proposed cognate words, including such basic vocabulary as the following:

Some progress has been made on the sound correspondences between Chinese and Tibeto-Burman languages, though hampered by the difficulty of reconstruction on both sides. Initial consonants generally correspond regarding place and manner of articulation, but voicing and aspiration are much less regular, and prefixal elements vary widely between languages. Some researchers believe that both these phenomena reflect lost minor syllables. Proto-Tibeto-Burman as reconstructed by Benedict and Matisoff lacks an aspiration distinction on initial stops and affricates. Aspiration in Old Chinese often corresponds to pre-initial consonants in Tibetan and Lolo-Burmese, and is believed to be a Chinese innovation arising from earlier prefixes. Proto-Sino-Tibetan is reconstructed with a six-vowel system as in recent reconstructions of Old Chinese, with Tibeto-Burman distinguished by the merger of the mid-central vowel *-ə- with *-a-. The other vowels are preserved by both, with some alternation between *-e- and *-i-, and between *-o- and *-u-.

Loanwords

During the Old Chinese period, Chinese civilization expanded from a compact area around the lower Wei River and middle Yellow River eastwards across the North China Plain to Shandong and then south into the valley of the Yangtze. There are no records of the non-Chinese languages formerly spoken in those areas and subsequently displaced by the Chinese expansion. However they are believed to have contributed to the vocabulary of Old Chinese, and may be the source of some of the many Chinese words whose origins are still unknown.

Jerry Norman and Mei Tsu-lin have identified early Austroasiatic loanwords in Old Chinese, possibly from the peoples of the lower Yangtze basin known to ancient Chinese as the Yue. For example, the early Chinese name *kroŋ (江 jiāng) for the Yangtze was later extended to a general word for "river" in south China. Norman and Mei suggest that the word is cognate with Vietnamese sông (from *krong) and Mon kruŋ "river".

Haudricourt and Strecker have proposed a number of borrowings from the Hmong–Mien languages. These include terms related to rice cultivation, which began in the middle Yangtze valley:

Other words are believed to have been borrowed from languages to the south of the Chinese area, but it is not clear which was the original source, e.g.

In ancient times, the Tarim Basin was occupied by speakers of Indo-European Tocharian languages, the source of *mjit (蜜 mì) "honey", from proto-Tocharian *ḿət(ə) (where *ḿ is palatalized; cf. Tocharian B mit), cognate with English mead. The northern neighbours of Chinese contributed such words as *dok (犢 dú) "calf" – compare Mongolian tuɣul and Manchu tuqšan.

Affixation

Chinese philologists have long noted words with related meanings and similar pronunciations, sometimes written using the same character. Henri Maspero attributed some of these alternations to consonant clusters resulting from derivational affixes. Subsequent work has identified several such affixes, some of which appear to have cognates in other Sino-Tibetan languages.

A common case is "derivation by tone change", in which words in the departing tone appear to be derived from words in other tones. If Haudricourt's theory of the origin of the departing tone is accepted, these tonal derivations can be interpreted as the result of a derivational suffix *-s. As Tibetan has a similar suffix, it may be inherited from Sino-Tibetan. Examples include:

Another alternation involves transitive verbs with an unvoiced initial and passive or stative verbs with a voiced initial:

Some scholars hold that the transitive verbs with voiceless initials are basic and the voiced initials reflect a de-transitivizing voiced prefix. Others suggest that the transitive verbs were derived by the addition of a causative prefix *s- to a stative verb, causing devoicing of the following voiced initial. Both postulated prefixes have parallels in Tibeto-Burman languages. Several other affixes have been proposed.

Reduplication and compounding

Old Chinese morphemes were originally monosyllabic, but during the Western Zhou period many new bisyllabic words entered the language. For example, over 30% of the vocabulary of the Mencius is polysyllabic, including 9% proper names, though monosyllabic words occur more frequently, accounting for 80–90% of the text. Many words, particularly expressive adjectives and adverbs, were formed by varieties of reduplication:

Other bisyllabic morphemes include the famous *ɡa-lep (蝴蝶 húdié) "butterfly" from the Zhuangzi. More words, especially nouns, were formed by compounding, including:

However the components of compounds were not bound morphemes: they could still be used separately.

A number of bimorphemic syllables appeared in the Classical period, resulting from the fusion of words with following unstressed particles or pronouns. Thus the negatives *pjut 弗 and *mjut 勿 are viewed as fusions of the negators *pjə 不 and *mjo 毋 respectively with a third-person pronoun *tjə 之.

Grammar

Little is known of the grammar of the language of the Oracular and pre-Classical periods, as the texts are often of a ritual or formulaic nature, and much of their vocabulary has not been deciphered. In contrast, the rich literature of the Warring States period has been extensively analysed. Having no inflection, Old Chinese was heavily reliant on word order, grammatical particles and inherent word classes.

Word classes

Classifying Old Chinese words is not always straightforward, as words were not marked for function, word classes overlapped, and words of one class could sometimes be used in roles normally reserved for a different class. The task is more difficult with written texts than it would have been for speakers of Old Chinese, because the derivational morphology is often hidden by the writing system. For example, the verb *sək "to block" and the derived noun *səks "frontier" were both written with the same character 塞.

Old Chinese nouns and pronouns did not indicate number or gender, but some personal pronouns showed case distinctions:

In the oracle bone inscriptions, the *l- pronouns were used by the king to refer to himself, and the *ŋ- forms for the Shang people as a whole. This distinction is largely absent in later texts, and the *l- forms disappeared during the classical period. In the post-Han period 我 and 其 came to be used as general first and third person pronouns respectively. The second person pronouns 汝 and 爾 continued to be used interchangeably until their replacement by the phonological variant 你 (modern Mandarin nǐ) in the Tang period.

There were also demonstrative and interrogative pronouns, but no indefinite pronouns with the meanings "something" or "nothing". The distributive pronouns were formed with a *-k suffix:

As in the modern language, localizers (compass directions, "above", "inside" and the like) could be placed after nouns to indicate relative positions. They could also precede verbs to indicate the direction of the action. Nouns denoting times were another special class (time words); they usually preceded the subject to specify the time of an action. However the classifiers so characteristic of Modern Chinese only became common in the Han period and the subsequent Northern and Southern dynasties.

Old Chinese verbs, like their modern counterparts, did not show tense or aspect; these could be indicated with adverbs or particles if required. Verbs could be transitive or intransitive. As in the modern language, adjectives were a special kind of intransitive verb, and a few transitive verbs could also function as modal auxiliaries or as prepositions.

Adverbs described the scope of a statement or various temporal relationships. They included two families of negatives starting with *p- and *m-, such as *pjə 不 and *mja 無. Modern northern varieties derive the usual negative from the first family, while southern varieties preserve the second. The language had no adverbs of degree until late in the Classical period.

Particles were function words serving a range of purposes. As in the modern language, there were sentence-final particles marking imperatives and yes/no questions. Other sentence-final particles expressed a range of connotations, the most important being *ljaj 也, expressing static factuality, and *ɦjəʔ 矣, implying a change. Other particles included the subordination marker *tjə 之 and the nominalizing particles *tjaʔ 者 (agent) and *srjaʔ 所 (object). Conjunctions could join nouns or clauses.

Sentence structure

As with English and modern Chinese, Old Chinese sentences can be analysed as a subject (a noun phrase, sometimes understood) followed by a predicate, which could be of either nominal or verbal type.

Before the Classical period, nominal predicates consisted of a copular particle *wjij 惟 followed by a noun phrase:

"I am a young person." (Book of Documents 27, 9)

The negated copula *pjə-wjij 不惟 is attested in oracle bone inscriptions, and later fused as *pjəj 非. In the Classical period, nominal predicates were constructed with the sentence-final particle *ljaj 也 instead of the copula 惟, but 非 was retained as the negative form, with which 也 was optional:

(of shooting at a mark a hundred paces distant) "That you reach it is owing to your strength, but that you hit the mark is not owing to your strength." (Mencius 10.1/51/13)

The copular verb 是 (shì) of Literary and Modern Chinese dates from the Han period. In Old Chinese the word was a near demonstrative ("this").

As in Modern Chinese, but unlike most Tibeto-Burman languages, the basic word order in a verbal sentence was subject–verb–object:

"Mencius saw King Hui of Liang." (Mencius 1.1/1/3)

Besides inversions for emphasis, there were two exceptions to this rule: a pronoun object of a negated sentence or an interrogative pronoun object would be placed before the verb:

"The years do not wait for us." (Analects 17.1/47/23)

An additional noun phrase could be placed before the subject to serve as the topic. As in the modern language, yes/no questions were formed by adding a sentence-final particle, and requests for information by substituting an interrogative pronoun for the requested element.

Modification

In general, Old Chinese modifiers preceded the words they modified. Thus relative clauses were placed before the noun, usually marked by the particle *tjə 之 (in a role similar to Modern Chinese de 的):

"... the heart that cannot bear the afflictions of others." (Mencius 3.6/18/4)

A common instance of this construction was adjectival modification, since the Old Chinese adjective was a type of verb (as on the modern language), but 之 was usually omitted after monosyllabic adjectives.

Similarly, adverbial modifiers, including various forms of negation, usually occurred before the verb. As in the modern language, time adjuncts occurred either at the start of the sentence or before the verb, depending on their scope, while duration adjuncts were placed after the verb. Instrumental and place adjuncts were usually placed after the verb phrase. These later moved to a position before the verb, as in the modern language.