Empires Of The Word Part 8

The Libyans first put pressure on Egypt in the thirteenth century BC, a generation after the fall of Akhenaten. We read of desert campaigns by the pharaohs Seti I and Ramses II, but there appears to have been a steady trickle of immigration. Companies of Libyan troops, notably the Qahaq, Shardana and Mashwash, were accepted into the Egyptian army as auxiliaries.4 Ramses' successor Merneptah (1211-1202) reports a ma.s.sive victory against would-be invading armies of Libyan peoples, Libu, Mashwash and Tjehenu. And Ramses III, a generation later, tells of similar defensive actions c. 1179 and 1176. Nevertheless, a steady infiltration into Egypt seems to have continued, and the Libyan presence became a fixture in the Delta area. Ramses III himself had a Libyan slave, Ynene, serving him at court.5 Two hundred and seventy years later, the Libyan faction had established itself with sufficient stability to marry into the royal family. The XXII dynasty, ruling not from Memphis but from Tanis in the Delta, was founded by the Libyan parvenu Shoshenq, a Mashwash. It lasted 230 years, although it was riven by family feuding and was forced to accept a joint (equally Libyan-dominated) kingdom with a separate dynasty set up in another Delta town, Taremu (Leontopolis).

The incoming Libyans would have spoken a language related to modern Berber or Tamazight, still spoken in much of North Africa. But the linguistic effect of their arrival is imperceptible. An Egyptian pharaoh of the twenty-first century, Inyotef, had had a dog called 'abaqero, which seems to be the Tuareg Berber name for a greyhound, abaikour.6 And among the Egyptian numerals, the word for 'ten', mudjaw, is reminiscent of the Berber mraw.7 This is not much.

To the Egyptian south was the land of Kush. In this direction aggression flowed in the opposite direction from that across the Libyan border. The Egyptian motive can be inferred from the transparent etymology of their name for Kush, Nubia-from nabaw (Coptic nub), 'gold'-although the chief mines were inconveniently sited in the eastern deserts. But like Egypt, it could also be seen as an integral part of k.u.mat, 'The Black Land', made up of fertile Nile silt, the kingdom that existed only as a ribbon development along the great river. Egypt had been operating south of the natural boundary at the first cataract throughout the Old Kingdom, mining gold and establishing a settlement at Buhen, by the second cataract. It gained full control of Nubia in the nineteenth century, lost it again in the eighteenth, re-established control in the sixteenth and then held it for five hundred years. The Egyptian viceroy was given the t.i.tle ZIR nasuwt kus, 'King's Son of Kush', to emphasise his centrality in the government. Around 1087 the holder of this office abused his position to occupy the Egyptian capital, Thebes, and then withdrew south of the first cataract to declare effective independence for Nubia.

Nothing more is then heard of Nubia for 260 years, but around 728 the ruler of Kush, now based at Napata but investing himself with full pharaonic splendour, a.s.serted a claim to celebrate the worship of the G.o.ds at Thebes, Memphis and Onw (Heliopolis). He was able to enforce his claim, and the next sixty years saw Kus.h.i.tes in (fairly loose) control of Egypt. The unity of the Black Land had come back to haunt its erstwhile masters.

This unity was ended, as it happened, by a full-scale a.s.syrian invasion, coming in from the opposite end of the country in 664 BC. In the aftermath, a new dynasty in Egypt restored indigenous control within its traditional borders,* while the Nubian kings returned to their own land and moved their capital from Napata to Meroe, 400 kilometres farther up the Nile. There they founded the Meroitic civilisation, which lasted until AD c.250, with an alphabetic script based on hieroglyphs. The language they wrote in this way is not related to Egyptian, and is not fully understood to this day.

Once again there was no known impact on Egyptian as used in Egypt itself, despite the long coexistence of Egypt with Nubia. The details of influence are difficult to judge since we have no direct evidence of the language spoken in Kush at the time. During the period of Egyptian control of Kush, Egyptian must have been used widely at elite levels in its northern regions, but use of Egyptian did not survive the withdrawal of links between the two countries, despite the evident enthusiasm for things Egyptian which persisted south of the border. The mutual imperial adventure had lasted, on and off, over two thousand years, but it had left both partners without any lasting linguistic link.

Another country where Egypt attempted conquest was the land of Canaan to its north-east. Since the earliest period there had been trade links with Palestine, and around the middle of the second millennium these became particularly strong with the Phoenician city of Byblos, which supplied cedar timber logged in Lebanon. Around 1830 BC, a pharaoh invaded the south of Palestine, but little is known of his motives or any consequences. Four centuries later, there was a sustained campaign to control the whole country as far north as the borders of Mitanni. This has been explained as an attempt to free Egypt once and for all from the threat of foreign domination, recently suffered under the so-called Hyk-sos kings (a Greek rendering of hqR hrst, 'ruler from abroad'). But there is no evidence, linguistic or other, that this dynasty, whoever they were, had come from the north-east.

Whatever the motive, Egypt did succeed in establishing Egyptian over-lordship throughout Palestine and Syria as far as Ugarit in the north. This is confirmed by the Amarna diplomatic correspondence, which relates to the years from 1345 to 1330 BC, and is largely taken up with exchanges of letters between the pharaoh and many of his Canaanite va.s.sals, notably Ribhadda, the ruler of Byblos. This part of the correspondence is exclusively in Akkadian. The letters from the Egyptian side are in quite good Akkadian, but the answers that came back are in a dialect heavily influenced by Canaanite languages.8 Neither side was fully at ease in this lingua franca. But the point for us is that after a century of political domination Egypt had not transmitted effective knowledge of its language, not even to kings and officials who were professing themselves servants of an Egyptian master.* Instead they communicated in the language of the princ.i.p.al eastern power.

Compet.i.tion from Aramaic and Greek.

That power, first focused in a.s.syria, later in Babylon, finally in Persia, continued to grow in influence over the next thousand years. As Egypt lost its control of Palestine (its last hurrah was the campaign of the Libyan pharaoh Shoshenq through Palestine around 925), and then the eighth century BC saw a.s.syria advance its control in the same region, Egypt began to attract refugees and exiles. The language they spoke was Aramaic, which by this time had spread all over the Semitic-speaking Middle East, and had even replaced Akkadian throughout the a.s.syrian empire.

In the seventh century BC, Aramaic entered Egypt in earnest, borne by the a.s.syrian invasion force of 671-667 which sacked Thebes and installed a puppet pharaoh. But a.s.syrian domination turned out to be transient, and Psamtek, the son of the quisling pharaoh Neko, was able to reclaim Egypt's independence by 639. He soon began to rea.s.sert Egypt's role in Palestine, occupying the Philistine capital Ashdod in 630, and defeating and killing Josiah, king of Judah, in 610. His successors continued the policy for another sixty-five years, taking advantage of the eclipse of a.s.syria by Babylon, and turning Palestine and Syria as a whole into a buffer zone for all the hostilities between Egypt and Babylon. The sack of Jerusalem in 587, and the exile of the Jews to Babylon, was one of the prices that others paid for this policy.

Probably the net effect of this on language was to bring into Egypt not Aramaic, but Greek. An opportunistic alliance with Ionian and Carian pirates had enabled Psamtek to shake off a.s.syria. This set the tone for the dynasty's practice of acting in consort with Greeks, both militarily and commercially. An Egyptian fleet of Greek-built triremes patrolled the Red Sea and Mediterranean coasts, and there was a Greek mercenary contingent with the Egyptian forces sent up the Nile on a last mission against Nubia in the 590s. The Greek trading colony of Naucratis was established close to Sais in the west of the Delta, as a treaty port very comparable to Shanghai in China in the nineteenth and twentieth centuries AD. There was a roaring trade, notably in Egyptian wheat and linen, paid for with Greek wine and silver. Greeks, when high on wine, says Bacchylides, a poet of the fifth century, would fantasise about ships from Egypt laden with wheat.9 This was the beginning of a rich, cosmopolitan atmosphere in the Delta that was to be fulfilled in the expansion of Alexandria as a Greek city three hundred years later. The sound of Greek would have become familiar in Egypt, even if few as yet would have been learning it.* But before Greek reached its acme, Egypt would undergo an involuntary infusion of Aramaic.

Aramaic, besides being the language of the Babylonians, was also adopted as the official language of the Persian empire, and it was this state which achieved the hitherto impossible task of subjecting Egypt durably to foreign rule. Egypt, drunk on Greek wine, was brought down to earth when the Persians marched in in 522 BC, deposed and killed the pharaoh Psamtek III, and set up a standard Persian administration with Egypt reduced to a province under a satrap.

Persian rule lasted for two centuries, tempered by a resurgence of Egyptian independence in the fourth century that was later crushed. The Aramaic language established itself not just as a language of government and law, but also as a widespread medium of private communication. In fact, an accident of climate rather distorts the record. Because of its dry climate, Egypt provides the vast bulk of doc.u.ments in Aramaic that have survived from this period, whether on papyrus, parchment, painted on stone or incised on metal.

Aramaic, then, was the first language in three millennia to make a significant inroad into Egypt. When Alexander took the country in 332 BC, initiating three centuries of Greek rule, he found an administration run in Aramaic; in some respects, for instance in the law courts, this language persisted under the Ptolemies,10 but in general Aramaic was replaced in official use by Greek. Although the Ptolemies took their role as Greek successors to the pharaohs seriously, and Greek Egypt became an autonomous and prosperous country again, the Egyptian language was henceforth relegated to the extremes of sacred and profane: in the temples, and on the lips of the common people. Alexandria, which replaced Athens as the academic centre of the ancient world, was a Greek-speaking city. Famously, Queen Cleopatra, the last Ptolemy to rule (51-30 BC), was also the first to learn Egyptian-and that apparently only because she had a pa.s.sion for languages.

There was pleasure in the very sound of her voice. Like a many-stringed instrument, she turned her tongue easily to whatever dialect she would, and few indeed were the foreigners with whom she conversed through an interpreter, since she answered most of them in her own words, whether Ethiopian, TroG.o.dyte, Hebrew, Arab, Syriac, Median or Parthian. The kings before her had not even had the patience to acquire Egyptian, and some had even been lacking in their Macedonian.*

Changes in writing.

The Egyptian language went through more radical revolutions in its written form than it did orally. The elegant and exact pictorial symbols familiar from Egyptian monuments were called (by the Greeks) hieroglyphs, 'sacred carvings', translating the Egyptian term ', maduww nats ar, 'words of G.o.d' (the phrase also used for Ptah's creative words in the text that heads this chapter). We have no indication as to how they arose, and they undergo essentially no modification in the 3400 or so years for which we see them in use, although in the last few centuries, when Egyptian religion was increasingly an antiquarian practice within a h.e.l.lenised and Christianised country, the scope the system gave for symbolism and imagery has increasing play. Vast numbers of new pictograms are invented, showing that the system is no longer bound by the constraint of being a practical script. The last inscription dates to AD 394, after which it was suppressed by the Christian authorities.*

They had, from the time of the first non-monumental doc.u.ments (c.2600 BC) been paralleled by an equivalent but more cursive script, called hieratic-'priestly'. These two scripts made up what was essentially a single system, which could be rendered either in monumental glyphs or a cursive scrawl, with about 175 signs interpreted as consonants or sequences of consonants, and a few hundred signs used in conjunction with them to specify meanings.

From the seventh century BC, a new style of writing, known as demotic-'popular'-began to be used: it began as a radically simplified form of hieratic writing, but soon diverged from the traditional system when the link with the original hieroglyphs was forgotten.

After the Greek conquest at the end of the fourth century BC, Greek glosses begin to appear in demotic texts, to clarify a difficult reading here and there. Literacy in Greek was becoming widespread. Despite this, the indigenous system of writing still had a very long way to run. The last dated demotic text is from AD 452,784 years after the Greek conquest, 482 years after the Ptolemies had been supplanted by Rome, and 310 years after the apostle St Mark is said to have first preached in the then Egyptian capital of Alexandria. Like the last hieroglyphs written fifty-eight years before, it was found on the last outpost of Egypt, the island of Philae.12

Final paradoxes.

As we have seen, Christianity was to put an end to hieroglyphic writing and with it the central stream of ancient Egyptian culture. But despite this it had a last perverse effect, ensuring the long-term survival of the Egyptian language itself. By the third century AD Egyptian had long lost any role in government or elite life, which were now conducted exclusively in Greek. Yet at this very point, the newly rising force of Christians saw the language as the best means to advance the conversion of the Egyptian people. As such, they made it the vehicle of a new sort of literature, in which the Greek alphabet would be used to represent Egyptian. Since the Egyptian language is more complex in its sound system than Greek, six new letters (borrowed from the demotic script) were added: and so the Coptic alphabet was created. The new tradition began with translations of the Bible, then expanded into original compositions, narrating the lives of the Fathers of the Egyptian Desert, St Pachomius and his followers. Coptic became a major channel for the development of the Christian doctrine, with homilies, letters and polemics all widely read in the Egyptian Church.

Egyptian was written in this way for another thousand years. Ironically, it was this late-acquired a.s.sociation with the Christian Church which saved it; by contrast, the lightning spread of Islam and Arabic in the seventh century soon blotted out the language of the previous masters, Greek.

Egyptian, now known as Coptic, had survived the first onslaught; but the threat from Arabic was always more insidious than that from Greek. Islam, after all, was an egalitarian religion; once Arabic was accepted, there were no other bars to social preferment under the new regime. Over the centuries, the fortunes of the Coptic language ebbed with its a.s.sociated religion. The last great work written in Coptic is the Triadon, a long poem composed shortly after 1300. Even a hundred years later, Christians in Upper Egypt were said to speak little else,13 but it seems that by the end of the sixteenth century Coptic conversation was gone, or almost gone. Its recitation, in the liturgy of the Coptic Church, has lasted to our day.

Language from Huang-he to Yangtze.

The Master said: Learning without thinking is useless. Thinking without learning is dangerous.

Confucius, Lunyu (a.n.a.lects), ii.15 The basic pattern of the history of the Chinese language is very similar to that of Egyptian, the maintenance of unity and linguistic stability despite repeated alien influxes.

Origins.

The language's closest relatives are found in Tibet and Burma, but they are not close: Chinese is generally seen as a separate branch of the Sino-Tibetan language family, with no special link to any of the other major languages in it: these include Tibetan, Karen, Burmese, and even such languages of southern China as Yi, Lisu and Jingpo. In their basic structure, all these languages are very similar, as tone languages, with most of the words or word roots monosyllabic, and no inflexion of nouns, adjectives or verbs. But this is not enough to define the family: rather it defines the area, since other unrelated languages in the neighbourhood, such as Thai, Zhuang, Hmong and Mien, are also like this.

The Chinese language first turns up in the valley of the Yellow River or Huang-he. The earliest record is now a matter of controversy. In 2000, Chinese scholars recognised written characters in the markings on some 4800-year-old wine cups, found at Juxian in Shandong ('Mountain East') province, where the river meets the sea. Whether that a.n.a.lysis is correct or not, the next-oldest characters are still a good 3400 years old: they were found written on bronze vessels, and on tortoisesh.e.l.ls and ox shoulder blades (heated until they cracked-a means of telling fortunes), near Anyang in Hebei ('River North') province. Although the symbols are by origin pictographic, the system as a whole clearly represents the Chinese language. Visual puns are used to convey words with more abstract meanings (e.g. , originally a symbol for lai, 'wheat', represents lai, 'come'), or more specific ones (e.g. lang, 'wolf, is shown as , a combination of qusn, 'dog', and liang, 'good').*

The subsequent history of standard Chinese as spoken is conventionally divided into many periods, Old Chinese (up to 500 BC, represented by the Shijing, 'Book of Poetry'), Middle Chinese (500 BC-seventh century AD, represented by the Qieyun rhyming dictionary), Old Mandarin (seventh to fourteenth centuries), Middle Mandarin (fourteenth to eighteenth centuries) and Modern Mandarin ever since. The prominence of poetry in the early part of this record is not a matter of aesthetics. Given the indirectness of the connection between Chinese script and its p.r.o.nunciation, the evidence for the development of the spoken language comes mostly from detailed a.n.a.lysis of verse, particularly looking at which words rhyme.

The written language itself does not give much away about language development over the last 2500 years, since the cla.s.sical language, known as wenyan (), was defined in the Chunqiu, 'Spring and Autumn', period (770-476 BC), when the great cla.s.sics such as Confucius's a.n.a.lects were written, and was kept unchanged virtually ever after. It was only in the twentieth century that wenyan ceased to be the usual means of written expression, and a new written style, based on the words and structures of Mandarin, became universal. But wenyan was formed in a region (the north-east) and at a time that is only a millennium or so after the beginning of Chinese progression across the country. As such, it gives a useful baseline for the changes that have affected different modern dialects over the two and a half millennia since it was the vernacular. For example, in one of the paradoxes typical of language history, it shows that the least-changed dialect of Chinese is the one spoken farthest away from the north-east: Cantonese, known to the Chinese by the name of the long-independent 'barbarian' kingdom of Yue.

The gap between written evidence and spoken reality means that a fair amount of the detail of how influences have played out in the history of the language must remain conjectural. We can only infer, and we cannot fully doc.u.ment, the forces that we shall be describing, some of them operating piecemeal on Chinese to produce the variety of dialects heard, especially in the south, but others keeping the vast majority of speakers in close touch with each other even as the spoken standard gradually moved, all down the ages.

The degree of political unity in China, although its cyclical rise and fall is the usual tick-tock of Chinese history, is not particularly useful in recounting the spread and transformations of the Chinese language. Following the archaeological evidence, Chinese culture spread out from the middle Yellow River valley in all directions, but most significantly towards the south. In the Shang period (middle to late second millennium) we already find artefacts south of the Yangtze, and these spread out upstream into central China in the Zhou (early first millennium BC). But we know that a language different from Chinese was still spoken in the kingdom of Chu (approximately modern Sichuan, north of the Yangtze) in the third century BC*

Geographically, Chinese was moving from the cold, dry northern plains where wheat and millet were cultivated into the warmer, wetter uplands where the staple was rice. As well as a difference in climate, there was a difference in terrain, which made the going much tougher in the south: nan chuan bei ms, 'south boat, north horse', as the proverb has it. In practice waterways, defined by nature rather than human resource, are the only way to travel easily in the south. This was not a barrier to Chinese cultural and linguistic spread, but it did mean that uniformity, cultural or linguistic, was never so easily imposed there.

The motive behind the movement southward was no doubt the quest for more fertile soil, and its success must have been backed by the advantages in technology that the northerners were acc.u.mulating, symbolised by possession of a written language and large-scale organisation. The first reflection of this on politics comes in 221 BC, with the command of Shi Huang Di, the First Emperor, who unified most of central China, to half a million colonists to go and fill his newly conquered territories 'among the various Yue peoples'. By this time there was a political motive to add to the economic one: the despot of a united China desired to separate the traditional families from their ancestral power bases; and the political push was renewed from time to time over the next millennium.

First Unity.

Shi Huang Di ('First Emperor'), who had converted his rule of the Qin state into the first overlordship of all the known states of the Chinese world, was for many reasons a significant figure. He reigned over China for only eleven years (221-210 BC), after thirty-seven on the throne of Qin, but what years they were: besides completing the Great Wall (invaders from the north were already a problem), abolishing the power of feudal lords, carrying through an intellectual purge in a notorious rampage of book-burning, and installing the Terracotta Army in his tomb in the then capital, Chang-an, he is also famous for the standardisation of Chinese characters, as part of a general programme to introduce common laws, weights and measures. This meant imposing the local standard of his (far western) state of Qin, which happened to be one of the most conservative in use at that time. It existed in two versions, the heavily pictorial zhuanshu, 'seal script', still occasionally seen on ornate inscriptions and official seals, and the more cursive lishu, 'clerical script'. This latter was taken up under the Han empire that followed, and codified in a dictionary of the time, the Shuowen Jiezi of Xu Shen. This system has been the basis of Chinese writing, ksishu (, 'standard script'), ever since.

Conscious of a common language in wenyan and a common script in ksishu, Chinese people took a millennium to begin to notice people diverging: early Tang literature (seventh century AD) talks of the south differing from the north in its fang-yan, 'regional speech', the normal word for a dialect: this came to be a pretty strong term, also to be applied (much later14) to refer to foreign languages such as Korean, j.a.panese, Mongolian, Manchu and Vietnamese.

The languages spoken by China's equestrian neighbours to its north and west were quite unrelated to the family that includes Chinese, the Sino-Tibetan languages already mentioned. Furthermore-and in this they differed from the tongues of China's southern neighbours-they were not like Chinese typologically either. Like their modern descendants, the so-called Altaic languages* of central Asia, including the Turkic, the Mongol and the Tungus families, they are all highly polysyllabic; their words, at least the nouns and verbs, are built up systematically and agglutinatively out of strings of short elements. They are not tone languages, but they make extensive use of the principle of vowel harmony, so that the vowels in the suffixes echo the vowels of the word's root. Their word order places the verb at the end of the sentence. In all these respects, they are radically different from Chinese, a monosyllabic tone language with little or no word formation, and a basic order in which the verb comes second in the sentence.

The Xiongnu* were the princ.i.p.al steppe nomads of Mongolia and Turkestan in the third century BC. Despite their major role in Chinese history it is extremely difficult to find evidence of what their language was like. There is, however, a single quotation of ten Chinese characters, giving some advice of the Buddhist monk Fotudeng to a Xiongnu king. The characters would have been read in the fourth century AD as syog tieg t'iei lid kang b'uok kuk g'iw t'uk tang.

If we follow Louis Bazin in reading this as sug tagti idqang, boqughigh tutqang your army send-out, warlord hold we can infer that their language was Turkic, rather than Mongol or Tungus.15 Three Chinese kingdoms of the north, the Qin, the Zhao and the Yan, had each built sections of wall to keep the Xiongnu out. The walls were unified and lengthened when the Qin emperor incorporated all the kingdoms into his realm. The Chinese also learnt how to oppose the Xiongnu with their own cavalry tactics. Hostilities continued for five hundred years, and for all this time the Chinese were successful in keeping the barbarians out of China, and in maintaining a forward policy that kept control throughout the western regions now known as Gansu and Qinghai; in this way, the Silk Road was secured, as well as access to the far-away horse-breeding grounds of Ferghana by the Pamirs, vital for the Chinese defence. However, their defence also depended on maintaining an active frontier garrison, and it was a costly exercise to keep the guards supplied. When the centralised government of China broke down at the end of the Han dynasty, this failed, and it became possible for the Xiongnu to penetrate the wall.

Retreat to the south.

A confused and increasingly b.l.o.o.d.y period ensued, leading in the fourth century AD to open compet.i.tion among a number of Turkic and Mongol hordes for control of the north, and exposing the total impotence there of the traditional government. The effect was to displace southward the centre of Chinese. In 317 a new dynasty was founded in Nanjing, 'Southern Capital', while different Turkic and Mongol hordes contested the north. Ultimately, the two centuries to 557 were dominated in the north by the Tabgach,* who at least proved effective in defending what they had won. These new lords were speakers of a Turkic language, but they soon endeavoured to take up local forms, adopting the Chinese name Wei. This policy appears to have needed some enforcement, or at least encouragement: six generations later, in 500, their ruler, Xiaowen, outlawed by decree the Turkic language, costume and customs.

It was rather similar, politically and linguistically, to what was going on in the old Roman empire at the same time, with Germans taking over its heartland in western Europe, changing but not supplanting its language as they attempted to adopt it, and the successors of the old Roman power retrenching into what had historically been non-Roman lands in the eastern territories, the Balkans, Greece and Anatolia. Yet the Chinese language faced no compet.i.tion from a potential equal, as Latin faced Greek in the eastern Mediterranean. The land, in all its parts, was dominated by Chinese, even if increasingly spoken by people with some very strange accents.

Down in the south a unified Chinese dynasty continued; there large numbers of well-to-do Chinese immigrants were gradually spreading the range of Chinese. They had moved partly to escape the invaders, but also to occupy the more fertile land drained by the Yangtze. The languages of the native population there, whether of the Tai, Sino-Tibetan or Hmong-Mien families, were all of a type quite similar, though often unrelated, to Chinese. The result was a relatively smooth take-up of Chinese by learners in the south: some of the new Chinese dialects that arose, especially the southernmost (called Yue, or Cantonese), sound very much like the original.

Middle Chinese of the seventh century AD had syllables that could end in m, n, ng, p, t, k or a vowel, and so does Modern Cantonese, just like the (unrelated but neighbouring) southern language Zhuang; in Mandarin final m has become n, and final p, t and k have all been dropped. Again Middle Chinese is inferred to have had three tone contours, and a separate, so-called 'entering', pattern for words ending in p, t or k. These later split to eight tones, with a high and a low onset, depending on whether they started with a voiced or voiceless consonant (b-d-g-z-j versus p-t-k-s-c). This is the basis of the system in modern Cantonese, and also in Zhuang; Mandarin has taken a different route, splitting only one of the original tones, but when it dropped final p, t and k it a.s.signed all the words affected to one of the other tones. It has ended up with four tones, while Cantonese (and Zhuang) have eight.16 In 589 it proved possible to reunite the country. A new Chinese golden age, of prosperity if not always peace, began under the Sui and then the Tang dynasties. Throughout this period, Chinese continued to spread southward.

The Tang dynasty lasted until the end of the ninth century, when it degenerated into a power struggle among regional warlords. Many foreign missions reached China in this period, including Buddhists from India, Nestorian Christians, Zoroastrians and Manichaeans, and Muslims. This would have spread the sounds of the Sanskrit, Aramaic, Persian and Arabic languages to the major centres, where they would have been used in worship; but the numbers actually speaking them must have remained tiny. In any case, by the end of the Tang all except the Buddhists and Muslims had been purged out of existence. During the eighth and ninth centuries there was an increasing threat of incursions from Tibet in the west, and stout resistance from the Nanzhao natives in Yunnan ('South of the Clouds', in the south-west) but no long-term loss of territory. This period (from 847) also saw another Turkic-speaking group, the Uighurs, settle in the northerly province of Gansu, and set up an independent kingdom, friendly to the Chinese, in the far west (modern Xinjiang).

The breakdown of central government was repaired after half a century (960) by the Song dynasty, but not before the extreme north, Manchuria and the lands north of the Great Wall, had been taken by the Khitan, a Mongolian tribe; Gansu too, in the north-west, was lost, invaded by the Tangut, who spoke a language related to Tibetan. The Tangut held on to this area; but the Khitan were in 1115 overwhelmed by another group from farther north-the Jurchen, a Tungus-speaking people, whom the Chinese, ill-advisedly, a.s.sisted. Although the Jurchen adopted the Chinese name and style of Jin (, 'golden'), they almost immediately turned on their allies and, after invading much of the south as well as the north, were left in control of the entire valley of the Huang-he, the traditional Chinese heartland. This they held (like the Tangut) until displaced by one greater, Genghis Khan himself, who led a Mongolian invasion in 1211.

As so often, it proved much easier for the invaders to overrun the north than the south. For two generations the Song dynasty maintained a defence of the southern empire, based on Hangzhou, until in 1279 the Mongols were able to take them in the rear, having first conquered Yunnan (and indeed the north of Vietnam) in the south-west.

For the first time, a non-Chinese speaking dynasty (Mongols, now known as the Yuan, , 'Original') controlled the full extent of China. Since the Mongols by this time also controlled most of the rest of Asia, it could be thought lucky for China that the Mongol Kublai Khan decided to move his capital from Kara Korum in Mongolia to Beijing (, 'Northern Capital'), since otherwise it might have suffered the fate of all colonies, to be disregarded by its ruler; but in any case, the unity of the Mongol empire was lost by 1295. The newly converted Muslim Khans of the west refused to accept the sovereignty of Kublai Khan's successor at Beijing, since he was a Buddhist.

Mongol control of China did not last much longer. Although Kublai was famous for his civility, his successors were less distinguished. It is worth mentioning the last of the dynasty, Togan Timur (1333-1369), since among much anti-Chinese legislation he pa.s.sed laws forbidding Chinese to read or write Mongolian. It is evident that a strict racial policy was being followed. One would have expected, by comparison with the Manchu who were to follow much later-or the contemporary, but of course quite unknown, example of the English in Ireland*-that the elite would be pa.s.sing laws to prevent their own members from taking up the language of the conquered people.

In 1369 Togan Timur and his Mongols ended up chased out by a popular Chinese warlord turned national hero, who established himself as the first Ming emperor. There was then for three centuries no interference by outsiders in the government of China.

Northern influences.

Then the Tungus-speaking Jurchen people, now to be known as the Manchus, gained a second chance to dominate China. This invasion was the last permanent penetration of China by speakers of a foreign language.

In the early seventeenth century, the Manchus had been reorganised, under two able leaders, and advanced into the northern marches of Chinese territory to establish a capital at Mukden. Then, in 1644, they had the luck to be invited into Beijing as a tactical move in a struggle between two generals contending to replace the Ming. The Manchus took the opportunity to install themselves, styling their new dynasty with the name Qing (, 'Pure'), and by 1651 had put down all resistance in the rest of China. Although they came speaking their own language, and it remained an official written language of the Chinese state until the end of the dynasty in 1911, it had died out in speech even at court by the eighteenth century. The language did not survive even in Manchuria itself, a curious victim of its people's successful takeover of China and its way of life. Today it is only spoken, under the name of Xibo, by the descendants of a detachment of troops dispatched from the Manchurian capital Mukden to Xinjiang in 1764-a north-eastern language now spoken only in the Chinese north-west.

It was into the north that the invaders came, and the Chinese spoken in the north went on to become the standard language for the country. But although the northern dialect underwent significant changes, they can only partly be put down to the particular difficulties that Xiongnu, Tabgach, Jurchen, Mongol or Manchu would have encountered as they tried to get by in Chinese.* There is the interesting fact that Mandarin Chinese can distinguish women, 'we (excluding you)', from zsnmen, 'we (including you)', just as Mongol and Manchu do; this is an innovation since Middle Chinese. And perhaps one can point to the absence of consonant cl.u.s.ters in modern Chinese, some of which were allowed in Middle Chinese. For example, sniwr, 'appease', and t'nwsr, 'secure', have become sui and tuo. Altaic languages cannot abide more than one consonant at the beginning of a syllable.

There are in fact a few written relics of the kind of Chinese that was spoken in one of the intermediate periods before the invaders were absorbed. The thirteenth-century Chinese translation of The Secret History of the Mongols is full of Altaic patterns such as postpositions instead of prepositions, verbs following the object, and existential verbs at the end of the sentence, all weird in Chinese, whose basic word order is much more like English: Da-fan nyu-hai-er sheng liao lao zai jia-li de.

Generally daughter born past always stay home-at particle.

li wu.

reason is-not.

There is no reason why a daughter, born to you, should always stay at home.

And there is copious evidence for mixtures of Manchu and Mandarin in the zi-di-shu, 'Son's Books', which are a written record of the narrative entertainment the Manchu enjoyed in their early days in Beijing (1736-96), though they are written more with Chinese word order scattered with Manchu vocabulary.

In northern dialects there is still a tendency for direct objects to occur rather often before the verb, and for than-phrases to occur before comparative adjectives, features that might be attributed to Altaic influence. But in general this mixed style of Chinese did not establish itself.17 Later generations of invader families picked up Chinese naturally from their Chinese mothers, nurses and schoolmasters; probably the Altaic patterns were just too far opposed to Chinese for any compromise to develop. This is typical enough of Chinese linguistic relations: in general, there are not many loan words in Chinese borrowed from other languages in any direction, and certainly no structural influences; du, 'calf, does seem to have come from Altaic, characteristically enough since its peoples lived by stockbreeding (cf. Mongol tuul, Manchu tuksan, Evenki tukucn, all meaning 'calf), but the many Mongol words that are found in the drama of the Yuan dynasty have since been lost again.18

Beyond the southern sea.

Although Chinese has spent its three and a half millennia almost wholly confined to East Asia, it did put out some feelers across the sea to its south. In the last thousand years, this led to some permanent residence of Chinese abroad; in the last two hundred, partly as a reaction to-or exploitation of-European settlement, serious overseas communities have grown up, which may be significant in the future spread of the language.

The earliest inklings of Chinese in Nan-yang, 'the Southern Ocean', as the Chinese called the sh.o.r.es of the South China Sea, are visits of merchants to Tongking (northern Vietnam) in the third century BC.19 They were followed up in 111 BC by troops, and China annexed Tongking, along with Nan-yue* (modern Guangxi and Guangdong). China was to hold Tongking for over a thousand years, in fact until AD 938, despite sporadic and increasing resistance. China attempted to a.s.similate it culturally, with Chinese cla.s.sics for the local elite, compet.i.tive examinations for administrators, and official use of wenyan. There was Chinese immigration, and some married into Vietnam's princely families, providing many later leaders. Mahayana Buddhism, introduced under the Tang dynasty, became the majority religion.20 Despite all this, the Chinese language did not spread permanently to this part of the world.

Somewhat later than the advance into Tongking, Chinese proceeded farther south, though apparently with instincts more scholarly than materialist. In the third century AD, two Chinese envoys, Kang Tai and Ju Ying, wrote a report on the foundation of Funan (in modern Cambodia).21 There is little more to be said of it, or what the Chinese were doing there; but the route via ri Vijaya (in Sumatra) to India became quite well travelled by China's Buddhist scholars a little later, in the fifth to eighth centuries. (See Chapter 5, 'Outsiders' views', p. 192.) After the eighth century, trade comes to the fore as a motive, but the links seem to have been maintained by foreign merchants, Arabs, Persians and Indians, and it is only in the eleventh century that we find the first reports of capital-raising by Chinese merchants to finance their own expeditions. This was under the Song dynasty, which actively backed the traders. Thereafter government support for overseas expansion wavered, the Mongol Yuan staunchly in favour, even making a failed endeavour to invade Java in 1293, the Ming who succeeded in 1368 preferring isolation: private trade was banned, and all contacts had to be made through diplomatic channels. There was a brief resurgence during the famous global voyages of Admiral Zheng-He (in the period 1405-33); but after that episode resident Chinese merchants had, for a time, to go underground.

Most of the Chinese who had taken to this life came from Fujian, with a smaller contingent from Guangdong, a fact which is explicitly recorded in a fifteenth-century report, Yingyai Shenglsn: The Overall Survey of the Ocean's Sh.o.r.es, by Ma Huan, one of the sailors with Zheng-He. Ma writes, of two states in Java, 'Many people from Guangdong and Zhangzhou are staying there,' and he mentions many other exiles from Fujian elsewhere in the island.22 The truth of this stands out very clearly in the predominance of Min, Hakka and Yue, south-eastern dialects, in the speech of overseas Chinese to this day.*