The Arabs going into battle for Islam can be seen in fact as an alloy of three very different preceding traditions among their fellow-speakers of Semitic languages: the abstract theology of the Jews, the embracing inclusiveness of the Aramaic Christians, and the military momentum of the a.s.syrians. Indeed, if one includes their propensity for long-distance navigation and speculative trading, they can also be ranged with the Phoenicians.

But there is one thing in the cultural background which does unite all the Semites, of whatever religion or desired level of opulence. However successful their cities, however developed their religions and philosophies, they never escaped the memory that they had all arisen from desert nomads. Arabic was the language of nomads, and Islam was founded by nomad aggression from Arabia. Aramaic penetrated the a.s.syrian and Babylonian empires, and so became established, through nomads spreading from Aram. The Hebrews and Phoenicians developed their cities and their cultures when Habiru nomads had finally settled down in the land of Canaan; explicitly, the Torah talks of the children of Israel wandering through the wilderness of Sinai for forty years. And the Akkadians might never have taken over from Sumer without the incursions of those little-known nomads of the west, the Amorites. Ultimately, surely, it must have been nomads who brought the Semitic languages in prehistoric times out of Africa and into the Fertile Crescent.

Nomads may be hard to find in the modern Semitic world. But aspects of nomadism are still central to the unsolved problems of the Arabs: the home-lessness of the Palestinians, the moral queasiness about the unearned riches welling up from the desert wastes of Arabia, the wild men of al-Qa'eda in self-imposed exile while they plan destruction for the iniquitous cities. In all this, speakers of Arabic are very true to their tradition. Indeed, the histories of Akkadian, Phoenician, Aramaic and Arabic are a five-thousand-year demonstration of the benefits of the desert-as a place to come in from.

* The family is named after Noah's second son, Shem, introduced in Genesis ix.18, and the linguistic use goes back to A. L. Schloezer, writing in 1781. He drew his inspiration from the fact that many of the peoples named as the descendants of Shem in Genesis x.21-31 spoke languages of this family, notably Hebrew (coming via Arphaxad), a.s.shur and Aram. But the term is not well chosen: Shem also had among his sons Elam and Lud, the patriarchs for Elamite and Lydian, which are quite unrelated languages; and Canaan (first of the Sidonians, as well as Amorites and Arwadites) and Nimrod (first of the Babylonians and Akkadians) are given as descendants of Ham, though their languages are in fact closely related to Hebrew, a.s.syrian and Aramaic.

*The first Semitic names (in fact from Akkadian) appear even earlier, in Sumerian doc.u.ments c.2800 BC (Caplice 1988: 3).

p.r.o.nounce s as English 'sh', h as the sound for blowing on gla.s.ses to mist them, as English 'th' in thin, ' as the clearing of a throat, a as a long 'a' as in father, and e as the long 'e' in Beethoven.

*The Greeks, on the scene too late to know any of these early origins, called the place Mesopotamia, 'Mid-River-land', emphasising the framing role of the rivers Euphrates and Tigris, Greek versions of the names Purattu and Idiqlat. But in this early period the Euphrates is much more central, flowing through Babylon and Ur, and watering the lands of both Akkad and Sumer. The Tigris, farther to the east, grows in importance with the rise of a.s.syria. 'a.s.syria' is the Greek attempt to name a.s.shur.

*The name Hitt.i.te (from the Hebrew itti) comes from their power centre in the land of Hatti, where the natives spoke a quite unrelated language, Hattic. The Hitt.i.tes in fact called their language Nesian (nesili), after their city of Nesas (or Kanesh, modern Kultepe, in south-eastern Turkey) but the biblical misnomer 'Hitt.i.te' has stuck.

Croesus, the proverbially rich last king of Lydia, fell to Cyrus the Persian in 547 BC. Linguistically, this was the ultimate death rattle of Hitt.i.te power.

*This is his name in Hebrew. His real name was Tukulti-apil-Esharra, meaning 'my trust is in the son of Esharra', namely the a.s.syrian G.o.d a.s.shur. The Mushki are equated b55555tty Igor Diakonov with the Mysians, Thracian settlers in western Anatolia, and also the Armenians, named Sa-mekhi by the Georgians. The Bible also speaks of Meshech as a foreign people.

*The Phoinikes, especially the Sidonians, are renowned in the Iliad for fine weaving and metalwork, and in the Odyssey as travelling merchants.

*There are 6 million tons of ancient slag, covering 3/4 of a square kilometre, at the silver mines of Rio Tinto, near Huelva (probably the site of Tartessos, believed to be the same as Tarshish in Hebrew). Despite this ma.s.sive activity, extending over centuries, archaeological evidence tends to show that Phoenician settlements in Spain were commercial enclaves rather than towns (Markoe 2000: 182-6).

Another ideographic system, invented at the other end of Asia, had similar effects. The j.a.panese, Korean and Vietnamese languages, all of which became literate through the use of Chinese characters, have sustained major linguistic (and cultural) borrowings from Chinese which are by and large still present today.

*The one exception is Bactrian, later to become the language of the Kushana empire (first to second centuries AD), written in the Greek alphabet. This shows the lasting cultural influence of the independent Greek dynasties in the far east, whom the Kushana supplanted.

* And this is precisely what we do with our number symbols, whether Arabic or Roman.

*The Amorites did not have their own literate tradition, but their language can be partially reconstructed when their names are quoted in other languages, usually Sumerian. This provides a link with the later western Semitic languages, such as Ugaritic, Phoenician and Hebrew, which do not show up in the written record for another five hundred years or more. Since there was a tendency to a.s.sign names that are full sentences, they give a fuller picture of the language than might have been expected: Aya-dadu, 'Where is Daddu?', Sub-addu, 'Return, Addu!', Yasub-'ilu, 'G.o.d returns', Samsu-'ilu-na, 'The Sun is our G.o.d.'

*This, after all, is exactly what happened to the various tribes of Angles, Saxons, Jutes and Danes who settled along with Frisians in Britain in the first millennium AD. Middle English, closest to Frisian, was the result.

* In Babylon some diehards were still writing Akkadian on clay six centuries later.

*As it happens, the last we hear about Akkadian is from a Syrian novelist writing in Greek in the second century AD: Iamblikhos (whose strange name is evidently Aramaic or Arabic, ya-mlik, 'may he rule') said he had learnt 'Babylonian' from his Babylonian tutor, a man 'learned in the wisdom of the barbarians'. (The third-hand source for this can be traced from Stephens and Winkler 1995:181.) Hebrew and Phoenician include some of the complexities of their grammar in their spelling: most of the stop consonants are p.r.o.nounced as fricatives in the middle of a word. In our romanisation, we represent this with an under- or overline: thus , , g, , p, are p.r.o.nounced v, th (as in then), gh (a gargling sound), ch (as in loch),f, th (as in thin). Dots under s, t and d in Phoenician, Hebrew and Arabic mean that they are p.r.o.nounced 'emphatically', giving them a somewhat dull, throaty quality.

*Agreement has never been reached on why the Greeks picked on phoinikes as their word for these roaming Semitic traders. Literally it means 'date palms' (or indeed the mythical phoenix birds), but the a.s.sociation with phoinos or phoinios, 'gory, blood red', was always kept in mind, since the Phoenicians were the purveyors par excellence of purple-dyed fabrics, and farmed the dye's raw material, murex sh.e.l.lfish, on an industrial scale. The a.s.sociation of the colour with this part of the world goes beyond Greek: the Akkadian word for 'purple' was kinau, derived from the place name Kina (n)I, 'Canaan' (Black et al. 2000: s.v.). Although the Hebrews lived in Canaan themselves, they used the word kna 'aniy, as Greeks did phoinix, to refer indifferently to a Phoenician or a merchant; and this seems to be what the Phoenicians called themselves.

* This is known as the TaNaK(for Torah Ni'im wa-Ksuim, 'Law, Prophets and Scriptures'). But besides that there is the commentary on the Torah known as the Mishnah (200 BC-AD 200), the supplement known as Tosephta (AD 300), and a verse-by-verse commentary on the TaNaK, known as the Midrash (AD 200-600). These show that Hebrew continued to be written as well as read.

* It was written on clay tablets; this is why it survived. But it was incised in an alphabet based on cuneiform, so graphically too it throws an interesting light on the Phoenicians, until then reputed to have been the first to use an alphabet. The simpler shapes of the Phoenician letters are due to their usually being written with ink on papyrus, rather than stamped with an angled stylus on clay.

El is simply the Semitic root for 'G.o.d', seen also in Hebrew elohim, one of the two words for G.o.d in Genesis, and Arabic Al-lah, literally "The G.o.d'.

* The poem continues with listings of characteristic products for all the major client nations: Tarshish (metals); Greece, Tubal, Meshech (slaves, bronze working); Beth Togarmah (equines); Rhodes (ivory and ebony); Aram (turquoise, fine cloth, coral and rubies); Judah and Israel (wheat, honey, oil and balm); Damascus (wine, wool); Danites, Greeks of Uzal (wrought iron, ca.s.sia, calamus); Dedan (saddle blankets); Arabia, Kedah (sheep and goats); Sheba, Raamah (spices, gems, gold); Haran, Canneh, Eden, a.s.shur, Kilmad (clothes, fabric, knotted rugs).

* Elimam (1977) suggests that the Punic story had a happier ending, and that Punic is still alive today, as the ancestor of Maghrebi 'Arabic' (maghreb is Arabic for 'west'). It is certainly true that this Semitic language, usually characterised as a dialect of Arabic, diverges strongly from the cla.s.sic language of the Koran; but this is true of all the Arab vernaculars. Where Punic did survive after the Roman period, it would very likely have made a significant contribution to Maghrebi. Unfortunately, the restricted evidence of what Punic was really like makes it hard to know to what extent this happened. Elimam himself suggests, on the basis of the longest Punic speech in Poenulus (ten lines, eighty-two words), that Punic has 62 per cent in common with Maghrebi, and a further 18 per cent has undergone some semantic evolution.

* The return is recorded at length in the books Ezra and Nehemiah of the Bible. They are in Hebrew, though much of the correspondence with the government is given in Aramaic (Ezra iv.8-vi. 18 and vii. 12-26). It is an amazing demonstration of the preservative power of a tradition consciously maintained that now, after an absence of two and a half thousand years, Hebrew is again the vernacular on the streets of Jerusalem.

* This momentum was known anyway, since India's original scripts, Kharoshthi and Brahmi, are both derived from Aramaic writing. Since Brahmi in turn is the origin of every other alphabet in South and South-East Asia, the Persian king Darius was in effect setting the writing systems of most of Asia for the next 2500 years when he chose Aramaic as the standard language for his empire.

The French historian Fernand Braudel can hardly forgive him for missing his opportunity to go west instead, and so take over the Mediterranean (Braudel 2001: 277-84, 'Alexander's mistake').

* Matthew xxvii.74. Sawyer (1999: 84) quotes a lot of evidence for att.i.tudes to Galilean.

The name Urfa is probably derived from Hurri (cf. the Greek name of its surrounding province, Orrhoene), with a history going back to the Mitannian period.

The Muslims in themselves were never a physical threat to the Aramaic speakers, since they saw them everywhere as millet, or distinct nationalities, separate but respected. But there was a tendency for Aramaic speakers everywhere to give up everyday use of the language in favour of Arabic.

* Christians were not the only people to go on speaking Aramaic, though they have lasted longest. The Gnostic sect of southern Mesopotamia also spoke another dialect of Aramaic, known as Mandate or Mandaean, at least until the eighth century. And for a few centuries AD, the Jews of Babylonia and Persia also continued, producing most notably the vast Babylonian Talmud. Both these communities were prolific in writing literature.

There is a considerable modern diaspora too, to the major cities of Israel, Lebanon, Syria, Iraq, Iran and Turkey. Many are said to have emigrated to Armenia and Georgia after the Russo-Persian war of 1827; and a sizeable number have gathered in the USA. The use of the Internet in binding them together is examined in McClure (2001). She quotes estimates of worldwide numbers around 1-3 million.

Its name is derived from Arabic qibt, 'Egyptian', a shortening of the Greek Aigyptios.

* This caused some philological problems, since Muhammad's dialect of Arabic was slightly nonstandard: it lacked the glottal stop ', known as hamza (the stop heard in place of the tt of the London p.r.o.nunciation of 'bitter'), had lost the -n ending of the nominative, and had turned the -t ending of feminine nouns into -h. The scholars wanted to retain the text exactly as written, but recite it according to the rules of standard Arabic. As a result, all these consonants of Arabic had to be inserted in the written text with special accent marks, as if they were vowels. These marks are all now standard in Arabic spelling.

* Arabic script turned out to be much more universally attractive than its language, and has been taken up wherever Islam was accepted. This has happened despite its functional weaknesses, with no marking of vowels or tones, and a need for elaborate accents even to distinguish all the consonants. Nevertheless, compromises have been found, and it has been applied to languages as various and as unrelated as Persian, Turkish, Kashmiri, Berber, Uighur, Somali, Hausa, Swahili and Malay, as well as Spanish and Serbo-Croat. It must owe this success to the fact that literacy in Muslim countries finds its alpha and its omega in the sacred text of the Qur'an in Arabic script; so any other writing system can only be an extra complication.

* It may be worth noting that the j in this word is p.r.o.nounced as in judge.

But one is left wondering why the linguistic approach of the Germans, notably the Visigoths, had been so different, when in 410 they likewise took over control of the neighbouring higher civilisation, the Roman empire, only to cast themselves, almost immediately, as its protectors. But in the European case, there was no third language playing the role of Persian: Latin was still the only language of temporal power, as well as the language of the Roman Church.

* Hausa, centred on Kano in northern Nigeria, is more of a problem for the constraint. It has certain features that are reminiscent of Arabic, e.g. two genders, masculine and feminine, the latter marked with -a (cf. Arabic -ah); and the absence of p-as in Arabic, it usually replaces p in loan-words with f. Moreover, its predominantly Muslim speakers have filled it with loan words from Arabic, including most of the numerals above ten, and the days of the week, and even some productive prefixes, such as ma-. ('School' is makaranta, formed from karanta, 'read', itself related to the word Qur 'an. In Arabic, 'school' is maktab, or madrasa, with the same prefix, but ktb, 'write', or drs, 'lesson', as the stem.) But it also has many features much more typical of its African neighbours, e.g. three contrasting tones, and explosive consonants. It may be that its own utility as a lingua franca, widely used in West Africa and not just among Muslims, has acted to maintain its independence.

* They even plied, especially in the early centuries, to South-East Asia and China. Abu Zayd of Siraf wrote that sea traffic in 851 was regular because of a great exchange of merchants between Iraq and markets in India and China: in fact, he said, a trade colony of 120,000 Westerners (including Muslims, Jews, Christians and Zoroastrians) were ma.s.sacred in Canton in 878 (Hourani 1995: 76-7).

* Zanzibar is in fact an Arabised form of Persian: Zangi-bar, 'blacks' land'.

* In Turkish spelling (introduced by Ataturk in 1928-9), c is [d] (j in judge), c is [c] (ch in church), i is i p.r.o.nounced with the tongue root drawn back (as in Scots kirk), and is either a gargling sound (like Greek gamma or Arabic ghain) or just a lengthening of the preceding vowel; o and u are as in German.

4.

Triumphs of Fertility: Egyptian and Chinese.

* In the interests of readability and realism, Egyptian words are given according to the reconstruction of Loprieno 1995 for early Middle Egyptian, with the addition that vowels that he believes indiscernible are represented here by . R is the French (or Israeli) uvular r, and j is p.r.o.nounced as in German, like English y in yet. is a deeper h, as when huffing on a pair of gla.s.ses; and is like ch in 'loch' or 'Bach'.' is ayn, notorious from Semitic languages, the throat-clearing sound at the beginning of English 'ahem'. It should be remembered, however, that as written in hieroglyphs, Egyptian words are totally without vowels.

1.

Zi-lu said, 'If the Prince of Wei were awaiting you, Sir, to take control of his administration, what would be the Master's priority?'

'The one thing needed is the correction of names!' the Master replied.

'Are you as wide of the mark as that, Sir?' said Zi-lu. 'Why this correcting?'

'How uncultivated you are, Yu!' responded the Master. 'A wise man, in regard to what he does not understand, maintains an att.i.tude of reserve. If names are not correct then statements do not accord with facts. And when statements and facts do not accord, then business cannot be properly executed. When business is not properly executed, order and harmony do not flourish. When order and harmony do not flourish, then justice becomes arbitrary. And when justice becomes arbitrary, people do not know how to move hand or foot. Hence whatever a wise man states he can always define, and what he so defines he can always carry into practice; for the wise man will on no account have anything remiss in his definitions.'

Confucius, Lunyu (a.n.a.lects), xiii:3 (Chinese, early fifth century BC)2 Two ancient languages, widely distant in their lands and their eras, are yet strangely similar in their careers. In their attributes they are unmatched, except by each other.

Egyptian and Chinese are both vehicles of single cultural traditions of immense prestige. For each, the role as universal language was uncontested in their homeland. By the dawn of their recorded histories they were already established over the central zone of the lands where they were to be spoken. Each maintained this position of solitary and basically unchanging dominance for an awesome period of over three thousand years, or more than 120 generations. Yet, in each case, despite the fame and prestige of the culture among neighbours, who were often dominated politically by these powers, the languages never a.s.sumed any role as lingua franca beyond the territory that they considered their homeland.

Another parallel concerns their scripts. Each language originated its own unique system of writing, based on pictograms in a particular style; and each of these scripts early attained a form that would not change. Each was later taken up by another people, and simplified to yield the basis for a phonetic writing system: Egyptian hieroglyphs were the starting point for the Phoenicians' alphabet, and the j.a.panese drew their kana syllabary from Chinese characters. But in each case the original language culture disregarded the innovation, and maintained its ancient system essentially unchanged, despite the vast overhead this entailed in continuing lengthy scribal education.

Their careers are parallel. For us, their main interest lies in considering how a language can achieve steady state, a kind of h.o.m.oeostasis where it appears to absorb any perturbation that might affect it. This steadiness is particularly interesting in the cases of Egypt and China, since the languages have not simply survived in isolation, but can be seen coping with human incursions for much of their history, and occupy s.p.a.ces large enough to pose difficulties for a unitary government.

Another aspect of this puzzling unity, especially in the case of Chinese, is the strange coherence of the language itself. Certainly Chinese has dialects, and they are different enough often to be considered distinct languages. But this famous fact is less interesting than a less noted one: over 70 per cent of Chinese speakers speak a single variety, known as Mandarin or Putonghua,* and this, the official language of the Chinese state, is spoken in more than 75 per cent of the country's area. It has some local accents but essentially no internal variation. Since both the Chinese population and surface area are vast, the degree of uniformity so achieved is unparalleled in any other known language. We need to consider how it could have come about.

The two also have some direct implications for the modern world.

Egyptian, after all, did ultimately succ.u.mb to the incursions of its neighbours, carried out with steadily increasing permanence by waves of a.s.syrians, Persians, Greeks, Romans and Arabs, and now survives, if at all, as Coptic, in the liturgy of what was a foreign religion, Christianity. There is evidence here of what it takes to obliterate a seemingly eternal tradition in the land of its birth. How is immortality undone?

By contrast, Chinese, for all the political reverses and atrocities its people have suffered at the hands of heartless foreigners in the last two centuries, has never been stronger than it is today. Its speakers make up one sixth of the world's population, and it has three native speakers for every one of English. Nevertheless, over 99 per cent of them live in China, so it cannot be considered a world language-unless China is your world. Those who speak it often call it zhong guo hua, 'centre realm speech': in that at least Chinese ethnocentrism is undiminished. There is still time to consider those forces that have kept the Chinese realm so firmly, and compactly, centred on its traditional homeland: will they still prevail in the modern world?

Careers in parallel.

The remarkable similarity of the careers of the Egyptian and Chinese languages can first be displayed in the form of two chronological charts. Foreign incursions and cultural influences are marked in boldface type.

Both Egyptian and Chinese history are made up of long periods of stable unitary government, interspersed with intervals of civil unrest, or at least disunity, when there were competing dynasties in different parts of the countries. Egypt has three such periods of stable self-government, the Archaic + Old, the Middle and the New Kingdoms, followed by a Late Period, when foreign rule was the norm rather than the exception. China also has three long periods of indigenous rule, the feudal age of the Shang and Zhou dynasties, the First Empire of the Qin and Han dynasties, and the Second Empire of the Sui, Tang and Song, which then were overlaid by a succession of partial or total alien invasions.

Both civilisations were formed originally along the valley of a single river, the Nile* and the Huang-he ('Yellow River') respectively, although China expanded to take in the next great river valley to the south, the Yangtze Kiang. And both civilisations demonstrated that, although they were not capable of defending their borders indefinitely, successful invaders stood to be absorbed in the long term. The linguistic a.n.a.logue of this was that no foreign invaders imposed their language on the population, nor indeed (until the Persians and then the Greeks took Egypt) managed to retain their own language for more than a generation after mastering the country.

These are both tales of solid growth and heroic maintenance, rather than ma.s.sive spread. This chapter first sketches each language's history, particularly noting the encounters with languages spoken by foreign intruders: these often came to stay, but tended not to supplant their hosts. Armed with the facts, we can then consider what might be the secrets of such language stability.

Language along the Nile.

Be a craftsman in speech, thou mayest be strong, the tongue is a sword to a man, and speech is more valorous than any fighting ...

Instruction for King Medicare, line 32 (Egyptian, mid-twentieth century BC)3 The origin of the Egyptian language must be found close at hand, in the Afro-Asiatic or Hamito-Semitic family whose descendant languages cover most of North Africa and the neighbouring areas of the Fertile Crescent (from Palestine round to Iraq) and Arabia. Egyptian has no close relatives in this large family, but its family origins do account for some of its characteristic features, mundane things such as the fact that feminine nouns end in -t.

A stately progress.

Archaeology shows that the Egyptian state was established first in the late fourth millennium BC, in the region surrounding the great salient of the Nile which was later dominated by the city of Wast (known to the Greeks as Thebes), hence in southern or 'Upper' Egypt. It is apparent that Egyptian was already the language spoken, since there are legible hieroglyphic captions on labels and pots in the royal cemetery in this area, at Abydos, from the early third millennium. In fact pre-dynastic sites, of this so-called Nagada culture, have been discovered along the whole length of the Nile from Aswan to the delta and including the Faiyum, showing that the whole area of ancient Egypt was already occupied. Since the surrounding desert remained uninhabitable, the kingdom of Egypt was always a ribbon development along the Nile. Traditionally, its history begins when King Menes unified the Upper and Lower lands, and set up his capital at Min Nafr (Memphis) in Lower Egypt.

This achievement remained a matter of legend rather than history, since the king's name cannot be identified with any of the hieroglyphic evidence, and there is no written evidence of separate kings in the north and south. Nevertheless, there was a tradition of differently shaped and coloured crowns for the two kingdoms, unified formally in the historical crown of the pharaoh (in a way reminiscent of the composite character of the Union Jack). And the name by which the Egyptians always knew their own country was TaRwj, 'the pair of lands'.

Thereafter, Egyptian has no history, in that it had achieved its historic domain, the Nile valley from the first cataract to the sea. Although Egyptian power would expand periodically and withdraw again, up the Nile into Kush and north-eastward over Palestine and Syria, the language did not spread with it. For almost four thousand years its range stayed the same.

Nevertheless, spoken Egyptian did change phonetically and syntactically over this time. The cla.s.sical language of Egyptian literature was refined and established in the third millennium. Known as 'Middle Egyptian', its use was maintained in writing as far as possible until the end of Egypt's civilisation, above all in formal and ritual texts. But evidently the language gradually changed on the lips of its speakers. Among a host of finer periods, linguists distinguish broadly an earlier era (3000-1300 BC) from a later one (1300 BC-AD 1500). From the middle of the second millennium, it is clear that the spoken language had moved on significantly.

On the simplest level, the sounds of the language change: r and the feminine ending t are lost at the end of words, and (ch in church) and d (j in judge) are simplified away, replaced by simple t and d. But there are structural changes too. They are reminiscent of the way in which Italian came to differ from Latin, or Middle English from Anglo-Saxon. In the older period, Egyptian had been highly inflected, with a set of endings for number and gender; it had had no definite or indefinite article (corresponding to English the or a); and the characteristic word order had the verb first in the sentence, followed by subject and then object. In the later period the noun endings tend to be lost, but articles come into play, expressing the distinctions in a different way. The verb system becomes more dependent on auxiliaries, and so less highly inflected. Furthermore, the subject now tends to come first in the sentence (as it does in modern English).

Take a single example, the Egyptian for 'Hallowed be thy name'. This changed from uw'obu rin-k. tomare pe-k-ran ouop.

shall-be-purified name-your let-do the-your-name be-pure.

The pieces of cla.s.sical Egyptian are still basically there, but now put together quite differently.

Charmingly, the first glimpse of this later language to appear in the record is the more popular style of writing seen under the religious reformer Pharaoh Akhenaten; this writing reform came along with official portraits that for the first time emphasised a pharaoh's home life, with his queen Nefert.i.ti and their daughters, around 1330 BC.

Although the state religion and the decorum of official iconography were restored after his reign, the antiquated style of written expression never fully came back. Religious texts (rituals, mythology and hymns) did continue to be written in the cla.s.sical form of the language; indeed it persisted until the end of hieroglyphic writing in the fourth century AD; but popular literature, school texts and administrative doc.u.ments show that a different variant of the language was now being used generally.

The language persisted in Egypt as the main medium of daily life for another two thousand years from the time of Akhenaten.

Against this underlying continuity, the main dramatic interest was provided by contact with other languages whose speakers came to live in Egypt. There were four such languages: Libyan, Kus.h.i.te, Aramaic and Greek.

Immigrants from Libya and Kush.