Kingdom of Kush

The land of the bow

Drafted from Welsby, The Kingdom of Kush: The Napatan and Meroitic Empires (1996)formation

Egypt's name for the country upstream of the first cataract was Ta-Seti, the land of the bow. Egypt's word for gold was nub, and Nubia is named from it.

Those two words are most of the relationship. For three thousand years, Egypt went south for gold and came back with archers, and periodically tried to hold the country directly — under the New Kingdom, Nubia was governed by a viceroy as far as the fourth cataract, because Nubia had gold, and it slipped away when the New Kingdom's grip failed.

When the New Kingdom's grip failed, what grew in the space was a Nubian kingdom that in about 727 BC came down the river and took Egypt.

The kingdom at Napata

What grew there is called the kingdom of Kush, ruled from Napata under the flat sandstone bluff of Gebel Barkal, which Egyptian religion had held for eight hundred years to be a dwelling of Amun. The Kushite kings took that over entirely: Amun of Napata was their god, they built him a temple under the mountain, and they buried their kings in pyramids nearby.

A Nubian dynasty adopting the gods, the script and the funerary architecture of the empire that had ruled and taxed them is not obviously what one would predict. What it produced was a line of kings who could argue, when they went north, that they were more properly Egyptian than the Egyptians.

Piye

Drafted from Welsby, The Kingdom of Kush: The Napatan and Meroitic Empires (1996)expansion

The campaign is on a granite stele from Gebel Barkal, and it is one of the longest and most detailed royal narratives from anywhere in the ancient Nile valley, at a length Egypt's own pharaohs did not attempt.

Piye came north, reduced the delta princes one at a time, and received their submission. The text is unusually interested in two things. It is pious to the point of piety being the argument: Piye stops to make offerings, insists on ritual correctness, and presents the campaign as restoring proper worship that the delta had let lapse. And it is furious about horses. At Hermopolis, having taken the city, he goes to inspect the stables, finds the animals have been left hungry during the siege, and tells the defeated prince that starving his horses is worse than anything else he has done.

A Nubian dynasty then ruled Egypt for about sixty years, as its twenty-fifth.

The only time the river ran the other way

That is the fact worth stating plainly, because the general histories mostly mention it and move on.

Conquest on the Nile ran north to south for three thousand years, and once — for sixty of them — it ran south to north. Taharqa was building at Karnak and campaigning in the Levant, and he appears in the Hebrew Bible as Tirhakah king of Cush, coming out to fight the Assyrians.

The dynasty's programme in Egypt was archaism, and it was not a compensation for being foreign so much as a policy. They restored temples, deliberately revived Middle Kingdom artistic conventions and Old Kingdom texts, and took the full five-name pharaonic titulature.

A Nubian king restoring Egyptian temples and styling himself the Horus of Gold is making a specific argument: that he is not an outsider taking Egypt but an Egyptian king putting right what Egyptians had let go. It was a better claim than most conquerors have, and the delta princes he had just reduced could not answer it.

And the Assyrians

Esarhaddon took Memphis in 671 and Ashurbanipal sacked Thebes in 663, and the twenty-fifth dynasty withdrew to Nubia and did not come back.

Esarhaddon's taking of Egypt in 671 was the first time any Mesopotamian power had managed it, and the Assyrian accounts of the campaign do not trouble to say who was holding Egypt at the time. It was a Nubian dynasty, and Assyria fought two Kushite pharaohs for a decade over the richest country in the ancient world.

What happened next is the more interesting half. A dynasty evicted from the richest country in the ancient world went home, kept the titles, moved its capital further from the frontier, stopped writing in Egyptian, and lasted another thousand years.

My horses have suffered hunger

Drafted from Breasted, Ancient Records of Egypt, Volume IV: The Twentieth to the Twenty-Sixth Dynasties (1906)expansion

The Kushite conquest of Egypt is known mainly from one object: a granite slab about six feet high, found in the temple at Gebel Barkal in 1862, carrying the longest historical inscription any Egyptian-language monument bears. Piye set it up at Napata to record what he had done in the north, and Breasted, who translated it, called it the most instructive surviving document on the internal condition of Egypt in a period without a central power.

It is also, unusually for a royal inscription, a text about piety, protocol and horses.

The orders to the army

Piye sent his troops north and gave them instructions, which the stela records before any fighting.

The military part is what a commander says: if the enemy's infantry and chariotry are in another city, wait for his army and fight as he says; if his allies are elsewhere, hurry to them and force battle on them first. Yoke the war horses, the best of thy stable; draw up the line of battle. Thou knowest that Amon is the god who has sent us.

Then the part nobody else writes down:

When ye arrive at Thebes, before Karnak, ye shall enter into the water, ye shall bathe in the river, ye shall dress in fine linen, unstring the bow, loosen the arrow. Let not the chief boast as a mighty man; there is no strength to the mighty without him. He maketh the weak-armed into the strong-armed, so that multitudes flee from the feeble, and one alone taketh a thousand men. Sprinkle yourselves with the water of his altars, sniff the ground before him.

An army on campaign is ordered to stop at the great temple of the god, disarm, wash, change into clean linen and lie down on the floor in front of him before going on to fight. That is the Kushite claim in a paragraph: not that a foreign power was taking Egypt, but that the proper servants of Amun were coming north to put right a country that had stopped serving him correctly.

The stables at Hermopolis

The most quoted passage in the inscription comes after the surrender of Nimlot, prince of Hermopolis, who had broken his allegiance and stood a siege.

Piye took the city. He refused to look at the women sent to intercede with him. Then:

His majesty proceeded to the stable of the horses and the quarters of the foals. When he saw that they had suffered hunger, he said: "I swear, as Re loves me, and as my nostrils are rejuvenated with life, it is more grievous in my heart that my horses have suffered hunger, than any evil deed that thou hast done, in the prosecution of thy desire."

A king who has just reduced a rebel city announces on the record that starving the horses during the siege was worse than the rebellion. Nothing else in Egyptian royal literature sounds like it, and it is the reason Piye is the one Kushite pharaoh with a personality attached.

The four kings and the fish

The stela ends with an image of submission that is also a statement about who these people were.

Everything had fallen; there was not a nome closed against him; the west, the east and the islands in the midst were on their bellies in fear of him. In the morning four rulers came to do homage, wearing their uraei — and the princes of the north, coming to behold the beauty of his majesty, had legs as the legs of women.

Then the last line of the reception:

They entered not into the king's-house, because they were unclean and eaters of fish; which is an abomination for the palace. Lo, King Nimlot, he entered into the king's-house, because he was pure, and he ate not fish. There stood three upon their feet, but only one entered the king's-house.

Three kings of Egypt were left standing in the courtyard by a Nubian because their diet made them ritually unfit to come in.

What the document is doing

Piye then loaded his ships with silver, gold, copper, clothing and every product of the north, sailed upstream, and went home. He did not stay in Egypt, did not rule from it, and did not put a governor over it; the Delta princes kept their titles.

That is what the inscription is for. It was cut at Napata, in Kush, for a Kushite audience, and what it says to them is that their king had gone north, found the country in the wrong hands and the temples neglected, corrected it with the correct rituals and returned. The taking of Egypt is presented as an act of observance, and the only thing in the whole document the king loses his temper about is a stable full of hungry horses.

Meroë

Drafted from Welsby, The Kingdom of Kush: The Napatan and Meroitic Empires (1996)peak

Around 300 BC the royal burials move from the Napata region three hundred kilometres upstream to Meroë, in the Butana between the Nile and the Atbara.

The reasons that are usually given are practical and probably right: better rainfall, so agriculture that does not depend entirely on the flood; grazing; iron ore and the wood to smelt it; and distance from Egypt, which by then was Ptolemaic and had a standing army whose southern frontier was Kush's northern one.

The pyramids nobody counts

There are more pyramids in Sudan than in Egypt. Something over two hundred of them, at Meroë and the earlier royal cemeteries, in three main groups on the desert ridge above the town.

They are not the Egyptian kind. They are much smaller — twenty to thirty metres at most against Khufu's hundred and forty-six — and much steeper, close to seventy degrees where the Giza pyramids are fifty-two, with a chapel attached at the east face and a shaft tomb beneath. Egypt had built pyramids for about a century and then stopped, fifteen hundred years earlier. This is that abandoned technology picked up by a different people, given a steeper and smaller shape of its own, and used for six hundred years after the country that invented it had lost interest.

About forty of them were blown open in the 1830s by an Italian treasure hunter, which is why so many stand today without their upper courses.

Iron

Meroë has slag heaps. Large ones, still there, and an early excavator called the site the Birmingham of Africa, which is the kind of phrase that gets repeated for a century and then gets qualified.

The qualification is that nobody has demonstrated industrial-scale production for export, and the quantity is consistent with a substantial local industry supplying a kingdom rather than an economy built on iron. What is not in question is that Meroë was working iron on a scale unusual for the period and the region, and that the kingdom's wealth came out of the ground — iron and, from the old Nubian workings, gold.

What the kingdom looked like from outside

Herodotus knew of it and got it mostly wrong. The Ptolemies traded with it and hunted elephants on its frontier. Rome fought it once, in 23 BC: a Kushite raid into Roman Egypt, a punitive expedition, a sack of Napata, and then a negotiated peace on terms Strabo is visibly uncomfortable reporting.

The Kushites had taken a bronze head from a statue of Augustus at the frontier, and they buried it under the threshold steps of a temple at Meroë so that anybody entering trod on the emperor's face. It stayed there until 1910. It is in the British Museum now, in a case, with an explanation.

It is the most eloquent surviving comment any ancient people made about Rome, and it consists of putting a step on a head.

The kandakes

Drafted from Welsby, The Kingdom of Kush: The Napatan and Meroitic Empires (1996)peak

Nobody can say for certain how a Kushite king was chosen, and the two candidate answers are both strange.

An oracle chose the king

The Kushite royal stelae describe the succession as a decision by Amun of Napata. The candidates — the king's brothers, in the accounts that survive — are brought before the god. The god's image, carried by priests, moves towards the one he wants. The chosen man is acclaimed.

Whatever was actually happening in that room, that is what the state said was happening, and it is not a version of any of the familiar mechanisms. It is not a council conferring an office, and it is not a predecessor naming an heir. The acclamation that follows is real and is described, but it comes after the decision and does not make it.

The obvious suspicion — that the priests moved the god towards whoever had already won the argument — is unprovable and probably beside the point. A succession system is a way of making a transfer of power unarguable, and a god who cannot be cross-examined does that job well.

And the royal women

The throne passed between brothers far more often than from father to son. The king's mother held a formal position of the first importance and appears in royal inscriptions and reliefs beside her son. From about the second century BC there are queens ruling in their own right — the kandakes, a Meroitic title for the queen mother which Greek and Latin writers mistook for a personal name and passed on to English as Candace.

They were noticed abroad. Strabo's account of the war with Rome has a one-eyed queen in command. The Acts of the Apostles has an official of "Candace queen of the Ethiopians".

What all this adds up to is argued. One reading is a matrilineal rule, with eligibility running through the royal women — an arrangement well attested elsewhere in Africa, from Ghana, where al-Bakri states the rule and gives the reason, to Mali, where Ibn Battuta records it as a general principle of inheritance and disapproves, to Makuria downriver, to Asante much later, where the matriclan defined who could rule.

The other reading is a succession running among a royal kin-group of both sexes, with the queen mother as its senior member. That is a different constitutional arrangement, and it would produce exactly the same evidence.

The distinction matters more than it looks. Queen mothers with formal office, ruling queens, and the matrilineal inheritance of property are all perfectly compatible with a throne that descends in the male line. Prominence is not a rule, and Kush is the standing example of how easily the first gets read as the second.

A script nobody can read

Drafted from Welsby, The Kingdom of Kush: The Napatan and Meroitic Empires (1996)aside?What aside means

Around 200 BC the Kushites stopped writing in Egyptian and started writing in their own language, in a script they made for it.

Meroitic has twenty-three signs. There are two forms — a hieroglyphic one for monuments, using Egyptian sign-shapes, and a cursive one derived from demotic for everything else — and they map one to one. It is alphabetic with syllabic features: consonant signs carry an inherent vowel, four vowel signs modify, and a word-divider of two or three dots separates units.

Francis Llewellyn Griffith worked out the sound values in 1909, from a bilingual inscription and from Egyptian transcriptions of Meroitic names. So the signs can be read aloud.

Nobody knows what the words mean.

A hundred and sixteen years

Meroitic is not a mystery like Linear A, where the script itself is unread. The phonetics have been settled for over a century. The problem is the language: it has no established relatives.

Egyptian is Afroasiatic and has cousins. Sumerian is an isolate too, and is read easily, because the Akkadians who took over Mesopotamia went on learning it in school for two thousand years and left dictionaries. Elamite is an isolate and is readable because the Achaemenids published the same text in three languages on a cliff at Behistun.

Meroitic has no long bilingual and no successor state that kept it as a learned language. What survives is a few thousand short texts, mostly funerary, mostly formulaic. Perhaps a couple of hundred words are understood — enough to tell that a stele records somebody's mother and father and the offerings they should receive, and not enough to read a sentence with confidence.

Proposals connecting it to Nilo-Saharan languages have been made and are not settled.

What that costs

Almost everything anybody would want to know.

What can be said about how Meroë chose its kings — that an oracle was involved, that eligibility may have run through the mother — comes from Egyptian-language stelae of the earlier Napatan period and from Greek and Latin outsiders looking in. For the six hundred years after the script changes, the kingdom is writing continuously about itself, and none of it can be used.

That is a rarer situation than it sounds. Most lost civilisations are lost because the writing did not survive, or was never made, or has not been found. This one is on museum walls in Khartoum, Berlin and London: legible, transcribed, pronounceable, and meaningless.

And what it left

Three letters.

When the Christian kingdom downstream needed to write Old Nubian, six hundred years later, it took the Coptic alphabet and added three signs out of Meroitic for sounds Coptic had no letters for. Those three characters are the only part of this kingdom's writing system whose value anybody can currently be sure of, because they are used in a language that can be read.

Egyptian hieroglyphs to Meroitic to Old Nubian, on one river, across two and a half thousand years — and the middle one is the only one nobody can read.

The last inscriptions

Drafted from Welsby, The Kingdom of Kush: The Napatan and Meroitic Empires (1996)contraction

Meroë's decline has no event in it. What it has is a set of curves, all of them going the same way from about the second century AD, and none of them individually decisive.

The royal pyramids get smaller and cruder. The grave goods thin out. The inscriptions get shorter and then stop. Building at the temple sites falls off. The imported Mediterranean pottery that had been coming up the river through Roman Egypt becomes scarcer and then rare. Settlement contracts.

Nobody in the kingdom explains any of it, and nobody outside it noticed enough to write it down.

What was probably happening

Three things, all of them slow, and the sources permit only informed guesses about how they weighed against each other.

The road moved. Meroë's wealth came from being the passage between Egypt and inner Africa, and in these centuries that traffic increasingly went by the Red Sea instead — up the coast to the Egyptian ports, through Adulis and the Aksumite highlands, on ships. A kingdom that is a road stops being wealthy when the traffic finds a different road, and there is nothing it can do about it, because the advantage was never anything the kingdom made.

The ground got worse. The Butana, the grassland between the Nile and the Atbara that fed Meroë, had been cultivated and grazed for centuries and stripped of wood to smelt iron. Environmental degradation of that kind is hard to demonstrate and easy to assert, and the case here is circumstantial: a large population on marginal rainfall, an industry that consumed timber, and a settlement pattern that shrinks.

People moved in. Groups from the western deserts came into the river valley in these centuries. Whether they arrived as raiders, as settlers or as both is not recoverable. The Noba who appear in the fourth-century sources are on ground the Kushites had held.

The Roman episode, and what it showed

The one dramatic thing in the late period had happened earlier, and it is worth reading as a measure of what the kingdom could still do.

In 23 BC a Kushite force raided into Roman Egypt. Rome sent a punitive expedition, which went a long way up the river and sacked Napata. Then there was a negotiated peace, and the terms — Strabo reports them with visible discomfort — were favourable to Kush: the frontier was fixed, and the tribute Rome had demanded was remitted.

An empire that had recently annexed Egypt fought a kingdom on its southern frontier and settled on terms rather than continue. Three hundred years later, nobody had to fight that kingdom at all.

What contraction looks like without a chronicle

This is what most states' declines actually consist of, and it is almost never what gets written about them.

There is no battle to date it to, no bad king to blame, no invasion, no catastrophe. There is a stretch of river that made money for a thousand years because of where it was, and then trade moved a few hundred miles east, and the kingdom got poorer every generation until it could not pay for the things that had made it a kingdom.

Its own writing continues throughout, in a script that can be pronounced and not understood, and it may well say precisely what was happening.

350

Drafted from Welsby, The Kingdom of Kush: The Napatan and Meroitic Empires (1996)end

The royal burials at Meroë stop. The inscriptions stop. The town continues to be occupied on a reduced scale and then is not. Nobody wrote down that the kingdom had ended, and 350 is a convention covering a fourth century in which the evidence simply runs out.

The tidy story, and why it is not here

The standard account for a long time was that Aksum destroyed Meroë.

Ezana of Aksum — the king whose conversion to Christianity is dated from a change of symbol on his coinage — left an inscription describing a campaign down the Nile against a people called the Noba, in which he took towns, killed and captured, and set up a throne. It mentions Kasu, which is Kush. It was read for most of the twentieth century as the destruction of Meroë by an Ethiopian army.

It is now generally thought that Meroë had already collapsed and that Ezana came down a river where somebody else had moved into the ruins. The Noba he describes fighting are the people who had displaced the Kushites, not the Kushites; the archaeological sequence at Meroë does not show a destruction of the right date; and the campaign reads as a punitive expedition against a nuisance rather than the felling of an empire.

So nobody destroyed Meroë, and the state that was credited with it arrived afterwards.

The shape of the mistake

It is worth naming the shape, because it is a common one and it has been made about African kingdoms more than most.

Ghana is supposed to have been destroyed by an Almoravid conquest in 1076. That rests on one late source, is unsupported by the contemporary material, and shows no destruction layer in the ground. Majapahit is supposed to have fallen to Demak and to Islam in 1478; in fact the capital was taken by a rival Hindu prince and the dynasty went on for fifty more years.

In each case the ending of a kingdom nobody was writing about has been attributed to whoever turned up next with an army, because a conquest is a story and a slow loss of revenue is not. The difference here is that Ezana's inscription is genuine and the campaign happened; what is wrong is only the identity of the people he was fighting, who were the ones that had displaced the Kushites already.

What actually happened, as far as anybody can say

The Nile trade routes shifted. Aksum and the Red Sea took the traffic that had been going up and down the river, and a kingdom whose income depended on being the road between Egypt and inner Africa stopped being the road.

The Butana's environment may have degraded under long cultivation and charcoal-burning for the iron. Groups from the west moved into the river valley.

The state stopped being able to pay for itself, and stopped.

Three hundred years of nothing

Then, in the sixth century, there are Christian kingdoms on the same stretch of river — Nobatia, Makuria and Alodia — with a Byzantine liturgy, an alphabet adapted from Coptic, and three letters borrowed out of Meroitic.

What happened in between cannot be said. Three centuries on the same ground, with wealthy burials and imported goods and no name, no king-list, no date and no text. It is the largest single gap on the Nile and there is no prospect of filling it.

The three borrowed letters are the only thing that crosses it.

A singular coincidence in sound

Drafted from Burckhardt, Travels in Nubia (1819)afterlife

In 1814 John Lewis Burckhardt, travelling up the Nile as a poor Syrian trader, was collecting the names of the villages in the country of the Sheygya from people who had been there. He wrote them down in a list — Gos, Hannek el Zebeyr, Koreyr, Koray, Abramnar, Wosta, Tongazy, Koro, Kadjeba — and then, against one of them, added a note:

Merawe, a singular coincidence in sound with the ancient Meroe.

He was right that there was a coincidence and wrong about which place it pointed to; the Merowe he was told about sits by Gebel Barkal, at the other royal centre. But the observation is the thing. In 1814 a well-read European on that river knew the name Meroë from Herodotus and Strabo, heard something like it in the mouths of people living there, noted the resemblance in half a line, and went on. He had no means of pursuing it and no expectation that anything would come of it.

What he found instead

The Sheygya, the people holding that stretch of the river, were in Burckhardt's account the most powerful state north of Sennar and a genuinely formidable one.

They fought entirely on horseback, in coats of mail bought from the merchants of Suakin and Sennar, on Dongola stallions, carrying four or five lances in the left hand when charging and throwing them a great distance with much dexterity. He compares their horsemanship to the Mamluks'. They trained their horses to make violent springs with the hind legs at the gallop, and rode, like the Abyssinians, with only the great toe in the stirrup.

They also, he notes, made books. He had seen manuscripts copied at Merawe in as fine a hand as any.

Two thousand five hundred years after the kings of Kush went north out of that country with an army of archers, the people living on the same ground were the best cavalry between Egypt and Ethiopia, fighting in mail with lances, and copying books.

What was done to the pyramids

Twenty years after Burckhardt, an Italian doctor named Giuseppe Ferlini went to Meroë looking for treasure and blew the tops off about forty of the pyramids.

He found gold in one — a hoard of a queen's jewellery, which he took to Europe — and demolished the rest for nothing.

That is a considerable part of why the site looks as it does. More than two hundred pyramids stand on the ridges above the river, in three groups, and a great many of them are missing their upper courses because of one man in the 1830s, working with no record kept and nothing to compare against afterwards.

Sound without sense

The script was cracked in 1909. Griffith worked out the values of the signs, and Meroitic became a language that anybody with the alphabet can read aloud.

Nobody can understand it. A hundred and some years later, a couple of hundred words are known — enough to see that a stele names somebody's mother and father and lists the offerings owed — and the language has no demonstrated relatives, no long bilingual, and no successor who kept it as a learned tongue.

So the kingdom that ruled Egypt, fought Assyria and Rome, buried two hundred kings under pyramids and wrote continuously about itself for six hundred years is recoverable in outline from its neighbours' documents, and its own voice is a sound with no meaning attached.