The Decipherment of Meroitic | Claude Rilly, Sorbonne

I'm glad to see everyone here and welcome this evening. We have a real treat for our Isaac lecture this evening, and I'm very pleased and honored to be able to introduce our speaker, Claude Rieu, who is a very distinguished scholar. We were having a conversation earlier this week, and he is really first and foremost a philologist, a language person, a text person, but he's also an archaeologist, so that makes him dear to those of us who are on that side of the fence in this building. He really brings, I think, both of these together. Someone who has worked deeply and extensively in the languages and the literature and the texts of a very important but also not well known and understood language and language group at this point, but also has spent an enormous amount of time in the field and has done and earned his spurs, as a field archaeologist. So I think we're in for a real treat this evening to hear about his work with the Ameritic language. Just to say a little bit about his scholarship, trained in the academy in France, up through the systems, reaching his PhD in 2003 in Egyptology and Linguistics at the École Pratique des Hautes Etudes. He's since then continued on in various capacities, as I said, as the director of various projects and also on behalf of the French foreign mission in the Sudan for much of the ensuing decades. And he also completed his habilitation recently. These are on the Ameritic lexical and grammatical data linguistic position. That was his PhD thesis. And then more recently with his habilitation, the quest of meaning toward the translation of Ameritic text. His publication record is incredible and impressive. A long, long list of articles, most of them in French, so I'm not going to try to go into them in detail here. But some of the monograph-length volumes, one that I found that kind of jumped out at me when I was looking at the length of it, is the epigraphic repertoire of Ameritic, which has multiple volumes totaling more than 2,000 pages. So just to give you a sense, this is, of course, a multi-authored work. But to give you the scale and the substance of his scholarship is truly impressive. Books also on the language of the Kingdom of Meroe, books on the language and writing systems, numerous volumes, and then, as I said, many, many articles on this subject. He's truly the world's leading scholar and authority in this. I have to, just as a kind of personal side, I was also really struck by his knowledge and connection with colleagues and scholars, many of whom I had known as either a student decades ago or during my time in Toronto with colleagues there. So he's very, very widely known and engaged in the scholarship of the Sudan and of the Meroitic culture specifically. So I don't want to say more other than just to welcome Claude and that we're really privileged to have you here and we look forward to your lecture on the decipherment of Meroitic, the forgotten language of ancient Sudan. If you may join me in welcoming Claude to the platform. Thank you, Tim. Dear colleagues, dear friends, ladies and gentlemen, good evening. First of all, I'd like to thank the Institute for the Study of Ancient Culture for this invitation. It's really a great honor for me to present the results of my research to an American audience for the first time live. Moreover, the Institute has always been at the forefront of the study of ancient students from James Henry Breasted's photographic coverage of the sites of the Sudan to the salvage excavation of the Oriental Institute Nubian expedition and more recently those of the fourth cataract. For nearly 30 years now, I've been working to decipher a little-known language, Meroitic, that is the ancient tongue of Sudan, the earliest written language in sub-Saharan Africa. Here is an outline of the topics I'll be covering in this lecture. We'll start then with a short historical overview followed by the history of writing in ancient Sudan. Then the decipherment of Meroitic scripts, a chapter on ancient languages that are undeciphered or poorly known, you will be surprised, and we will look at methods for decoding the Meroitic language, be they external approaches, internal methods, and finally the famous comparative method. So the ancient kingdom of Sudan lies immediately south of Egypt along the Nile. Here the river is punctuated by six cataracts which is areas of rapid which limited river traffic unlike in Egypt. The country was called Kush by its inhabitants, a name which was taken by the ancient Egyptian alongside with the old name Toseti, that is the land of the bow, a weapon that the Kushites wielded with unrivaled dexterity. Around the middle of the third century BC, the first kingdom of Kush was created. We call it kingdom of Kerma after the modern name of the capital. For almost a thousand years, it was Egypt's rival in the Nile Valley, alternately economic partner or mortal enemy. Around 1450 BC, after a century of fighting, the pharaohs of the 18th dynasty conquered the kingdom of Kerma. They established there a colony for economic purposes. Every year the tribute of Kush was brought before pharaoh. It included all the luxury goods imported for inland Africa, such as ivory and ebony, animal skins, ostrich feathers, and above all gold, of which the land of Kush was the main source for Egypt. After the withdrawal of the Egyptians between 1900 BC, a native kingdom was reconstituted around the town of Napata, founded at the foot of this impressive tabular mountain, the Jebel Barkal, called Juwab, the pure mountain. The new state equipped itself with an efficient administration and a formidable army. They imitated the Egyptian institution and adopted a royal ideology in which the link between the ruler and god Amun played a central role. Around 723, the Kushite king Pyanke, sometimes erroneously called Piye, who already held Upper Egypt, conquered the whole country as far as the delta and established the 25th dynasty of Egypt. For almost 70 years, Kush and Egypt formed a single kingdom in the hands of Sudanese rulers, the famous black pharaohs. In 664, the last king of this Kushite dynasty, Tanwetamani, was driven out of Egypt by Assyrian invaders and had to retreat to Napata, where his successors were buried for four centuries. Around 270 BC, a new dynasty from the south ascended the throne of Kush. The rulers were now buried beneath the pyramids of the necropolis of Merui, 200 miles southwest of Napata. The kingdom of Merui was to last until the middle of the 4th century AD. It was destroyed by the invasion of tribes from western Sudan, known as the Nuba, who gave their name to the country, which from Kush became Nubia, but not before. Like Egypt, Sudan is a land of writing, although it appeared a bit later. The earliest attestation of Meroitic language dates back to the 2nd millennium before Christ. It seems to have been the main language of the kingdom of Kerma from 2100 BC, when the name Kush appeared for the first time, probably as a result of an elite replacement. A list of dozens of name of Kerma notables was written by an Egyptian scribe on the back of the Golanesh of Papyrus around 1600 BC. These names are transcribed phonetically in hieratic script and are composed of elements that are unmistakably Meroitic. However, this proto-Meroitic language did not yet have its own script, and the rare depiction of Kerma kings are not accompanied by any inscription such as this one, in which the Kushite ruler holds a bow, the emblematic weapon of Nubia. During the Egyptian colonization that followed the defeat of Kerma, the pharaohs covered Nubia with temples featuring numerous hieroglyphic texts such as here, the Temple of Samna, which cartraces of King Thutmose III. The local kinglet, under Egyptian domination, such as Hecanefer, prince of Mayan in Lower Nubia, not only adopted Egyptian names, but also used Egyptian language and script in their funerary trousseau, here on a shabti, that is a servant statuette. When, after the departure of the Egyptians, a new kingdom was formed around Napata, the Kushite king employed Egyptian scribes, notably when they ruled Egypt during the 25th dynasty, as can be seen here in the burial chamber of the last of these pharaohs, Thanuatamani. Even after they were expelled from Egypt by the Assyrians, the Napatan kings continued to use Egyptian exclusively in their official texts, such as this teller commemorating the enthronement of King Aspelza. We might compare more or less the use of Egyptian in Napata to the use of Latin in early medieval Europe, when it was the sole written language. It was only when a new dynasty of southern origin came to power around 270, creating the Kingdom of Meroë, that the local form of Egyptian demotic, known as cursive, was adapted to the Meroitic kingdom. The oldest dated inscription appears on a recently discovered sistrum, that is a music instrument, which includes the name in Egyptian script of King Arnechamani, who reigned around 220 before Christ, and on the instrument handle a very archaic inscription in Meroitic cursive is engraved. Half a century later, a second script was created, differing from the cursive only in the shape of the letters, known as hieroglyphic, and of course adapted from Egyptian hieroglyphs. It was strictly reserved for temples and royal artifacts. There was a kind of taboo on this script, unlike in Egypt where any private person could use, in fact, this script. After the fall of Meroë, Greek became the language of writing, as attested by the proclamation of King Sirko, the first known name of the Nubian chieftains who invaded the middle line. Alongside Greek, after the Christianization of Nubia in the 6th century, a new script was adapted from Coptic to record Old Nubian, the language that replaced Meroitic. It lasted until the complete conversion to Islam, that is not before the 16th century. Contrary to popular belief, the Meroitic script has been deciphered. It is the language, and the language alone, that still resists translation. Both script, cursive, and hieroglyphic were deciphered between 1907 and 1911 by the British Egyptologist Francis Cleveland Griffiths. He used text newly found in Egyptian Nubia at the sites of Araica, Chabloul, and Karanok. The publication of Karanok in 1911 presented the final results of this admirable work. What was all the more admirable in that Griffiths didn't just give the value of the sign, he described elements of grammar and translated a few dozen words enough to understand a simple epitaph of which the Karanok excavation has provided over a hundred examples. Both Meroitic scripts use the same system, an alpha syllabary or abugida, similar in principle to Indian scripts derived from Brahmi. The writing is therefore purely phonetic. The number of signs is minimal, 23 in each of the two inventories, so that sometimes some signs, you know, could have several values. For example, the vowel sign A could represent initial A or initial U. Words, or more often phrases, are separated by two or three dots. As you can see, two dots in cursive on the left and three dots in hieroglyphic on the right. The number of known texts exceeds 2,000 and increases every year due to new inscriptions discovered on Sudanese archaeological sites. The vast majority of these texts are in cursive. Hieroglyphic texts amount to less than five percent of the corpus. Look at those three sentences. You can read them. You can read them if reading means identifying the letters. Yet no one, no one person in the world, for the time being, can translate them. They belong to three unknown or little-known languages of the past. The first is in North Piscine, an ancient language of Italy. The second is Pictish, the ancient language of northern Scotland. Finally, the third is Meroitic, and it comes from one of the most obscure stele written in that language. Meroitic belongs to a category of ancient languages whose writing is known but which are incompletely deciphered. In fact, in concern, I would say, the majority of ancient languages attested by texts. So, alongside Meroitic, there is Etruscan, Pictish, and Gaulish. In the Iberian Peninsula, Lusitanian in Portugal, Iberian, Celtiberian, and Tartessian. In northern Italy, Rischian, Camunic, and Ligurian. In southern Italy, North Piscine and Mesapian. In Sicily, no less than three, Elimian, Sicanian, and Cyclonian. In Greece, Lemnian. In Asia Minor, Phrygian, Lydian, Carian, and Hattic. In Crete, Minoan, and Ethio-Cretan. In Cyprus, Ethio-Cypriot. In Turkey, Uartan. In Syria, Hurrian. And in the Persian Gulf, Kassite and Elamite. And the list goes on, you know. Knowledge of this language varies widely. Hurrian, for example, is fairly well understood, while Tartessian is completely unknown. Generally speaking, the conditions on which the deciphering of a forgotten language depends are, first and foremost, the limiting factors represented by the quantity of written material we possess. Numerous or rare, long or short texts. And secondly, the variety of genres and archaeological contexts. Lusitanian, for example, an ancient language of Portugal, is only attested by five short inscriptions, which explain why, apart from its probable belonging to the Indo-European family, and possibly to Celtic, there is little hope of making progress without new discoveries. Camunic, the language of the ancient mountain dwellers of eastern Piedmont, is known from over 240 inscriptions, which seems a decent number. But the corpus is made up entirely of graffiti engraved on rocks. Probably the least informative of all the archaeological contexts, which explains the obscurity that still covers the language. If the number, length, and variety of an inscription of a lost language are sufficient, then two main factors come into play. Firstly, are there any bilingual texts, including the language to be translated, and another language, which is known, and are they sufficient in length or number? Secondly, does the forgotten language belong to a linguistic family whose close members are unknown? Deciphering by bilingual alone is rare, as it requires a large number of bilingual texts. But that is how Sumerian, a language which has no known relative, was deciphered, because of the prodigious number of bilingual texts and even complete Sumerian Akkadian glossaries that have been discovered. Decipherment by linguistic comparison is more frequent. Akkadian itself has been deciphered from known Semitic languages such as Hebrew, Arabic, or Syriac. This method has also played an important, though not completely exclusive, role in the decipherment of Hittite and Tocharian. So Hittite in Turkey and Tocharian in Scythia. Both of them are Indo-European. On the other hand, the absence of a known related language is the main reason why the deciphering of Etruscan in Italy is so slow. It started in the 15th century, not finished. If we take these conditions into account, we see that Meroitic is in a rather favorable situation. There is a large number of texts, many of them rather long. They were first collected in this catalogue, the Repertoire d'Epigraphie Meroitique, which we published in 2000 with my master Jean Leclan. So more recently, my German colleague Jochen Halloff, the other specialist of Meroitic in the world, okay, don't laugh, has published an index of Meroitic sequences totaling almost 7,000 words, most of which, unfortunately, are of a known meaning. So several textual genres are represented in Meroitic. So we have funerary inscriptions, many of them, royal chronicles, temple inscriptions like in Egypt, object dedication, magical protection texts, ostraca accounting documents, pious graffiti, and many others. The number of texts is large and growing every year, and they are often of good length. Bilingual texts are short and rare. This one is a caption inscribed on several jars found in the pyramids of the Queen of Meroe, which mention in Demotic Irep and Kemi. That was a time where it was possible to drink wine in Sudan. And in Meroitic, an approximate equivalent, Kumushau, it comes from Egypt, something like made in Egypt, a far cry from the Rosetta Stone, indeed. On the other hand, the linguistic family of Meroitic has only recently come to light, giving us hope for further progress. We'll come back on this point later. What do we know about Meroitic in 2024? The writing system has been deciphered, the transcription of numerals has recently been improved, and only a few rare ideograms remain to be elucidated in commodity counts. Paradoxically, word pronunciation is fairly known, not least because the writing system is phonetic. A few points remain to be clarified. The vowel O seems to be transcribed by several means, and the writing of the two nasal consonants, that is, the vowel consonants of English king and the palatal consonant of Spanish niño, has not been fully elucidated. Text typology. The typology of text is also well defined, and almost all the new inscriptions discovered each year can be immediately classified under a defined genre. Vocabulary is the weak point in the deciphering process. Too few words have a reliable translation, even if semantic approximations are proposed for several hundred terms. The linguistic family is well known. Finally, grammar juxtaposes areas of syntax that are very well understood and analyzed, with huge gaps in knowledge of morphology, particularly verbal morphology. And very recently, a pronominal system has been partly elucidated, but still a work in progress. The final element of optimism is the steady increase in the corpus. We found several dozen texts in the Sdenga necropolis, whose mission I am leading, notably in 2016 and 2017. On the right hand, you can see the last stele discovered by a Sudanese archaeologist in 2023, that is just a few months ago. What methods are used to make progress and can give hopes to continue along on the path of translation? As we shall see, almost each decipherment is a special case. But the methods employed for Etruscan, among others, have a number of points in common with the approaches of the Meroiticists. Griffith's initial progress, once the script had been deciphered, was made thanks to Egyptian. While bilingual texts of sufficient size have yet to be discovered, as you have seen, the similarities between certain Egyptian texts that served as models and the Meroitic equivalent can be exploited. This is what Etruscologists call the bilinguistic methods. And here you got an example, but for Etruscan. The article described, which you can see here, describes three parallel donations for Muley in Faliscan, an ancient language close to Latin, Greek, and Etruscan. And all those formulas, that is, duenum, duenas, kalos, kalo, or in Etruscan, melakas, all those formulas means a beautiful object for a beautiful person. So, how can we do for Meroitic? But already Griffith found such a parallel text in the There are several pious graffiti in which pilgrims engrave a fruit or two, with or without juice, as a symbol of their sacred journey, accompanied by an inscription in Demotic or Greek, declaring that so and so's feet are here in the presence of Isis or another deity. It also exists in Meroitic with the same wording, which has led to the deduction of the Meroitic words for feet and the postpositional phrase in the presence of. Egyptian is also the source of large number of Meroitic borrowing, including the names of gods, for instance, Amanai, which is coming from Egyptian Imen, Amun, or Usuri, which is coming from Egyptian Usir. We got also Celtic and administrative function, such as Annata, priest from the Egyptian Hemnetia, or Upute, envoy from the Egyptian Uputi, and even some unexpected elements, such as precious metals. So Nabara, gold, from the Egyptian Nebu, pronounced Nabal, and Edra, silver, from the Egyptian Khetch. From such elements, Griffith was able to infer the meaning of certain propositions and compare it from text to text on similar formulations. This progressive method of elucidation through textual context, called combinatorial analysis by etruscologist and philological methods in or discipline, remains a safe bet. It can be compared with crossword puzzles, whose crossword puzzles, where most of the work consists of filling the gaps, once certain words have already been found. In this game, information is conveyed in two directions, across and down. Similarly, to fill the gaps in Meroitic inscription, we can combine two textual dimensions, syntax and semantics. For instance, an unknown word at the end of a full sentence must be a verb, because it always occurs last according to strict Meroitic syntactic rules. On the other hand, an unknown adjective following the word water, in the context of offering to the deceased, must have a positive meaning, such as good or abundant water. And by the way, these two adjectives are attested in this context. Because images are ubiquitous in Meroe, like in Egypt, recourse to the iconography accompanying the inscriptions often serves as a starting point for the philological method. Let's take an example from the great enclosure of Moussavarat, an immense cultic complex where pilgrims flocked for ceremonies and oftentimes inscribed various graffiti. This graffiti depicts a dog chasing a hare. The facsimile was published in 1979. At the time, the sentence was difficult to understand and only the second phrase could be translated. It belongs to Netarora, Netarora being of course a personal name. So the word which is written T-L-T was attested in other texts, but it wasn't until 2000 that my late colleague Nick Millett and I proposed reading the word Talanta with a nasal never written as a rule before another consonant. The term is borrowed from Greek. It refers to a talent, an ancient unit of weight for metals. Here it is probably a bronze talent, less expensive. The caption includes an adjective, which was long erroneously translated as noble. Later, it was interpreted as a demonstrative pronoun, translated as this one. In fact, as I have shown in my grammar, it can also be a demonstrative adjective, meaning this, a bit like in English where this can be a pronoun or an adjective. The slight difference in translation is of the utmost importance because the adjectival function makes it possible to match the image to this caption. This something and this something is of course represented on the image. Thanks to it, we know now the name of the dog in Meroitic, W-L-E, pronounced wild. The inscription is therefore a wish addressed to the gods and the written verb can be translated from the context as to earn, to yield. Indeed, the dog's owner hopes to earn this considerable sum probably by selling games rather than by betting on greyhound races, of which that existed in Greece and in Gaul. But we have no evidence of this kind of thing in the Nile Valley. The problem with these internal methods is their slowness. In fact, they yield a meager booty, just a few words per year. The most effective approach is linguistic comparison, that is comparing the unknown language with known language of the same family. However, it was still necessary to prove the relatedness of Meroitic with other languages. And yet, for almost a century, no such link could be demonstrated. Was Meroitic related to Old Nubian, as Griffith initially thought? After further analysis, he abandoned this idea, which was a wrong idea, as we will see. Did Meroitic belong to the Hamitic family, as Ernst Schwarz suggested? This Hamitic language family was later renamed Cushitic, from Cush, because of this hypothesis, which was wrong. But it soon became clear that Meroitic was not Hamitic or Cushitic. Cushitic languages, just to give you a few examples, it's Bedja in Sudan, maybe you know better Somali, or languages that way, Oromo in Ethiopia, etc. But finally, it's not the family of Meroitic. So was Meroitic an isolated language like Sumerian, or did it belong to an extinct family like Etruscan? In 1963, American linguist Joseph Greenberg published a seminal book entitled Languages of Africa, in which he divided African languages into four super families, or phyla. So you can see those four families here, Afro-Asiatic in blue, Niger-Congo in red and orange, orange for Bantu languages, Khoisan, then in green, and Nilo-Saharan in yellow. This latter group included a certain branch called East-Sudanic, which comprised languages from Sudan, such as Nubian, which had not been at this time completely ruled out as a possible relative of Meroitic. A young Canadian archaeologist working in Nubia, he was young at this time at least, Bruce Trigger, saw in this new classification the key to the problem. He published in 1964, just one year after Greenberg, an article entitled Meroitic and East-Sudanic, a linguistic relationship. He attempted to show correspondences between the lexicon and morphology of Meroitic on the one hand, and Nubian and Nara, a small language from Eritrea, on the other. Unfortunately, he relied heavily on erroneous data from Ernst Suchiarz, which completely discredited the analysis he did among specialists at the time. In fact, the progress made in recent decades on Meroitic on the one hand, and on the description of African languages on the other, and here I must salute my German colleagues who have done great work for describing African languages, and especially Nubian languages, has made it possible to assign the language of Cush to a specific family, which I called Northern East-Sudanic. In fact, it is the northern part of Eastern-Sudanic. Bruce Trigger was right, even if he used partially erroneous data. As expected, the closest relatives of Meroitic are the Nubian languages spoken in Egypt and the Sudan. So, let's go back to the Moussawa graphite. We can see that the recently identified word dog has a close cognate in Nubian. So, wal in Meroitic, and in Dongolese Nubian, it is wel. So, it's not borrowed from Meroitic, because it has reflexes, cognates, in Darfur Nubian, where no Meroite ever put their feet. So, Ness, to call them with a simple word, Northern East-Sudanic, Ness languages form a subgroup of Eastern-Sudanic languages, which include the Nilotic language, which you might know, such as Maasai, and Luwo. Luwo was spoken by President Obama's Patalon family. So, these languages, ultimately, belong to one of Africa's four great phylum, Nilo-Saharan. So, it's true that the unity of Nilo-Saharan is currently being called into question, but the Eastern-Sudanic subgroup remains unchallenged and solid. On the other hand, there is no doubt about the kinship of the Northern East-Sudanic languages, which comprise three groups. The Eastern Ness group, with Nara from Eritrea, formerly called Barya. In fact, Barya was used by the Ethiopian of the plateaus, you know, to say slaves, so they don't want to be called anymore Barya. We must call them Nara. Also, the Nubian group, and Neroitic, on the one hand. Then there is the Taban group, which is spoken in Wadai, that is one of the eastern provinces of Chad, and in Darfur. And the Nima group, in the Nuba mountains in Sudan, including two small languages, Afeti, something like 3,000 speakers, you see that it's really an emergency case to describe this language. And the second being Niman. Here is a map of these four groups, Neroitic, so Taban, Nima, Nara, and Nubian. The study of these languages is therefore crucial to the decipherment of Neroitic. However, with the exception of Nubian, which has had dictionaries and grammar for centuries, just imagine that the first Nubian-Italian-Mexican dates back to 1650. These languages are poorly described, if at all. That's why I had to swap the epigraphist magnifying glass for the linguist recorder and survey the Nuba mountain to document Nima language as Afeti here, or Niman, which is one of my main informants, Sheikh Hamreka, which means Sheikh of America, you know. And people told me, but strange, we thought they had a president. And also Nara, here is a wedding in the Nara tribe in then eastern Eritrea. So in 2010, I published Le Neroitique et sa famille linguistique, Neroitic and its linguistic family, a 600-page book that shows how Neroitic fits in perfectly, both in terms of vocabulary and grammar, with the Nass family. In 2012, with Alex de Voort's assistance, I published a shorter English version. Here, a small example of lexical correspondence between Neroitic and Nara. I could do the same with Nubian, with Niman, but of course it must be short, you know. The words for sister and slaughter are extremely close. So sister in Neroitic is kadit, and in Nara is kadé. What is most astonishing is the obvious similarity and correspondence between the two names of the creator god, Ape-de-mak, which means the god Ape-de-mak, maka means god, in Neroitic, and the name of the supreme deity of the Nara, used today to translate Allah, since their conversion to Islam. And this word then is Ebere. Similar names for creator also exist in Nubian, Ebeto, and in Niman, Abede, showing that these people not only share a related vocabulary, but have also inherited similar cultural concepts from their common ancestors. Finally, we can show that these languages are not only lexically related, but also have important grammatical similarities, which in linguistics is more conclusive. So there are no grammatical gender, like for example in English, a singer can be a female or a male singer, which then can designate a male or a female. The word order is subject, object, and the verb at the end, as I already told you. There are no prepositions, they are postpositions. An example in English, in English we have prepositions of course, but when you say for example 10 years ago, ago is a postposition. So the adjective is placed after the noun, like in French or Arabic, but unlike for example English or German. The genitive construction is possessor-possessee, like in English for example Tom's house, and not like in House of Commons. The number marking is not singular plural, it is collective-singulative. It means that to get a singular, you got to add something, which is just the opposite of English or French. The plurality of object is never marked on the noun, but it is marked on the verb, which is quite exotic of course for Europeans. Finally, direct and indirect object, that is what we call for example in Latin grammar accusative and dative, are marked by a case suffix g plus voile. So is linguistic comparison the miracle cure for translating all those ether too obscure neurotic texts? Obviously it's not that simple. The dispersal of the nest languages globe began over 400, 4000 years ago, corresponding to depths of time almost equivalent to that separating the Indo-European languages, for example the distance between Russian and English. As a result, these languages are not sufficiently close to one another for similarities to be immediately obvious, as it is the case for example between Romance languages. It is often necessary to resort to reconstructed form to show how words are related. Linguistic comparison is therefore not a miracle cure, but a powerful tool which added to internal methods and to the use of archaeological and iconographic contexts, enables us not only to progress towards the understanding of the text, but also provides first-hand information on the Kushite civilization. When I began working on Merotic over 20 years ago, people thought the case was hopeless. I've been told so. It is now reasonable to hope that the text of the earliest written culture south of the Sahara will be understood in the next future. Thank you.