What does a king do in his spare time? Sultan al-Afdal al-Abbas, who ruled Yemen from 1363 until his death in 1377, seems to have enjoyed many of the leisure activities commonly favored by medieval Muslim kings and princes: falconry and archery; horsemanship and swordmanship; and, like many Islamic sovereigns, literary and scholarly pursuits. Well-versed in a broad spectrum of religious and secular disciplines, he wrote on such topics as eminent personages in Yemenite history, the genealogies of the Arabs and the cultivation of grains and cereals.
Nothing unusual there; in medieval Islamic countries we would expect a learned man to have an interest in such subjects. But in the early 1970s, historians were stunned to discover that this scholarly monarch had been responsible for yet another work: a spectacular polyglot dictionary, in which he listed about 1,200 Arabic words and then, alongside, in parallel columns, supplied their equivalents in five other languages: Persian, Turkish, Greek, Armenian and Mongolian.
This was unusual. Though multi-language dictionaries were widely circulated throughout the Muslim world in the Middle Ages, most were lists of Arabic, Persian and Turkish words - a total of three. Never before had a dictionary involving six languages been found.
Such vocabulary lists, obviously, would have been of use to merchants and diplomats—not to mention those in intelligence work - yet al-Afdal's dictionary, apparently, was compiled with no immediate utilitarian aim in mind. Commercial terms are conspicuously absent, and the king's choice of languages, categories and words makes it highly doubtful that he gave any thought to the linguistic needs of diplomats, military men, civil servants or spies. To the contrary, all the evidence leads to the refreshing conclusion that the dictionary was compiled strictly as an intellectual pastime - a scholarly hobby. Al-Afdal al-Abbas, the sixth Rasulid sultan of Yemen (the Rasulid dynasty ruled Yemen from 1229 to 1454), apparently collected words in other languages the way some people collect stamps, coins and butterflies - for fun.
The "Rasulid Hexaglof' - as the dictionary has been dubbed by researchers - is one of a group of works by al-Afdal that were discovered in Yemen; the works were bound together to form a single manuscript. In 1974, a Lebanese scholar showed a microfilm copy of this manuscript to Professor Tibor Halasi-Kun, an expert in Turkic languages and history in the Department of Middle Eastern Languages and Cultures at Columbia University.
Immediately recognizing the historical value of the unique six-language dictionary, Halasi-Kun obtained a microfilm of that particular portion of the manuscript - the manuscript itself has never left the Yemen Arab Republic - and set about assembling a team of scholars to edit, translate and analyze the text of the dictionary. This is more difficult than it sounds because the text is written entirely in Arabic script (used but rarely to transcribe the Greek, Armenian, and Mongolian languages) and presents a number of paleographical, philological and historical problems.
That team, consisting of four internationally respected philologists, two in the United States and two in Hungary, is still at work on the dictionary, but hopes to publish its results in book form in the near future. Halasi-Kun is working on the Turkish section and his former student, Professor Peter B. Golden, currently associate professor of history at Rutgers University's Newark campus, is responsible for the Greek. Working together, Halasi-Kun and Golden are also translating the Arabic and Persian entries, which are linguistically far less difficult than the others. The Mongolian portion is being studied by Academician Lajos Ligeti of the University of Budapest, while his Hungarian colleague, Professor Edmond Schiitz, is handling the Armenian.
The whole tale has a wonderfully cosmopolitan flavor to it: a Yemenite king; Lebanese, American and Hungarian scholars; texts in Arabic, Persian, Turkish, Greek, Armenian and Mongolian. In a sense, though, it reflects medieval Yemen itself. Situated at the southern end of the Red Sea astride ancient maritime trade routes, Yemen had a long history of commercial relations with, to name a few, Egypt, Ethiopia, the Arabian Gulf countries, even India and Ceylon, and in its rulers and peoples, was multi-ethnic.
Prior to its takeover by the Rasulids, Yemen had been ruled by the Ayyubids, the dynastic house to which the great Saladin belonged, and then in 1239, after the Ayyubids lost their foothold, by the Rasulids. Of Turkic origin, either Turkomans or Turkmens, the Rasulids' present-day homeland in Central Asia is one of the 15 constituent republics of the Soviet Union. Among Yemen's ethnic groupings there were also Arabs, Kurds, and Turks, along with a sizeable Jewish population, a fair number of Persians and a smattering of Greeks, Armenians, Georgians and Circassians. The dictionary, then, should be viewed not only as the handiwork of a distinctly intellectual monarch, but also as the product of an exceptionally international milieu.
In his office on the Newark campus, Professor Golden enthusiastically discussed the dictionary and its 14th-century compiler. At 40, the cigar-puffing Golden is by far the youngest member of the team that will publish the manuscript, but he has already carved out a formidable reputation as a historian and philologist. For his doctoral dissertation on the Khazars - who inhabited the lower Don-Volga region, north of the Caucasus Mountains, and were one of the most important Turkic peoples of the Middle Ages - Golden studied medieval texts written in Arabic, Persian, Russian, Greek, Hebrew, Armenian and Georgian. He has contributed articles to the Encyclopaedia of Islam and the forthcoming Cambridge History of Inner Asia, and his future plans include research on the nomadic Cuman, or Kipchak, Turks of the Black Sea steppe in the Middle Ages, and on medieval Georgian chronicles dealing with the Mongol conquests. During the past seven years he has familiarized himself with all the columns in the Hexaglot, including those (Turkish, Armenian, Mongolian) for which he is not formally responsible.
"I think it is purely a scholarly hobby," said Golden when asked about al-Afdal's reason for compiling the dictionary. "Here is a very cultured, very learned gentleman, this Yemenite king. He is simply putting together a kind of vocabularium of what for him were the politically and culturally significant languages of the area of the world that was of interest to him, which means basically the eastern Mediterranean... He is talking about what were the great languages of this era, in essence."
To Golden, the sultan's dictionary is an impressive achievement, though in fact medieval rulers with academic tastes are nothing out of the ordinary. Ulugh-beg, grandson of Tamerlane, who governed Samarkand and its environs from 1409 to 1446, was an outstanding astronomer-mathematician, and many other kings and princes found periodic respite from their official duties and obligations by engaging in the study of history or literature. But, says Golden, al-Afdal al-Abbas's passion for lexicography - dictionary-making - though splendidly offbeat, is not what makes this Yemenite sultan so special. "The interest is unusual - but the guy was good!" declared Golden.
When work on the dictionary began, however, it was not at all obvious that al-Afdal had been a good scholar, because the Rasulid Hexaglot manuscript is not written in the king's own hand, but is a scribal copy i.e. a copy of a copy of a copy of a copy of... etc. The team assumed automatically that variations in spelling were scribal errors.
The assumption was logical. Like modern typesetters, copyists almost invariably introduced some errors, and the likelihood of error, naturally, would have increased whenever a scribe was required to copy data in a language that he did not know—especially in uncommon languages like Mongolian and Armenian.
Under such circumstances, it's easy to make a spelling mistake - and much harder to catch it afterwards, even when proofreading the manuscript. The team, therefore, frequently thought that the spelling of various words had been garbled during the copying of the dictionary. As Professor Golden said, "When you work with these things - and you're dealing with a copy - there is a tendency to sometimes see yourself as a little bit cleverer than the copyist. You say, "Well, it's obvious he made a mistake here.' "Now sometimes, indeed, it was the copyist's error. But, nine times out of 10, when the copyist was being faithful to the original and we had doubts about the genuineness of the given form of a word -our doubts proved to be completely unfounded. The king was a first-rate philologist, with a fantastic ear, and if he says something is pronounced this way, believe it. Because if we would search around, we would find the evidence that this isn't any aberration on his part; ifs a bona fide form."
The Arabic and Persian columns of the Hexaglot appear to be typical examples of the written Arabic and Persian of the late Middle Ages, but each of the other languages in the dictionary exhibits features which invite special comment. For instance, the Turkish words in the latter part of the dictionary, beginning with folio 7 (there are 10 folios in all), come from a dialect that is closely related to the kinds of Turkish spoken in modern Turkey, Azerbaijan and Soviet Turkmenistan, whereas the Turkish of the beginning section displays affinities with the group of Turkic languages that includes, among others, Tatar, Bashkir, Kazakh, and Karakalpak, all of which are presently spoken in the Soviet Union. This discontinuity is the tip-off- though there are other hints - that the Rasulid Hexaglot may, in fact, consist of two separate dictionaries which have been spliced together to form a single whole. In the Greek and Armenian portions, there are a number of common characteristics; both languages, for example, are written in Arabic script, a rare find. (The best-known example of Greek in Arabic script is a series of poems by the 13th century mystic, Jalalal-DinRumi.)
That is most interesting about the Greek and Armenian entries is that in both cases they bear witness to now-extinct spoken dialects of the Middle Ages. Like Arabic today, medieval Greek and Armenian each had a standard written form that adhered closely to classical models, so that two Byzantine historians living in separate provinces of the Byzantine empire and speaking distinctly different dialects of Greek would use much the same literary language in composing their respective histories. But the spellings that al-Afdal gives for Greek and Armenian words suggest strongly that he had heard those words pronounced by native speakers, rather than relying on written texts.
Furthermore, the Greek in the Hexaglot, which constitutes one of the few surviving records of a spoken Byzantine dialect, provides scholars with long sought-after clues concerning the evolution of modern Greek. Professor Golden believes that the particular dialect represented in the dictionary is Cypriot, or else is from the western coast of Asia Minor. Similarly, the Armenian found in the Hexaglot is clearly distinguishable from the usual literary Armenian of the late Middle Ages. Moreover, the dialect recorded apparently was not spoken in the Armenian motherland in Transcaucasia, but stems instead from the medieval Armenian kingdom in Cilicia along the coast of what is today southeastern Turkey. Cilician Armenia fell to the Mamluks during the lifetime of al-Afdal al-Abbas, but it had played a pivotal role in the history of both the Crusades and the Mongol conquests. It was logical, therefore, that the Yemenite king considered Cilician Armenian to be one of "the great languages of this era," to use Professor Golden's expression.
From the standpoint of philological scholarship, the Mongolian section of the dictionary is the most valuable column of all. Though Mongol rule in the Near East, which extended over Iran and Iraq and was centered in Persian Azerbaijan, lasted from 1256 to about 1335, all of the surviving literary works are in Persian, Arabic or Syriac. And though there was never any doubt among scholars that Mongolian was spoken in this part of the Islamic world for at least a few decades, little was known about the dialect until the discovery of the Rasulid Hexaglot.
As yet, it has not been definitely established whether al-Afdal used written sources alone in compiling this column of his dictionary, or whether, 30 or 40 years after the Mongols had ceased to rule in the Near East, he was able to find someone in Yemen who still spoke Mongolian. Either way, the king's curiosity about the language of Genghis Khan is a stroke of luck for modern specialists, who will profit immensely from his diligent labors.
The vocabulary items in the Hexaglot are grouped systematically according to subject. Professor Golden called the classification scheme "very, very scholarly and internally logical." Included among the many categories are the following: anatomy, bodily functions, beasts of burden, fur-bearing animals, insects, birds, kinship terms, parts of the day/week/year, numerals, weights and measures, currency, bodies of water, topography, trees, fruits, grains and cereals, colors, illnesses and afflictions, horses, household implements and tools, weapons, archery equipment, horse paraphernalia, assorted foods, clothing, precious metals and gems, crafts and craftsmen.
Indeed the Hexaglot research team has been overwhelmed by the orderly method followed by al-Afdal. "Since all of us are philologists by training," Golden pointed out, "we go through it and say, 'By God, that's the way to do it!' I mean, if I were doing something like that I might want to arrange it in this way. In a sense, we almost felt this king was a kind of a kindred spirit. He was interested in a number of things we're interested in."
The dictionary's detailed listings in such subject areas as falconry, archery and horsemanship indicate that these interests were especially dear to the king's heart. For example, he gives the Arabic, Persian, Turkish, Greek, and Armenian words (no Mongolian column is supplied here) for arrow, bow, quiver, bowstring, feather of an arrow, arrowhead and target. We also learn that a white-gray falcon is called al-bazal-ashhab in Arabic, baz-isaped in Persian, aq toghan in Turkish, aspron yerakin in Greek, and spidak baza in Armenian.
As expected, the most fundamental terms were included in the Hexaglot: God, man, woman, living, dead, earth, sun, moon, friend, enemy, bread, meat, milk, head, heart, Paradise, Hell. But ifs the linguistic oddities that have delighted the scholars translating the dictionary. And so we also encounter listings for cottage cheese, gnats, pasta, raincoat, shoemaker's awl, louse, and second stomach of a camel. Professor Golden's favorite entry is the Arabic al-'ukna, meaning "a single fold of skin across the abdomen, caused by fat." The only otherianguage of the six in which a single word denotes the same concept is Turkish. In the Persian, Greek and Armenian columns, the king was forced to insert two-word expressions meaning "fold of the belly" (the Mongolian is absent). A good lexicographer has to make do with what's available, of course.
The question remains - what does the Rasulid Hexaglot tell us about life in Yemen in the 14th century? The honest answer is, not a great deal. Instead, it tells us about the interests and concerns of a talented Yemenite king, and, in doing so, prodigiously enhances our knowledge of late medieval Turkic, Cilician Armenian, late Byzantine Greek, and an extinct Mongolian dialect of western Asia.
Yet the Hexaglot also raises as many questions about al-Afdal as it answers. What was he like as a person? When and how did he become interested in the world of words? Did he have friends with whom he could share his deep interest in the subject? And did he feel regret when the performance of kingly tasks kept him from his beloved research and writing for days or weeks at a stretch?
About these things and many others we can only wonder. Perhaps it is enough, though, that we have his dictionary - a gift to posterity. And the careful preparation that has gone into the forthcoming edition and translation of the Rasulid Hexaglot ensures that it will be a publication meeting the highest standards of scholarship - a publication in the tradition of al-Afdal al-Abbas.
Barry Hoberman free-lances from Somerville, Mass, and contributes frequently to Aramco World.
