An interesting passage from The Poison King: The Life and Legend of Mithradates, Rome’s Deadliest Enemy (see this post):

Mithradates’ dazzling memory and facility with languages were legendary in his own time…. Only one other individual in antiquity had linguistic abilities that even approached those of Mithradates. According to Plutarch, Queen Cleopatra of Egypt “spoke many languages and gave audiences to most foreign ambassadors without the help of interpreters.” Aulus Gellius remarked that “he was thoroughly conversant in the dialects of the 25 nations that he ruled, and spoke each language as if it were his native tongue.” …
Which languages did Mithradates speak or read with ease? These are certain: Greek, Macedonian, Persian, Latin, Aramaic/Hebrew, Parthian, Armenian, Old and New Phrygian, Cappadocian, and the Gaulish dialect of his Galatian lover Adobogiona. Other languages may have included Avestan (Old Iranian, used in Zoroastrian prayers); Sanskrit (Hindu medical texts); Egyptian and Punic; Celtic/Gallic (perhaps Allobrogesean, the language of his bodyguard Bituitus). He knew some Anatolian tongues, such as Carian, Mysian, Isaurian, Lydian, Lycian (and Pisidian), and maybe had a smattering of Syriac, Elamite, and Sumerian (used in religious texts of the Seleucid era). He could have learned Italian dialects, Marsic, Oscan, and Umbrian; Thracian (spoken by many of his cavalry regiments; and Getic (spoken in Tomis on the Danube). Other possibilities include vestig[i]al forms of Assyrian or Hittite and dialects of Colchis, Sarmatia, and Scythia.

There are obviously heaping dollops of speculation in that passage (one somehow doubts the Sumerian), but it’s a useful rundown of the linguistic situation in that part of the world a couple of millennia ago. (One wonders, though: why the parentheses around “and Pisidian”?)


  1. The author may have a special gift for piling up huge lists of names which may be united by their exotic sounds more than by any real connection?
    I have an issue with the Elamite too. Not only was it extinct for about half a millenium already … it was also so completey unrelated to both Indo-European and Semitic subsets of the list!
    (If she wanted the king to memorize a dozen or two of non-IE, non-Semitic languages, then he wouldn’t have any problems finding them among his subjects on both sides of the Caucasus Range, beyond the Colchis!)

  2. First of all, “Italian” in this passage is obviously a typo for “Italic”. Second of all, is there any evidence that Mithradates spoke any language *known to us today* except Greek (probably his mother tongue)? For all we know the score of languages he knew may well have all been regional languages of Anatolia of which we have no knowledge (and which might not even be related to any of the languages of the region known to us today).
    Just because he dealt with Romans didn’t mean he knew Latin: it’s a well-known fact that the Romans used Greek a great deal in the Eastern Mediterranean. Furthermore, just because he was of Macedonian descent does not mean that he spoke Macedonian. Also, I am baffled that Sanskrit, the language of “Hindu medical texts”, is listed: I am unaware of any evidence that Sanskrit manuscripts (on any topic) were in circulation this far to the west of India during this period.

  3. Yes, both Sanskrit and Elamite bothered me as well. But if she says it’s “certain” he knew Latin, I assume she’s not just making it up but that Plutarch or somebody says he did.

  4. komfo,amonan says

    Is there any evidence for the Isaurian and Mysian languages themselves at all? Is the assertion that he knew these languages a wholesale fabrication?

  5. komfo,amonan says

    And whatever language the Allobroges spoke, no one would dub it ‘Allobrogesean’. Shame they didn’t catch that.

  6. Looking at a map, I don’t think either Sumerian or Elamite are plausible. Pontus itself was originally along the Black Sea coast, north of Mesopotamia and outside the region where Sumerian was a prestige language. It was also much further outside the region where Elamite was spoken in southern Iran.
    It’s not impossible, I suppose, but it does seem pretty unlikely.

  7. it is possible that he learnt some sumerian/ sanskrit, for example, in order to read some classics? perhaps copies through merchants? the past was not as ‘closed’ a society as we sometimes think..

  8. Mithridates was in the wrong area to have learned Sumerian. He grew up well outside the area where scribes had learned Sumerian, and at its largest the Pontic kingdom was only at the edge of that area.
    At that time in history Sumerian had been dead for roughly two thousand years. However, the language was used like Latin was up until roughly 1 AD, so there were people still around who knew it. As far as we know, though, they lived in Mesopotamia. It just doesn’t seem likely that Mithridates would have had anyone to learn it from.
    If he’d been in Babylon, sure, it wouldn’t have been such a stretch.

  9. Mithridates’ empire was on the eastern Black Sea and it expanded into the western Caucasus. Pliny the Elder claimed 300 languages were spoken in the Caucasus and Rome needed 130 interpreters to deal with its business there. Strabo said that 70 different languages could be heard in the Greek colony of Dioscurias (now Sukhumi, the capital of Abkhazia). Dioscurias was one of the cities Mithridates conquered. So I think it’s far more likely that he knew the ancestors of, say, Abkhazian, Georgian and Laz (Proto-Zan?) than Elamite or Sumerian.

  10. Yeah, I’ve just been looking up the history of Georgia. Mithridates ruled over a lot of “proto-Georgian” peoples such as the Macrones (possible ancestors of the Mingrelians), the Chalybes, the Sanni, the Mossynoeci etc. He ruled Colchis itself, of course. He also ruled the Lazi (Lazica became quite an important state in Late Antiquity; it was a bone of contention between Rome and Persia, leading to the Lazic War described in Procopius’ De Bello Persico).
    So I think it’s safe to surmise that Mithridates had one or two Kartvelian languages in his arsenal. She should probably have paid closer attention to the comment “he was thoroughly conversant in the dialects of the 25 nations that he ruled.” I don’t imagine they included Umbrians or Marsians.

  11. Just to pile on – wasn’t Macedonian already more or less extinct by 88 BC? Even in Alexander’s time it was beginning to lose ground to Koine. What evidence do we have that Mithridates spoke it, and to whom?

  12. Yeah, I think she was just trawling for anything that might have been spoken or studied within some large circumference around Pontus and saying “Hey, he could have been acquainted with these,” rather than making any effort to decide what was actually plausible. It’s too bad she’s so sloppy—she writes pretty well and has an interesting perspective, but she’s as careless with the facts as, say, BBC Science News.
    For Isaurian, see this 2005 LH thread.

  13. aqilluqaaq says

    καὶ τὴν γλῶτταν ὥσπερ ὄργανόν τι πολύχορδον εὐπετῶς τρέπουσα καθ’ ἣν βούλοιτο διάλεκτον, ὀλίγοις παντάπασι δι’ ἑρμηνέως ἐνετύγχανε βαρβάροις, τοῖς δὲ πλείστοις αὐτὴ δι’ αὑτῆς ἀπεδίδου τὰς ἀποκρίσεις, οἷον Αἰθίοψι Τρωγλοδύταις Ἑβραίοις Ἄραψι Σύροις Μήδοις Παρθυαίοις. πολλῶν δὲ λέγεται καὶ ἄλλων ἐκμαθεῖν γλώττας, τῶν πρὸ αὐτῆς βασιλέων οὐδὲ τὴν Αἰγυπτίαν ἀνασχομένων παραλαβεῖν διάλεκτον, ἐνίων δὲ καὶ τὸ μακεδονίζειν ἐκλιπόντων. (Plutarchus, Antonius, 27.2)
    Mithridates, duarum et viginti gentium rex, totidem linguis iura dixit, pro contione singulas sine interprete adfatus. (C. Plinius Secundus, Naturalis Historia, 7.88)
    illum solum mortalium certum est XXII linguis locutum, nec e subiectis gentibus ullum hominem per interpretem appellatum ab eo annis LVI, quibus regnavit. (C. Plinius Secundus, Naturalis Historia, 25.6)
    Mitridates autem, Ponti atque Bithyniae rex inclutus, qui a Cn. Pompeio bello superatus est, duarum et uiginti gentium, quas sub dicione habuit, linguas percalluit earumque omnium gentium uiris haut umquam per interpretem conlocutus est, sed ut quemque ab eo appellari usus fuit, proinde lingua et oratione ipsius non minus scite, quam si gentilis eius esset, locutus est. (A. Gellius, Noctes Atticae, 17.17.2)
    Cuius utriusque industriae laudem duo reges partiti sunt, Cyrus omnium militum suorum nomina, Mitridates duarum et XX gentium, quae sub regno eius erant, linguas ediscendo, ille, ut sine monitore exercitum salutaret, hic, ut eos, quibus imperabat, sine interprete adloqui posset. (Valerius Maximus, , 8.7(ext).16)
    uel Mithridates, cui duas et uiginti linguas, quot nationibus imperabat, traditur notas fuisse, ( M. Fabius Quintilianus, Institutio Oratoria, 11.2.50)
    Mithridates rex Ponti oriundus a septem Persis, magna ui animi et corporis, ut sexiuges equos regeret quinquaginta gentium ore loqueretur. (Pseudo-Aurelius Victor, De viris illustribus urbis Romae, 76.1)

  14. I’ve just looked up the passage you quoted from The Poison King in Google Books and the previous paragraph actually mentions the facts in my first post here, then Mayor ignores them and jumps to the far-fetched conclusion about Mithridates being some sort of universal polymath. Odd. I suppose she’s got the problem that Mithridates is a fascinating figure but there’s just not enough hard data about him to fill a 300 or 400-page biography.
    One book I read and really enjoyed earlier this year about another of Rome’s enemies was Richard Miles’ Carthage Must Be Destroyed. It’s probably a lot more scholarly (while still being readable).

  15. as careless with the facts as, say, BBC Science News
    LOL … but IMVHO it transcends the “mere carelessness” because the loosely-compiled lists actually serve the author as a useful rhetorical means to advance her title assertion about “Poison King”.
    When she mentions medical texts in Sanskrit, and more extinct languages of the sacred lore, it’s like, “see now, the king surely have mastered all the ancient secrets about health, cure, and poison”.
    When she lumps together pigments and poisons of all types, it also help her assert that Sinope, the king’s seat, wasn’t merely a transit port for a fine ocher, but more like a hub of illicit knowledge and exchange of all countless uncontrolled substances.
    These lists are not mere streams of exotic words and sounds; they actually do attempt to support the author’s case. Each of the lists is like adding another clay foot to a colossus, in a sense.

  16. Bill Walderman says

    The sources quoted by Aqilluqaaq suggest that the 22 languages Mithridates is supposed to have spoken (leaving aside the extreme 50 figure put forward by Pseudo-Aurelius about 400 years after M’s death) were all languages spoken within his dominions. Pliny the Younger, writing about 150 years after M’s death, seems to be the earliest surviving source for 22 languages.
    You have to wonder what authority Pliny had for 22 languages, as well as whether many of the 22 couldn’t simply be lumped together as slightly differentiated dialects of the same language (recognizing the fuzziness of “dialect” and “language”).
    At any rate, the quoted passage from the book (“25 languages”) doesn’t give much confidence in the author.
    Incidentally, the Pontic region seems to have been associated in ancient myth with poisons and magical herbs and potions–witness Medea.
    You can’t really trust ancient historians about these matters, especially where a well-known and feared enemy of Rome is concerned.

  17. marie-lucie says

    I was glad to learn about the Isaurians, whose name and existence I had not known before.
    I find the following quote interesting:

    The Isaurians however were one of these small but tenacious groups of peoples like the Basques and Bretons, the Welsh and the Picts, and the Albanians whom the Romans couldn’t completely suppress.

    (from “brian”, responding in the 2005 thread LH links to).
    All these peoples have in common the fact that they inhabit(ed) coastal regions with rugged landscapes. Mountains protect their inhabitants by providing obstacles to invasion by would-be conquerors (consider the Caucasus, Switzerland or Afghanistan). The proximity of the sea means a ready source of food, making the coastal people relatively independent of the hinterland, and boats also enable not only exchange but piracy. Coastal mountains have always provided the ideal setting for the survival of a “small but tenacious” population, and therefore of its language (see Albania, Norway, Korea). Open plains and flat coastlines leave their inhabitants open to invasions, massacres and eventual assimilation.

  18. Residual zones and spread zones in Johanna Nichols’s scheme.

  19. just a thought: wouldn’t there be someone 2,000 years from us, who would claim that the Pope who blesses the pilgrims from his balcony in 22 plus languages, could speak 22 plus languages?
    One of the first books I read was ‘The Legends of Crimea’, on a holiday in Eupatoria, I think, in the West of the peninsula, a holiday town named after Mitradates the Eupator. The book, spanning Antiquity to WWII, had a novella on him. I remember how amazed I was that he trained himself to be resistant to all kinds of poison. So he had to have himself killed in Pantikapei-Kerch, where they still dig the ruins of the akropolis today.

  20. John Emerson says

    That area just forces you to make lists. To quote myself:
    At one time or another Gothia lived under the control of (or within the sphere of influence of) the Scythians, the Greeks, the Romans, the Byzantines, the Huns, the Khazars, the Comans / Polovtsi / Kipchaks, the Varangians / Rus / Russians, the Genoese, the Mongols, the Tatars, Tamerlane, the Cossacks, and Russia, and Gothia also had significant dealings with the Bosporan Greeks, the Greeks of Trebizond, the Petchenegs, the Alans, the Avars, the Bulgars, the Hungarians, the Crusaders (from “Romania”), the Wallachians, the Zikhians (whoever they were), and the Lithuanians.
    So: who were the Zikhians?

  21. Bill Walderman: you’re quite right, many of the 22 “languages” might have been what we today would call dialects within a continuum. This is doubly likely if we remember that Mithradates, it is claimed, spoke the languages of the various peoples he governed: many peoples (=tribal groupings) in those days shared a common language, with small dialect differences being sufficient as an “identity marker”: thus, there existed a Gaulish language at that time, but many Gaulish-speaking peoples.
    Actually, now that I think about it…could some (or even all?) of these 22 languages have been dialects of Greek? Greek had been spreading along the Black Sea Coast for at least five centuries before Mithradates’ time, which is more than enough time for significant dialect diversity to arise, especially if language contact played a major role.
    JCass: at this time were there separate Kartvelian languages already, or was Proto-Kartvelian a more or less unified language at the time? Also: just because some ethnic names attested then are attested later needn’t imply linguistic continuity: the French today derive their name from the Franks, but not their language.
    Marie-Lucie: I agree with your general point, but it is frankly anachronistic (on Brian’s part) to refer to “Welsh” and “Breton” peoples in Roman times: Breton was transplanted to the continent from Britain in the fifth century AD, and the Anglo-Saxon invasions are what eliminated Celtic from England and restricted it to Wales: both events are post-Roman. Also: while Albanian is spoken along the Mediterranean coast today, it almost certainly was not in the days of the Roman Empire.

  22. Well, the ancestors of Georgian and Laz may indeed have been one Proto-Kartvelian language (although I find it hard to imagine it not being split into differing dialects given the mountainous terrain, but I’m no expert). The stuff about the Macrones etc. being proto-Georgians (or proto-Kartvelians) I got from modern historians of the region.
    For a sceptical view of Mithridates’ linguistic abilities, there’s the 19th-century German historian Wilhelm Ihne. Judging by the snippet view from Ihne’s History of Rome at Google books, Ihne thinks “we must characterize as a ridiculous exaggeration what is related of his knowledge of twenty-two languages. Probably somebody had declared that Mithridates did not understand Greek alone, but also the language or languages of his native Asiatic subjects. To make this vague statement circumstantial, some over-clever narrator counted up twenty-two names of tribes subject to his sway at one time or another, imagined that each of these spoke a distinct language, and inferred that Mithridates must have…” (here the snippet breaks off but the conclusion is obvious).

  23. aqilluqaaq says

    Actually, now that I think about it…could some (or even all?) of these 22 languages have been dialects of Greek?
    If we have Cleopatra knowing Ethiopian, Trog(l)odyte, Hebrew, Arabian, Syrian, Median, Parthian, Egyptian, Macedonian, and Greek, I see no reason to suppose Mithridates’ twenty-two all Greek. And in any case Quintilian explicitly contrasts Mithridates’ knowledge of 22 languages with Crassus’ knowledge of 5 Greek dialects in the same sentence.
    uel Mithridates, cui duas et uiginti linguas, quot nationibus imperabat, traditur notas fuisse, uel Crassus ille diues, qui cum Asiae praeesset quinque Graeci sermonis differentias sic tenuit ut qua quisque apud eum lingua postulasset eadem ius sibi redditum ferret,

  24. Also: while Albanian is spoken along the Mediterranean coast today, it almost certainly was not in the days of the Roman Empire.
    Different Albanian.

  25. Different Albanian.
    But predecessor of the Albanian of today’s Albania was probably spoken along the Adriatic coast (Illyrian or smth. related?).
    (There is also the third Albania, of the Sicilian Greeks, if I recall it right)

  26. Alba, the old name for Scotland, was also called Albania by the Romans.

  27. marie-lucie says

    Etienne: I was making a general point, and so was “brian”, I think. I didn’t get the impression that he was restricting his comment to languages existing at the time of the Isaurians, I understood he was giving modern analogs of the Isaurian situation in antiquity, which is what I did too. Obviously my mention of Korea had nothing to do with the Roman Empire.
    It does not matter to the current status of Breton that the language came from across the Channel centuries ago: the point is that its geographical situation was probably the main factor in its survival to modern times.

  28. Etienne: J. Markale proposes in Celtic Civilization that of the three dialects of Breton, one may show survivals of a Gaulish substrate.
    The implication is that despite the assertion that Gaul had been totally Latinized, there were still Gaulish-speaking pockets in isolated regions. Thus the British refugees chose to go to an area where there were already people speaking a closely related language.
    I’m not nearly informed enough to have my own opinion on this topic, so I’m merely passing on what I’ve read.

  29. I think Ihne has nailed it. It was doubtless sufficiently unusual in those days, as in these, for rulers of multilingual empires to know any language but their own to make the matter worth mentioning.

  30. komfo,amonan says

    Alba, the old name for Scotland, was also called Albania by the Romans.

    Alba is still the name for Scotland. The Romans called it Caledonia or Pictavia, IIRC, while Albania shows up in the Middle Ages.

  31. Hat, Mockba: I was referring to the Albanian language of present-day Albania, which indirect evidence suggests spread to the Adriatic coast, and indeed to present-day Albanian territory, well after the fall of the Western Roman Empire: its original homeland seems to have been located further North, and inland. In the thread “The most interesting language” I gave a reference on the topic, as well as another on the evidence that Gaulish in Brittany had been wholly replaced by Romance on the eve of the invasion of Celtic speakers from the British isles.
    Maidhc: I have seen several claims that Breton and Gaulish came into contact in Brittany (as for the claim that a particular dialect of Breton only was affected by Gaulish: I believe Fleuriot was the first to make such a claim).
    Unfortunately, these claims seem to be motivated by regional/Celtic identity sentiments rather than by any actual linguistic data. The dialect diversity of Breton –a language that has been spoken in its present-day location for some fifteen centuries with little if any standardisation– can readily be explained without assuming that a Gaulish substrate played a role in the genesis of said dialect diversity.
    Aqilluqaaq: many thanks for all the textual attestations. I’m glad to see my hunch was right: there’s no clear evidence that Mithradates even spoke Latin (something Romans would have noticed and mentioned, I imagine). The contrast Quintilian makes between Crassus’ knowledge of five Greek dialects and Mithradates’ knowledge of 22 languages is interesting. But Quintilian would have had first-hand acquaintance with Crassus’ TRUE linguistic ability/repertoire, and would have known of Mithradates’ only through second-hand sources. So I wouldn’t quite shut the door on the idea that some or all of Mithridates’ 22 languages (assuming this claim has any truth to it at all) were what we’d call Greek dialects.

  32. marie-lucie says

    Gaulish and Breton
    I too have heard (or rather seen in print) something about the Vannetais dialect of Breton (the southernmost one) being (at least partially) a continuation of Gaulish, but I have not seen the arguments in favour of the hypothesis and I totally lack expertise in this area. I agree with Etienne that identity politics probably played a role in this hypothesis.
    The language closest to Breton is claimed to be Welsh (not Cornish which might seem logical because of geography).

  33. marie-lucie says

    caffeind: Residual zones and spread zones
    Yes, that’s basically what I was referring to, without using these terms. Johanna Nichols mentions mountains for residual zones, not coasts (she is a Caucasian specialist), but for the Pacific side of the Americas, as well as for Basque and the surviving Celtic languages, I think that it is telling that the ‘residual’ languages are situated in areas with mostly mountainous or at least very rocky coasts.

  34. The language closest to Breton is claimed to be Welsh (not Cornish which might seem logical because of geography)
    Claimed by whom? I believe it’s generally held that Cornish is much closer to Breton than Welsh is, and my slight acquaintance with all three languages supports that.

  35. marie-lucie says

    OK, Rodger C, it is possible that what I read (I don’t remember where) referred only to Welsh and Breton because Cornish was practically extinct at the time.
    By the way, what would be a good reference for the Celtic languages in general?

  36. Marie-Lucie: Alas, my knowledge of Celtic scholarship is mostly from graduate school days in the Seventies. I can hardly tell Lot from Loth nowadays.

  37. Marie-Lucie: everything I have read about the classifiction of British Celtic indicated that Breton and Cornish form a subgroup standing in opposition to Welsh, so I suspect your hunch above is correct.
    In answer to your question (speaking as a non-Celticist who has done some exploratory reading on the topic): Martin Ball and James Fife’s THE CELTIC LANGUAGES (Second edition of 2009) would probably be the best introduction to Celtic languages and linguistics: the various sketches of individual (living and extinct) languages, as well as the articles on Celtic typolology and classification, are all excellent. Furthermore, the bibliographies are quite thorough, which is very useful to the curious reader.

  38. marie-lucie says

    Merci, Etienne. That’s the kind of book I am after.

  39. Etienne: I also thank you for the reference.

  40. Trond Engen says

    And thanks from me for the reminder. I got it a couple of years ago and started reading, but tucked it away when something else came along. Yesterday morning I picked it out from the shelf again.

  41. US$55 at Amazon, more at ABEBooks. Oh well, as someone else who’s adopted Hat’s rule of at most US$10 per book, I am doomed to remain in ignorance.

  42. marie-lucie says

    Since LH already owns thousands of books, he is wise to limit himself to buying the cheaper ones, otherwise he would really run out of shelves if he indulged in more expensive books too.

  43. Al ber quer quay says

    John Emerson says:’ So: who were the Zikhians?’

    The Zikhians are the Circassians of today.

  44. Thanks! The key sentence from that link: “In the Middle Ages Circassia was known as Zikhia after the Zikhs, a Circassian tribe living on the shores of the Black Sea.

  45. David Marjanović says

    I’m sure all Greek dialects would have been considered Greek by their speakers and by other Greeks, so it’s highly unlikely that Greek dialects were counted separately. However, the claim is that he could act as a judge to the 22 peoples he ruled without needing interpreters; that doesn’t preclude the possibility that several of those peoples could have spoken the same language or mutually intelligible dialects, or the possibility that he spoke a respectable but single-digit number of, uh, lingue franche that were widely enough used – we are, in part, talking about the Caucasus after all.

    the classifiction of British Celtic

    Nothing to see here, I just had to quote this out of context. 🙂

Speak Your Mind