I’ve been reading one of my birthday gifts, Heath W. Lowry’s The Nature of the Early Ottoman State (thanks, Jim!), a brilliant reanalysis of the early Ottomans that proves (to my satisfaction, anyway) that far from being fearsome warriors for Islam who presented conquered people with the famous “convert or die” choice, they were highly pragmatic rulers who allowed those they conquered to keep both their faith and in many cases their arms and former positions, which makes it easy to see why their rule spread so quickly among people crushed by late Byzantine taxes and misgovernment.

But on page 100 I hit one of those linguistic misjudgments that make me grind my teeth and reach for the Languagehat soapbox. Lowry is talking about a study he did of “a series of surviving Tahrir Defters, or Ottoman tax registers, from the Aegean island of Lemnos (Limnos), dating from the years 1490-1520″:

That members of the island’s late-Byzantine aristocracy were likewise performing military duties on Limnos is inferable from the fact that the peasant auxiliaries were serving under the command of their own officers, some of whom even appear in the 1490 tahrir with their former Byzantine military titles, for example, Kondostavlo, or, the “Count of the Stables.” From the Latin comes stabuli, or “count of the stable,” this was adopted by the Byzantines as the military title Konostaulos in the late thirteenth century.

Now, this is confused in more than one way; to take one obvious point, which was the title, Konostaulos or Kondostavlo? Apparently the former; compare The Late Byzantine Army: Arms and Society, 1204-1453 by Mark C. Bartusis, p. 28:

The increasing importance of Latin mercenaries during the reign of John Vatatzes [1221-1254] was symbolized by the creation of the office of megas konostaulos (“grand constable”), the “chief of the Frankish mercenaries.” The fact that the first megas konostaulos, Michael Palaiologos, became emperor is not without a certain significance.

Note that Bartusis translates konostaulos as “constable,” and this is correct, because the late medieval Byzantines obviously got the word from those Latin mercenaries; compare Old French conestable, which gives both the English word and modern French connétable. The OED says: “The early development of the sense, whereby the comes stabuli, from being the head groom of the stable, became the principal officer of the household of the Frankish kings, and of the great feudatories, and the field-marshal or commander-general of the army, had taken place before the word came into English; the development was parallel to that of marshal.” It’s just silly to suppose that comes stabuli would have been borrowed, with its original sense ‘officer of the stable,’ by Byzantines who hadn’t used Latin for centuries. (I suppose it’s possible that the phrase had been preserved in moth-eaten official registers since the fifth century, when Latin was still used in Constantinople, and readopted in its Frenchified form when the barbarians from the West showed up, but I’m not sure how far one could talk about continuity in such a case. I am not a Byzantinist, however, and will gladly defer to those who know more about such things.)

If you’re curious, the borrowed word turns up in one of my favorite Byzantine historians, the loyal, resentful, and vivid Anna Comnena towards the end of her Alexiad, describing a dilemma confronted by Bohemond at the Battle of Dyrrachium (1081): “He was debating what ought to be done, turning over in his mind numerous possible courses: should the constables appear before him [εί χρή παραστήναι τους κονοσταύλους]?” These, of course, are Frankish constables, not Byzantine officials.
Another curious linguistic sidelight: Speros Vryonis, Jr., in his The Decline of Medieval Hellenism in Asia Minor and the Process of Islamization from the Eleventh Through the Fifteenth Century (heretofore my vade mecum for the early Ottoman period), says on p. 234 “Michael Paleologus served as kondistabl in charge of Christian troops of the sultan after he had fled the kingdom of the Lascarids.” What the heck kind of word is kondistabl? The footnote refers us to Acropolites, a Greek historian; why would he be using a form like kondistabl? More mysteries.


  1. “Kondostavlo” does get a few Google hits, as a proper noun or in Turkish phrases, so it’s not something Lowry just made up. Maybe it’s an Ottoman and/or later Greek form.
    A loan from French (or from the Latin nominative form) would not have the “d”, but a loan from Italian or Catalan would have a “d” or “t” in that position.

  2. Excellent point: Spanish condestable and Portuguese condestavel could have been sources for an alternate form with -d- (but Catalan conestable, Italian conestabile could not, unless there were medieval variants with -d-). I had assumed that the Greek name Κοντόσταυλος (Kondostavlos) had Greek κοντός (kondos) ‘short,’ a common element in Greek names, but maybe it’s from the -d- variant.

  3. John Emerson says

    Pelliot’s “Notes on Marco Polo” (more or less impossible to find) gives variorum spellings of a lot of key words. A detailed study might be fruitful, but the conclusion I came to was something like “nothing was impossible in those days”. Most often it was Italian or French spellings of Persian, Turkish, or Mongol words, but IIRC even French and Italian words were variously spelled.
    The textual history of Marco Polo is an absolute monster. I’m not sure that a base text exists; I think that three different base texts have been theorized.

  4. The Spectator says

    I realize this is a language log, but in the interests of academic honesty, I couldn’t let Lowry’s name go without comment.
    I haven’t read this particular book, but take what he says with a grain of salt. Lowry is a notorious Genocide denier with a spotty academic record, and his chair is funded by the Turkish government.
    For links, etc, on what other academics have to say about him, visit http://users.ids.net/~gregan/.
    The Spectator

  5. Sigh. In the first place, the “notorious Genocide denier” accusation rests on his having signed (along with many other scholars) the following petition concerning a House of Representatives resolution:
    “No signatory of this statement wishes to minimize the scope of Armenian suffering. We are likewise cognizant that it cannot be viewed as separate from the suffering experienced by the Muslim inhabitants of the region. The weight of evidence so far uncovered points in the direction of serious inter-communal warfare (perpetrated by Muslim and Christian irregular forces), complicated by disease, famine, suffering and massacres in Anatolia and adjoining areas during the First World War.”
    As another signer, Alan Fisher, wrote at the time: “The main purpose of the petition was to encourage the Turkish Government to open completely its archives related to this subject, and it subsequently did so.” I think it’s ridiculous to equate such a position with “Genocide denial”; that kind of smear diminishes the crime of true deniers like Arthur Butz and Willis Carto.
    In the second place, whatever the merits of his position on the Armenian genocide, they are irrelevant to his researches on Ottoman history, which are groundbreaking. I despise the kind of politicization of academics that uses anger at a political position as a bludgeon to try to destroy an academic’s reputation entirely. This was done, for instance, to Anastasia Karakasidou by people who didn’t like her superbly researched book on Greek Macedonia (see this LH post). Why can’t people disagree without trying to destroy each other?

  6. Without in any way intending to reopen the can of wriggling worms in the comments of the Karakasidou post, I would point out that it was probably the C.U.P. lawyers who told them not to publish the book: since it mentioned living persons, there was a good chance that C.U.P. would be sued in the U.K. for libel. The U. of Chicago Press can take advantage of the United States’s much more robust guarantees of press freedom.
    See Geoffrey Pullum’s essay “Trench-mouth Comes to Trumpington Street” (in The Great Eskimo Vocabulary Hoax) about a similar incident where C.U.P. refused to publish a book on syntax, of all things, because of concerns about possible defamation in the examples.

  7. Looks like Vryonis has two more or less equivalent uses of kondistabl.
    p. 234 footnoted to Acropolites I, 137.
    p. 468 footnoted to Nicephorus Gregoras I, 58.
    Corpus Scriptorum Historiae Byzantinae is all in Google Books, and here is a handy index.
    Georgii Acropolitae, p. 137
    Nicephori Gregorae, Vol. I, p. 58
    I’m suspicious of the first one; looks like p. 146 might be a better shot at where Michael Comnenus is leading some Christian troops, but I didn’t search exhaustively. (Also possible is that I completely misunderstood those references and they aren’t pages at all.)
    Anyway, I don’t see any trace of the Turkish word here, so I’d guess it to come from Vryonis himself.

  8. The Rambler says

    I enjoy this blog, and don’t want to start a war. However, to say that the one thing is irrelevant to the other is specious reasoning. If a scientist didn’t believe in evolution (and there are a handful who don’t), you would probably be not a little suspicious of everything else they say, no?

  9. John Emerson says

    In Portland Oregon the attempt to set up a chair of Turkish history, subsidized by the Turkish government, was attacked without any regard for anything the proposed professor had said about the Armenian Holocaust or anything else. The opponents opposed the plan simply because the Turkish government was involved.
    Until the Turkish government did whatever they were supposed to do about events of 70 years earlier, these people would attack anything the Turks ever did anywhere.
    A disgraceful episode, in my opinion.

  10. jamessal says

    Rambler: Nothing specious about what Hat said. A person can be wrong about (A) and right about (B). If (A) is really off the walls, then, sure, another person would be more skeptical (B); but now you’re talking about that person’s subjective experience, not the objective merits of (B).

  11. I’d have to go digging for the exact passage, but Gibbon says something similar to this about the origin of the term when describing the fully developed Imperial establishment under the successors of Constantine–with the implication that the term was used in the early Byzantine Empire, which would strengthen the idea that the later Byzantines were merely borrowing a term used earlier, rather than borrowing from a contemporary European language. But I don’t recall him indicating what the Greek form was.

  12. Gibbon’s footnote: κονόσταυλος or κοντόσταυλος.

  13. Thanks—always a pleasure to read Gibbon!

  14. caffeind says

    How do you tell if Κοντόσταυλος is Kondostavlos or Kodostavlos, or if κοντός is kondos or kodos, given that ντ is the Greek spelling of D?
    The form Google suggests as most popular is Contostavlos, which appears to be a popular last name.

  15. caffeind says

    Britannica claims “The title comes stabuli is found in the Roman and particularly in the Eastern Roman, or Byzantine, Empire from the 5th century ad as that of the head of the stables at the imperial court. The Franks borrowed the title,”

  16. John Emerson says

    At this point, just for fun it’s worth looking up the words “stable”, “stall”, and “stale” in the OED.

  17. How do you tell if Κοντόσταυλος is Kondostavlos or Kodostavlos, or if κοντός is kondos or kodos, given that ντ is the Greek spelling of D?
    There’s not really a difference, if I understand the situation correctly; some Greeks pronounce ντ as /nd/, others as /d/, but the difference never distinguishes two words.

  18. Terry Collmann says

    At this point, just for fun it’s worth looking up the words “stable”, “stall”, and “stale” in the OED
    And, of course, “stallion”.
    The stallion that stood in the stall in the stables was stale, said the constable.

  19. Why not assume the word to be a direct loan from Vulgar Latin *COM(I)TEM STAB(U)LI (vowels in parentheses may have already been dropped in Vulgar Latin)? Many loanwords from Latin dating back to the Imperial era were borrowed in the form of the accusative singular: this etymon would have yielded Greek *KONDESTAVLI: The shift from /e/ to /o/ in the second syllable may well be a case of folk etymology due to perception of a relationship between this word and /kondos/. Likewise, I wonder whether its singular form (KONDOSTAVL-OS) might not be a back-formation from a form KONDOSTAVL-I perceived to be a (Greek) nominative plural (as otherwise a Greek nominative singular KONDOSTAVL-IS is what I would have expected to find: likewise, had the word been borrowed from Italian or Catalan I would have expected a nominative singular KONDOSTAVL-ES).

  20. Cherie Woodworth says

    Though I am not a Byzantinist, my friends who are have often referred me to the (now) definitive quick reference for all things Byzantine, the Oxford Dictionary of Byzantium. The online version, which is regularly updated, can be found here: http://www.oxford-byzantium.com/
    As the full entries may not be accessible to those who do not have a university portal, I will hazard to quote the entry, what may be rather lengthy for a blog.
    About konostaulos, Alexander Kazhdan of the ODB says the following:
    Konostaulos (κονοστα υ̑λος, from Lat. comes stabuli, “count of the stable,” Fr. connétable), a term that entered Byz. in the 11th C. under Norman influence. A 12th-C. historian (An. Komn. 2:28.5–7) speaks of a Latin phalangarches Bryenne “called konostaulos.” Guilland (Institutions 1:471) mistakenly ascribes a seal of an anthypatos Isaac (?) to a konostaulos of the 11th C.; in fact, the seal belonged to a komes tou staulou (St. Maslev, Izv Bŭlg Arch Inst 20 [Sofia 1955] 452f, no.3; Laurent, Corpus 2, no.924). The office/title, predominantly in the form megas konostaulos, is known only from the 13th C. onward. Pachymeres (Pachym., ed. Failler, 1:37.4–7) defines him as the commander of Italian mercenaries. The first megas konostaulos mentioned in the sources is Michael Komnenos Palaiologos under John III Vatatzes (Akrop. 1:134.10–11). In the hierarchy of pseudo-Kodinos the megas konostaulos follows the megas primikerios. From the 13th to the 15th C. members of noble families (Palaiologoi, Tarchaneiotai, Monomachoi) held this post as did Western seigneurs such as Licario of Verona; the title was also conferred on Leonardo Tocco.

  21. Cherie Woodworth says

    The Nature of the Early Ottoman State (thanks, Jim!), a brilliant reanalysis of the early Ottomans that proves (to my satisfaction, anyway) that far from being fearsome warriors for Islam who presented conquered people with the famous “convert or die” choice, they were highly pragmatic rulers
    On the formation of the early Ottoman state, the “ghazi” hypothesis (that the Ottomans were jihadis or religious warriors for Islam — ghazis) was, as far as I know, put forward convincingly by Paul Wittek in his book The Rise of the Ottoman (1938).
    It was a pioneering book for intellectual history and epigraphy, but not surprisingly, ripe for an update. In that, Lowry is in line with others currently working on the early Ottomans. Check Lowry’s bibliography and see if he cites the revisionists Lindner and Barkey ; see also Patricia Crone, Slaves on Horses.

  22. Thanks very much, Cherie, that’s most enlightening! And it’s interesting that they had both konostaulos and komes tou staulou.
    Lowry discusses all the scholars you mention, and his book is explicitly a refutation of Wittek.

  23. John Emerson says

    What I’ve read about the Ottomans, coming from the Turko-Mongol direction, has made it seem that they were much like many other coalition raiding armies of various times and places — military adventurers and swashbbucklers of mixed origin with predominantly material goals.

  24. Just to wrap up loose ends, here is the Acropolites reference as given by the ODB that Cherie Woodworth quotes, and here is that same passage in the volume to which I referred earlier. Vryonis evidently had some third edition, though the introductory materials indicate that CSHB is the default source.

  25. Thanks! I wonder if I’ll ever cease being amazed and grateful that all this stuff is instantly and freely available on the internet. If I’d wanted to check Acropolites in the old days, I’d have had to wait for the NYPL to dig an edition out of their dustiest stacks.

  26. caffeind says

    Probably too long a shot, but could the “kono” form have been influenced by the Slavic word for horse?
    >There’s not really a difference, if I understand the situation correctly; some Greeks pronounce ντ as /nd/, others as /d/
    But how do we decide whether to transcribe ντ in a Greek word into the Roman alphabet as nd or d?

  27. It should be transcribed as nd, because that’s the standard pronunciation.

  28. Well, except initially.

  29. Picky, picky.

  30. Lars (the original one) says

    Speaking of Greek transcription I just discovered that the company I work for (until Sunday) is marketing its Greek classified ads site as AggelioPolis — in Greek. Check your app store if you don’t believe me!

    AggelioPolis: ο καλεύτεροσ τρόπος να αγoραζετε και να πουλάτε διαδικτυακά Δωρεάν αγγελίες που διαβάζονται από χιλιάδες ’Ελληνες καθημερινά, στην AggelioPolis

    Are Greeks so used to furriners mangling their alphabet that they will read AggelioPolis out as /aŋʝɛli’opolis/?

  31. Either you or your phone have screwed up the Greek a bit: “καλεύτεροσ” should be καλύτερος and “αγoραζετε” s/b αγοράζετε. But never mind that — I’ve learned a new Greek word, διαδίκτυο ‘internet’ with its derivatives διαδικτυακός (adj.) and διαδικτυακά (adv. ‘online’).

  32. Oh, and I’m sure the Greeks don’t even bat an eye at “AggelioPolis.”

  33. Lars (the original one) says

    Actually I copied it by hand (well, keyboard) since the iTunes store didn’t make it selectable, so I can’t blame my phone. And my visual glitch finder doesn’t work for Greek text. I did copy my version to google translate and the result looked plausible — originally I missed the ο in αγοράζετε as well and that mangled the translation somewhat.

    Can also be seen outside of iTunes, though you might be denied access if the app is not available in your market.

    Anyway, I just asked a few of my colleagues here in Sweden, and they of course didn’t have the faintest suspicion that ‘agge’ could be anything by /’ag:e/ or possibly /’æʒi/.

  34. I’m sure the Greeks don’t even bat an eye at “AggelioPolis.”

    All systems of writing are only approximations of the sounds of the words, and some of Beattie’s [a prominent Linear B skeptic] arguments on this score are disingenuous. ‘Pylians,’ he writes, ‘… would hardly know what to make of pu-ro.’ There would be as much risk of a Pylian making a mistake as there would be of a Scot misreading E’boro.

Speak Your Mind