Shiloh, Silom.

This is one of those selfish posts of no general interest; I’m just hoping someone out there can satisfy my curiosity about a trivial etymological point. The Russian equivalent of Shiloh (the ancient city, Hebrew שִׁילֹה‎) is Силом [Silom]. The first part of the word is entirely understandable, because the Greek version is Σηλώ, which has been pronounced /silo/ since the Byzantine period, when the Russians would have borrowed it. (I say “Russians” because even Ukrainian has Шіло [Shilo], without the -m.) I asked Sashura what he thought, and he suggested it was contamination from Силоамская купель, the Pool of Siloam in Jerusalem. That’s a perfectly reasonable suggestion, and I’m provisionally adopting it to ease my mind, but if anybody knows anything more definite, I’m all ears. (Curiously, the -m of Siloam is not original, since the Hebrew is שִּׁילוֹחַ [Shiloakh]; it’s from Greek Σιλωάμ, and I guess I’m curious about why the Greeks stuck an -m on there as well.)


  1. Always surprised me too, would love to find an explanation. As to купель силомская, I don’t think it exists for real – looks like Google Book’s OCR failure of an “a” in the Orthodox-compliant font.

  2. Where are you getting “силомская”? I have “Силоамская” in my post.

  3. Well, that explains Milton’s “Siloa’s brook that flowed / Fast by the oracle of God.”

  4. OCR error exampes:

    I have “Силоамская” in my post
    Now *that* one is for real.

  5. Sorry the link was too long to paste it straight. Hopefully it works now…

  6. Yeah, those are clearly OCR errors — if you click through to the books, they have the а.

  7. Re: Siloam. Habakkuk also became Αββακούμ in Greek and Аввакум in Russian. I’m guessing Koine Greek was just really not into ending nouns in certain consonants, and this was the standard way of avoiding that?

  8. Looks like it, though one wonders why they would choose μ, which did not occur as a Greek final, rather than ν, which did.

  9. The present-day Arab village of Silwan gets its name, I presume, from Byzantine Greek.

  10. The Arabic name of the biblical town of Shilo is Khirbet Seilun. According to Bible Walks, the site was sparsely populated through the later biblical and through the Hellenic period, and only in late Roman or Byzantine times is there much sign of habitation.

    The first part of the name means ruin. The parallel word in Hebrew is חורבה khorba. At root is a word for sword, חרב kherev, also seen in the traditional Muslim division of the world into Dar al-Islam and Dar al-Harb. (Some say the carob fruit חרוב is so named because of its resemblance to a sword.)

    I don’t know enough about Arabic to say anything definitive as to how the second element of the name might have come about. It may have been a natural development from the original Hebrew, or it may be an adoption of a Roman or Byzantine name for the place.

  11. Perhaps /-m/ was a Greek convention for “funny consonant ending we can’t pronounce”, being a consonant ending they (as I suppose) could pronounce even though it didn’t happen to occur in native words.

  12. J.W. Brewer says

    I would expect (unless something odd went on, which is not impossible) that the toponym was mediated to both modern Russian and modern Ukrainian from LXX Greek via Church Slavonic. So that raises the question of what spelling Church Slavonic used and why (unless one of them got it via a different path) Russian and Ukrainian subsequently diverged. One might look at other languages who you would think might have also gotten their toponym via Church Slavonic (Bulgarian and Serbian, for starters) to see how they handle it.

  13. Church Slavonic has “Silom”, not sure if I can pit a link in Cyrillic-named domain here though:
    и принесша ко їисyсу въ п0лкъ въ силомъ.

    Ditto Bulgarian: Верно го чуваше ковчегот на заветот во Силом

    Ukrainian may be a Polish borrowing?

  14. Here we go: The Septuagint has Σηλώμ. (May be an incorrect accent on the omega.) Plugging that term into Greek Wiki, however, brings up not the place but a biblical personality: Shela, a son of Judah.

  15. J.W. Brewer says

    Modern Ukrainian ethnolinguistic nationalism has sometimes been mixed up with Catholic-Orthodox tensions, with the former often more likely to be nationalistic in a way that stresses differences from rather than commonalities with Russianness. Since the Ukrainian Catholics are overwhelmingly “Uniate” (a name now apparently thought in some circles to be pejorative for reasons that are unclear to me, but I don’t have a polite synonym at hand) and Byzantine-rite, the LXX/Church-Slavonic tradition ought to be normative for them on onomastic issues, but the possibility of influence from some busybody 17th century Jesuits who were more familiar with the Vulgate (which has “Silo”) than the LXX/Slavonic perhaps shouldn’t be discounted.

  16. Yes, that’s an excellent point.

  17. “Khirbat” (ruin) has no direct connection to “ḥarb” (war); the former has kh, the latter ḥ (which is original, cf. Ugaritic ḥrb “sword”, so it’s unlikely to be a doublet resulting from Aramaic influence). Kharrūb, carob, also has kh rather than ḥ in Arabic, though I don’t know its broader history.

  18. The Polish term for Pool of Siloam is Sadzawka Siloe, per

    “Ukrainian Catholics” all by itself entails “Uniates”, since the latter term means an Eastern Christian in communion with Rome, and no group that is not Roman calls itself “Catholic” (though to be sure the Orthodox affirm that they are a catholic church, just as the Catholic Church affirms that it is orthodox). So the Right Thing is to say that the overwhelming majority of Ukrainian Christians are Eastern Catholics. Or rather, it would be if it were true. In the Ukrainian diaspora, Eastern Catholics may well dominate, but only 15% of Ukrainian citizens who claim to belong to any church (some 62% of the population does not) are Ukrainian Catholics; more than 70% are divided between the three competing Orthodox Churches, and the remainder is the usual mix of Latin Catholics, Protestants, Jews, Muslims, and Buddhists.

    As to what is wrong with Uniate, well, mysterious are the ways of taboo. Why should colored people be seen nowadays as patronizing, and people of color as acceptable?

  19. As Lameen says, the roots for חרב ‘sword’ and חרב ‘desolation’ are ultimately unrelated. The first one, according to Klein, has an Akkadian cognate ḫarbu ‘a kind of plow’. The second one also means ‘dried up’, which I would guess is the source of the word for carob, whose pod looks shrivelled up.

  20. marie-lucie says

    Why should colored people be seen nowadays as patronizing, and people of color as acceptable?

    Because colored was used as an official euphemism for “non-white”, especially (in the US) ‘fully or partly of African origin’ and in South Africa ‘brown-skinned’ (mixed race, or most Asians)’. People of color must be a calque from French, where it has long been a descriptive term for people not obviously black or white, with no official connotation since the government did not (and does not) discriminate on the basis of appearance or origin.

  21. J.W. Brewer says

    I expect the “Ukrainian Catholic” percentage is somewhat higher in the percentage of the current Ukrainian population (67% if you believe wikipedia, and one can think of reasons why the underlying government census data might overstate that by a bit) whose L1 is Ukrainian and, more importantly, may have been higher still among the relevant players at the historical point when a standardized/normative orthography for “Ukrainian” that is self-consciously different from Russian orthography was being developed and promoted. Once upon a time in English, the Douay-Rheims translation used different spellings for quite a number of Old Testament proper names than the King James did, and at least some Anglophone Roman Catholics rather self-consciously used the Douay-Rheims variants. I have no idea whether the Ukrainian spelling as given above has any sort of sectarian/factional valence or is used universally, and of course the notion that the absence of the final -m is the result of sectarian factors it itself just a hypothesis I made up out of whole cloth earlier in the thread (which I think is plausible without necessarily asserting that it is >50% likely to be accurate).

  22. J.W. Brewer says

    marie-lucie, I don’t know if “people of color” in modern AmEng was a calque from French or an independent coinage, but presumably all that matters is that it it not widely perceived as having uncomfortable historical baggage. “Gens du couleur libres” (= free mulattoes who sometimes themselves owned slaves) is a term known by historians of race relations in Louisiana, which were sufficiently non-idyllic that the reason “persons of color” is not taboo due to historical baggage is probably simply the comparative obscurity of that history.

  23. I believe it is generally accepted that it was indeed a calque of gens du couleur, which was used throughout Francophone America. For instance, Wikipedia, Safire On Language.

    the government did not (and does not) discriminate on the basis of appearance or origin.

    I’m sure you know that this is all a matter of degrees.

    For instance, from Louis XVI:
    défense d’amener aucun noir, mulâtre, ou autres gens de couleur.
    défenses à ses sujets blancs de contracter mariage avec des noirs, mulâtres, & autres gens de couleur.
    ils conservent toujous la tache de l’esclavage.

  24. Because colored was used as an official euphemism

    I’m sure JC knows the history perfectly well; he’s simply making the point that, setting aside the history, there seems no obvious reason to prefer one phrase over the other.

  25. I was speaking only of the U.S. context. WP says that although “citizen of color” is found as long ago as 1963, the commonplace use of “person of color” only dates to about 1980. It’s possible that it is a calque, but I doubt it.

  26. MMcM, yes, the Code Noir was pretty bad, but it applied in the Caribbean colonies, not in France. I did not know about the other provision, which applied within France. Slavery was illegal in France, so the provision might have been intended to prevent immigration of fugitive slaves. The Code was abolished with the Revolution, but restored by Napoleon (his wife Joséphine being from a planters’ family), something which precipitated the revolt and ultimate victory of the slaves in Saint-Domingue and the independence of Haiti.

    Some French families who came back to France still sometimes brought slaves with them, (especially children’s nurses, I think) who were free once on French soil. The family of Alexandre Dumas is one example: his grandfather was a planter who had some children with one of his slaves. Having to come back to France for some business, this man took with him the eldest of these children (needing money, he sold the mother and the other children: I think he sold them to a friend so that they would still be cared for). This child was brought up in France and ultimately became one of the top generals of the Revolution, but he did not get along with Napoleon (whose reinstatement of the Code cannot have helped their relations). He married a French woman and they had one child, who became the famous novelist.

  27. Another clue could be in that Shiloh is two – the place-name of where the Ark of the Covenant was and the name of a biblical figure, the messenger, the one who professed the Messiah. The two may have had different paths of adoption into other languages.
    The French page on homonymie says that Shiloh means havre de paix – haven of peace. Which makes it suspiciously close to shalom/salam. I am not sure if it’s another plausible clue.

  28. As Lameen says, the roots for חרב ‘sword’ and חרב ‘desolation’ are ultimately unrelated.

    My bad. Occasioned by engaging hands at keyboard before brain fully in gear, aka not checking sources before posting.

    Yet, yet . . .(from Wiki): “ח Heth originally represented a voiceless fricative, either pharyngeal /ħ/, or velar /x/ (the two Proto-Semitic phonemes having merged in Canaanite). In Arabic, two corresponding letters were created for both phonemic sounds: unmodified ḥāʾ ح represents /ħ/, while ḫāʾ خ represents /x/.” So the two different Hebrew roots חרב must share identical spelling, though not so in Arabic.

    Hebrew also has כ kaph, which can be pronounced as a voiceless uvular fricative or as a voiced velar plosive depending on position within the word. Again from Wiki: “(Arabic) كـ kaf is almost universally pronounced as the voiceless velar plosive /k/, but in rural Palestinian, Iraqi, Kuwaiti and Gulf Arabic in general, it is sometimes pronounced as a voiceless postalveolar affricate.”

  29. The French page on homonymie says that Shiloh means havre de paix – haven of peace.

    Probably not. The idea seems to be that שִׁילֹה‎ comes from the root š-l-w, which gives words like shalev “tranquil” and shalva “tranquility”, but I don’t think this is morphologically possible (that final w can’t just disappear). In any case, this root is presumably unrelated to shalom, which is from a root meaning “complete, whole”.

  30. A very earnestly liberal college I once taught at had recycle bins labeled PAPER OF COLOR. I kid you not.

  31. Surely that was in jest, not in earnest.

  32. TR, there is another form of the root, šlh, with Aramaic cognates and the meaning ‘to be tranquil’, so the etymology is not formally impossible. I am still skeptical about a Biblical town called ‘haven of tranquility’ or whatever, because real estate developers didn’t exist back then, and old place names are usually more prosaic.

    More on Shiloh and Siloam. First, how did šiloax and šiloh yield, respectively, Arabic silwan and seilun? Second, I note the adjective šiloni שִׂילֹנִי ‘person of Shiloh’, in the name of Ahiyah the Shilonite (1 Kings 11:29, @ Chr 9:29). Assuming that is really the meaning of šiloni, does this n have anything to do with the -n in either the Arabic or the Greek versions? Lastly, a search for words in the LXX ending in μ yields several other proper names with for -n, for example Γοσομ for Goshen (Joshua), Εδεμ for Eden, Μαδιαμ for Midian, etc. However, if Shiloh and Shiloah acquired a Greek neuter suffix -ν, it’s hard to imagine them undergoing the same process.
    (Hat, so much for your “trivial etymological point”…)

  33. “why they would choose μ, which did not occur as a Greek final, rather than ν, which did.”
    The same question might be asked about Eden, which is pronounced in Greek as Edem, Εδέμ.

  34. ah! that’s what I was thinking from the very beginning! could it be that the -m is simply the Genitive? Of Shiloh, of Siloa?

  35. The problem with that theory is that Hebrew doesn’t have a genitive… But maybe what happened was that final -m (which would have been quite marked to Greek ears) got associated with Hebrew names generally, based on relatively familiar words like Jerusalem, Abraham, etc., and all the plurals in -im, so that it got extended where it didn’t belong. “Oh, you know those Hebrew names, they all end in -m — it must be Ἐδέμ.”

  36. TR, I bet you’re right, unless -m > -n is abundant as well, which I haven’t checked.

    Greek without the diacritics, as I rahly copied/pasted it, looks odd.

  37. TR, I love this theory so much I’m ready to accept it, even though it sounds more as a joke than a theory.

    In Sholokhov’s ‘Virgin Soil,’ there is a character who, in expectation of the world revolution, spends evenings studying English. ‘Oh, they’ve taken lots of words from us,’ he says, ‘proletariat, communism, revolution. Except they slap weird endings on them, sounds like hissing. They hate the revolution, so they say revolushshn.’

  38. @John Cowan: This was a college where a sense of humor is regarded as politically and culturally suspect.

  39. On Ἐδέμ; Saenz-Badillos’ A History of the Hebrew Language mentions assimilation of final mem and nun from Qumran on (page 140: ‘”As in R[abbinic] H[ebrew], final _mem_ and _nun_ are interchangeable”) – presumably merging as [n], perhaps under some Aramaic influence?. So it is possible that there was some hypercorrection going on, as well as the association of the exotic -m with Hebrew in a Greek mind, which is very likely; it reminds me of medieval Latinists trying to put extra ‘h’s and ‘y’s in words known to be Greek, irrespective of the original Greek spelling.

  40. An excellent comparison, and I agree, association of the exotic -m with Hebrew is very likely the explanation.

  41. Re: exotic -m ending also see this LH thread about cherubim as singular (which is actually shared in Ukrainian too)
    and also about phares / parsin interpreted as plural for “Persian”

  42. “Your comment is awaiting moderation” – what, links from this very site are suspect now? 😉 Both links I pasted were LH’s old threads

  43. It’s not the quality of the links, it’s the quantity. More than one link puts you in quarantine no matter what. Fortunately, Hat the Liberator comes along fairly quickly to rescue your posting.

    I have mostly taken this as a challenge to provide only the most salient link, or simply accepted the delay. When your comment is approved, it will appear in date order like all others.

  44. It’s not the quality of the links, it’s the quantity. More than one link puts you in quarantine no matter what.

    Well observed! I had no idea why some comments wound up in moderation and others not. (I still don’t understand why some comments that essentially go “spam spammity spam spam SPAM SPAM!” wind up in moderation; why can the sophisticated Akismet software automatically reject zillions of spam comments but be unable to make up its mind about these?)

  45. Akismet (which is short for Automattic [sic] Kismet, by the way, where Automattic is the name of the company) uses an algorithm that is intentionally kept secret; if it were known, it would be much easier to defeat it. As such, improvements can be made only by the (surely) small number of Akismet developers. Still we persist in wondering whether folly must always be our kismet.

  46. A very nice piece; why do I always forget about Pangborn when coming up with lists of good sf writers to recommend to people who are not steeped in the conventions of sf?

  47. I looked some more into the Shiloh/Shiloah issue. The most helpful source has been Yoel Elitzur’s magisterial book Ancient toponyms in the Land of Israel, which plows deeply in the field of place names from antiquity, through the Aramaic and the Greek, and to the present-day Arabic, although it does not discuss Shiloh and Shiloah in detail.

    First, the -m suffix is not just a Septuagint oddity. -n and -m suffixes begin to appear after vowel-final names and other words in Jewish sources several centuries earlier, e.g. Megiddon for Megiddo in Zachariah 12:11, and Keisrin for Caesarea in the Mishna. The source and the phonetics of these finals are disputed.
    A paper by Kutscher, which I haven’t seen, suggests that the proximate source for the Greek Σιλωάμ was not the Hebrew šilˈloaħ שִׁלֹּחַ, but rather the later Aramaic šilōˈħa, presumably to explain the long ω and the final stress. The shift š>s predates the Septuagint in the local speech south of the Galilee, and is regular. The ħ regularly becomes zero in Greek, so that fits too.
    For Shiloh, i.e. šiˈlo שִׁלֹה (note the unusual ungeminated l), the story is similar. We start with the variant form šīlō, which appears already in Judges 21:21, which yields the Septuagint Σηλω ~ Σηλωμ as above, but now with η for the long ī.

  48. Ah, thanks. So Shilo/Силом is from Aramaic and Siloam is from Hebrew?

  49. I add my thanks — excellent detective work, and it’s good to know the final nasal is a Hebrew/Aramaic thing.

  50. Sashura, not exactly. Shilo/Shiloah are the older biblical Hebrew forms. Silo/Silom and Siloam are later forms, with Siloam < Aramaic šiloħa < Hebrew šiloaħ. But if the Hebrew šiloaħ had undergone the same transformations without passing through Aramaic, I guess it would have ended up as Σιλόαμ, with short o and penultimate stress, which would then turn into Latin/English Siloam and Cyrillic Силоам just the same.

  51. David Marjanović says

    but now with η for the long ī

    Good to know – it’s somewhat earlier than I’d have expected.

  52. It’s irregular: usually both i and ī show up as Greek iota. But in a few cases, ī is reflected as a diphthong or η.

Speak Your Mind