HANC PONTEM.

Allan Metcalf tells the sad tale of “Professor William Merrill, a distinguished scholar who is said to have read almost everything extant in classical Latin,” who made a schoolboy howler in an inscription for a bridge: “Hanc pontem dono dedit classis studentum quae in anno MDCCCCX foras exiit ne memoria sua apud posteros pereat” has the feminine hanc instead of the correct masculine hunc. “And once the arch was up, it didn’t take long for someone to notice. According to Professor Joseph Fontenrose in his memoir Classics at Berkeley: The First Century 1869-1970, ‘At once his error was pointed out, and someone said that this was the only feminine bridge in the world.'” But “Merrill defended the gender as written, having found feminine pons in some late ancient or early medieval writings (perhaps in Hisperica Famina, which has female bridges).” For the exciting conclusion to the story, visit Metcalf’s post!

Comments

  1. dearieme says

    “a schoolboy howler”: getting stuff to agree wasn’t the problem when I was a schoolboy – it was fighting off the waves of sheer bloody tedium. Why Latin was the worst taught subject I don’t know but the phenomenon didn’t seem to be restricted to my school. I suppose it had a rival in “gym”: whether that was widespread too I don’t know. So much for two of the legacies of the classical world.

  2. From this photo of the arch and bridge, you can see that the design was based on Trajan’s Arch. They have roughly similar proportions, steps leading up to them (it would have to be a ramp nowadays), they both have the enormous inscription on the frieze, decorative pilasters on the face of the archway (meanly-projecting ones at Berkeley), and parapet walls that run perpendicular to the archway.
    The two things I noticed were a) how the details have been ignored or reworked to the cost of the design – see how the parapet walls at Berkeley run into the pedestal of the arch so that the arch has nothing to sit on, whereas at Ancona the parapet loops around the outside of the arch, making a sturdy looking base; and b) how much money people will spend to build monuments to themselves – see the size of the arch relative to the size of the bridge over the creek.

  3. “how much money people will spend to build monuments to themselves”
    Let’s be thankful for that – if people wouldn’t like to put up monuments and leave inscriptions commemorating their existence, we would know even less than we know about history and about many ancient languages!

  4. AJP “Rosie” M. Banks: it was designed by Brown and Bakewell, former students of Maybeck’s there.

  5. These things happen. Part of the Treaty of Paris (1814) saw the island of Malta handed over to the British empire, resulting in a suitable Latin inscription in Valletta. The plaque was re-carved in 1851 to read: MAGNAE ET INVICTAE BRITANNIAE MELITENSIUM AMOR ET EUROPAE VOX HAS INSULAS CONFIRMAT A.D. MDCCCXIV. (It should read CONFIRMANT.) What gave? There was a brief discussion of this in the Times of Malta two years ago, correspondents taking the matter to heart, as indeed they should. See here, here, here, here, here, and here for more.

  6. (Then again, perhaps that “should” of mine should be in quotes.)

  7. saw the island of Malta handed over to the British empire
    Thus pissing off the Russians; the tsar considered himself the protector of the Knights Hospitaller.

  8. The first commenter wrote ‘[I]n the inscription STUDENTUM should be STUDENTIUM, as the normal present participle of STUDEO.’ But in fact the inscription does say ‘studentium’; see http://bltnotjustasandwich.files.wordpress.com/2012/06/hunc_thumb.jpg . Alan Metcalf mistranscribed it. Muphry’s (Skitt’s, Hartman’s, Davidson’s, White’s, Studiolum’s) Law strikes again!

  9. Funny, the Arch of Trajan actually looks rather gracile. I always pictured something overmuscled like the Arch of Titus > the Arc de Triomphe > Washington Square Arch. The last of these is in my own back yard, so to speak; alas, the last I saw it you couldn’t walk under it any more, perhaps to keep the graffiti off (it had to be cleansed in the 1980s), or perhaps to keep out suicide bombers.

  10. Here is the article copy / pasted with better photos.

  11. Thus pissing off the Russians; the tsar considered himself the protector of the Knights Hospitaller
    Not just protector! Paul was actually elected Grand Master by some renegade knights, mostly French, after their expulsion from Malta (1798)- but the title was never recognized by the pope. Paul set up a parallel order for Orthodox Christians and handed out crosses to worthy men and their ladies (Lady Hamilton, for one). After Paul’s assassination (1801), Tsar Alexander told the knights to find their own Grand Master, and dissolved his own crew once they did.
    I’d not heard that he was particularly undone by Britain’s taking Malta, however. I mean to say, the knights themselves hadn’t been there for sixteen years, and they were more Paul’s hobby than Alexander’s. Or at least, so I understood it. Perhaps Alexander wanted its magnificent harbor for themselves? Tell me more!

  12. No, you’re right; I was remembering the start of War and Peace (“Англия с своим коммерческим духом не поймет и не может понять всю высоту души императора Александра. Она отказалась очистить Мальту.”) and assumed Alex would have been upset by the treaty, but (according to Malta, Britain, and the European Powers, 1793-1815, by Desmond Gregory) “He renounced any Russian claim to Malta and showed no further interest in the islands when Napoleon was overthrown.”

  13. Dearieme,
    “Why Latin was the worst taught subject I don’t know but the phenomenon didn’t seem to be restricted to my school.”
    That happens all the time with language instruction and it’s a goddamned shame. It’s theft – a student puts in his time and deserves to get some decent instruction and come out with some skills. If he doesn’t get that, he has been cheated.
    John – how hard can it really be to determine the sex of a bridge? After all you just have to walk up under it and look.

  14. The Order of Knights Hospitaller, or Malta, or St. John, now exists as the Sovereign Military Order of Malta, the Catholic branch, and four Protestant branches: German, Dutch, Swedish, and British organizations descended from St. John commanderies that adopted Anglican or Lutheran or Reformed theology but maintained continuity with the Catholic past. It’s debatable whether the Orthodox branch still exists, and if so, which of the various claimant organizations is actually the real thing.
    The Sovereign Military Order of Malta has special status under international law. Since 1798 it is no longer a territorial state, nor a government in exile either, having renounced all claim to Malta. But it is still a sovereign entity with internal self-government legislative, executive and judicial, beholden to no nation for its existence. (Compare the International Committee of the Red Cross, which is also de facto independent but legally is just a private Swiss corporation.) The SMOM has diplomatic relations with 104 countries plus the E.U., and observer status at the U.N. (as does the ICRC). It issues passports, license plates, coins (scudi), and euro-denominated stamps. At one time it even had a small air force.
    There are a lot of other bogus St. John orders around who have no history and are in it for the money.

  15. (And “U” isn’t any harder to chisel than S, for example.)

  16. John – how hard can it really be to determine the sex of a bridge? After all you just have to walk up under it and look.
    But that’s just what we New Yorkers can no longer do!

  17. But “Merrill defended the gender as written,

    Of course he did. So few ‘experts’ have the gonads to admit their mistakes.
    The mark of the good scientist is their ability to do so, but in reality what happens is that the people who are wrong just die eventually, and the world moves on.
    Look up the story of the determination of the charge of the electron some time. And weep.

  18. The Sovereign Military Order of Malta has special status under international law. Since 1798 it is no longer a territorial state, nor a government in exile either, having renounced all claim to Malta. But it is still a sovereign entity with internal self-government legislative, executive and judicial, beholden to no nation for its existence. (Compare the International Committee of the Red Cross, which is also de facto independent but legally is just a private Swiss corporation.) The SMOM has diplomatic relations with 104 countries plus the E.U., and observer status at the U.N. (as does the ICRC). It issues passports, license plates, coins (scudi), and euro-denominated stamps. At one time it even had a small air force.
    The things I learn around here!

  19. marie-lucie says

    But “Merrill defended the gender as written,
    Perhaps he had been immersed for a while in Late Latin from a remote corner of the empire, and forgotten a lot of the Classical Latin which is what the inscription was supposed to be in.
    It would be interesting to know how he would have graded the work of a student who wrote hanc pontem.

  20. Wikipedia tells us the Order held Malta from 1530 to 1798, and that it defended the island against an Ottoman invasion force in 1565. The last is a rousing story I read many decades ago.

  21. befuggled says

    Also per wikipedia, in 1998 the SMOM made an agreement with the Maltese government to return to Malta. While their HQ is apparently still in Rome, they apparently have a 99-year lease on Fort St. Angelo in Birgu.

  22. The last is a rousing story I read many decades ago.
    I’m guessing in the form of Ernle Bradford’s The Great Siege, a ripping yarn.

  23. If Merrill knew Welsh he might have been misled by the fact that the derivative is feminine, cf. Pen-y-Bont. Perhaps a borrowing from “Late Latin from a remote corner of the empire.”
    @iakon: Was that rousing story Chesterton’s “Lepanto”?

  24. stormboy says

    Interesting (to me, at least) that ‘bridge’ is masculine in French and Spanish but feminine in Portuguese (a ponte). But I’m not implying he had a knowledge of Portuguese (which might have caused some interference in his translation).

  25. marie-lucie says

    stormboy: ‘bridge’ is masculine in French and Spanish but feminine in Portuguese (a ponte)
    Precisely my point: Portugal is more remote from Rome (where the word was masculine) than Spain and France. Pont (currently pronounced [pun(t)] is also masculine in Occitan.
    What about other cognates (eg Italian dialects, Rumanian)? Do they all have ponte (or a cognate), and if so, what gender is it?

  26. Romanian has pod, which while it looks enticingly like a cognate is in fact one of the many borrowings from Slavic (pod ‘under,’ from PIE *pod– ‘foot’). For what it may be worth, it’s neuter.

  27. It. ponte is masculine. Spanish puente is also sometimes feminine, historically and dialectically. Witness La Puente in L.A.
    This confusion over the third declension can sometimes even lead to the same word having both genders in the same language, sometimes for different senses. For example, French gens.

  28. marie-lucie says

    MMcM, les gens. Yes. This is one of several words which are one gender in the singular and another in the plural.
    gens (pronounced like the French name “Jean”): That word, a plural, started out as feminine in the singular: la gent ‘the people (of a nation, ethnic group, etc”), a usage which is now obsolete except in a jocular literary context, where it is followed by an adjective: la gent moutonnière could mean ‘the sheep, the sheep “race” or “nation”‘ (un mouton ‘a sheep’), or (more likely in the modern world) ‘people who act like sheep, blindly following a leader or fahion’. The word has remained feminine in the plural in a few set phrases where it is preceded by an adjective (an archaism) eg les bonnes gens ‘the “good old” people’, les petites gens ‘the “poor old” people, people of limited means and outlook’ (not “old people who are good/poor” but with the English connotation of “old” as an affectionate but somewhat deprecative term). These phrases are now rather old-fashioned, and otherwise the very common les gens ‘people’ (in general) is masculine, eg les gens du Midi ‘(the) people from Southern France’, les gens mal intentionnés ‘(the) people with nasty intentions, always ready to think the worst of other people and cause them trouble if possible’.
    Another such word is oeuvre, from the Latin neuter opus, operis, which became a French masculine: in building construction le gros oeuvre means ‘the basic structure’ (the foundation, main walls and roof beams), and there are other archaic usages in the technical vocabulary of some professions. In old compounds such as chef-d’oeuvre ‘masterpiece’ (lit. ‘head/top of the work’, hors-d’oeubre ‘appetizer’ (lit, ‘outside the (main) work’) and main-d’oeuvre ‘manpower’ (lit. ‘hand of work’) it seems that the word was still masculine, but it is impossible to tell since oeuvre here is not the head word and does not take an article. Otherwise, oeuvre is feminine (from opera, the originally neuter plural Latin form, mistaken as a feminine singular) and it means mostly ‘(a) work’ (of an author or artist)’ or ‘(a) charity’ (set up to work for the good of society). In the plural, les oeuvres (still feminine) can mean ‘the works’ (of an author, etc – not what you might put on a pizza!), or ‘the charities’.

  29. Sir JCass says

    Romanian has pod
    Romanian also has punte. It seems to be mainly used for the bridge of a ship, but you also get things like Puntea Suspinelor (“The Bridge of Sighs”). The gender of punte is feminine.

  30. There is also La Puente restaurant chain in Utah, and La Puente shelter in Colorado

  31. marie-lucie says

    Perhaps the shift from masculine to feminine for ‘bridge’ (Lat pons, pontis) in some areas was favoured, if not triggered, by the resemblance with the word for ‘well, fountain’ (Lat fons, fontis). Both had to do with a location near water, and in the past fountains and wells were very much more common than bridges. There are many places in Spanish lands called La Fuente.

  32. marie-lucie says

    Somehow this discussion of gender change (NOT “transgendering” in the modern sense) makes me think of the equivalence of the male name Pierre ‘Peter’ and the feminine noun pierre ‘stone, rock’ in French, and the pun in one of the gospels (which I learned in catechism as a child), when Jesus allegedly said to his disciple Simon: Tu es Pierre, et sur cette pierre je bâtirai mon Eglise, as translated from Latin (I rquote from memory) Petrus es, et super hanc petram edificabo ecclesiam meam, literally ‘You are Rock, and on this rock I will build my church’ (I don’t know how the English translations deal with this). Latin petra, borrowed from Greek, is feminine, as signaled by the final -a, and in order to make it into a male name the ending has been changed into a masculine one, -os in Greek, borrowed as the Latin equivalent -us. (The name Petros must have been around long before Jesus allegedly used it).
    Now Jesus probably did not speak Latin but had at least rudiments of Greek (the lingua franca in the area) if he knew about Petros, petra and ecclesia (assuming the quotation is genuine and not invented by a Greek-speaking Gospel writer). The TLFI says that Petros was a nickname given by Jesus to Simon, perhaps because the name Simon resembled Aramaic (or grecified (?) Aramaic) kêphas “pierre” (sorry, I can’t reproduce all the Greek transcriptions given in the TLFI under pierre). If so, this suggests that initial s:k and medial m:ph or m:f are or were regular correspondences between Hebrew and Aramaic. Does anyone know more about the quotation? Are these explanations plausible?

  33. I don’t know how the English translations deal with this
    Basically by ignoring it. All of the ones I have checked on line say something close to “Thou art/You are Peter, and upon this rock I will build my church” (Matt 6:18), leaving the reader to check a footnote or a separate commentary for an explanation. The second half, like the parable of building on sand (Matt 7:26), clearly alludes to the Middle Eastern practice of constructing buildings only on stone outcrops in order to prevent the foundation from being undermined by wind or rain.
    The name Petros must have been around long before Jesus allegedly used it
    I am not convinced of that; I think it got into the world’s stock of names solely on the strength of this story. I note that in Bulgarian, Камен ‘stone’ (cognate with hammer) is also a proper name, and considered equivalent to Петър (note that in the Bulgarian alphabet, ъ is the vowel /ə/, not a mere sign of non-palatalization).
    Jesus […] had at least rudiments of Greek
    More than that, if his examination before Pilate is to be believed. Greek is the only language in which they could reasonably have communicated. In addition, Jesus is said to have preached in the Decapolis, a region of ten Greek-speaking cities in the eastern part of Roman Judaea, and if so, surely he did so in Greek. This is where the incident of driving the devils into a herd of swine (Mark 5:1-10) takes place: no Jew or Samaritan would keep pigs.
    TLFI says that Petros was a nickname given by Jesus to Simon, perhaps because the name Simon resembled Aramaic (or grecified (?) Aramaic) kêphas
    Definitely Grecized. The original Aramaic form כיפא did not end in -s: the Greeks usually added -σ to masculine foreign names which they heard as ending in a vowel (thus Mωϋσῆς, Moÿses, Moses for Hebrew מֹשֶׁה,‎ Mosheh) to make them look more Greek and less feminine. As you note, there is a gender difference in the Greek text: ὅτι σὺ εἶ Πέτρος καὶ ἐπὶ ταύτῃ τῇ πέτρ οἰκοδομήσω μου τὴν ἐκκλησίαν (emphasis added), but not as far as I know in Aramaic.
    But any such phonological resemblance as the TLFI postulates seems very unlikely, even farfetched. I think the reference was an ordinary metaphorical one: calling someone a “rock” to refer to their loyalty, stability, and strength is very common in English to this day, and clearly most people who say so are not thinking of Peter, since the linguistic connection is not apparent in English. The best colloquial translation of Jesus’ words is probably “You’re a rock” or (at least to people of my generation) “You’re my rock, man.”
    In a few places in the New Testament (John 1:42, 1 Cor 1:12, Gal 1:18) the name appears transliterated as Κηφας (and so Cephas in the KJV and other English Bibles) rather than as Peter. This is sometimes explained away as being a different Cephas altogether, but I think that’s unlikely, especially if Cephas/Petros was not yet an ordinary name but a nickname given to a specific person. I don’t know whether כיפא ever became an ordinary name among Aramaic-speaking Christians.
    (By the way, you can type Greek letters here at LH and elsewhere on the Web by typing &theta; to get θ and &Sigma; to get Σ, analogous to &lt; for < and &amp; for &.)

  34. Hmm. So we’ve Portuguese, Romanian and Welsh (with a Romance loan!) with feminine gender on the one hand, and French, Italian and Spanish (mostly) with masculine gender on the other. The geographical distribution of the gender of the noun in Romance-speaking Europe needs to be looked at more closely before any conclusion can be drawn, but the above distribution is suggestive: feminine gender in the peripheral languages, masculine gender closer to the center. Could PONTEM have originally been feminine in spoken Latin, and only became masculine in more central areas at a later date?

  35. Addendum: If the Wikipedia articles on “Bridge” in various languages are to be believed, the Breton reflex of PONTEM is masculine, in Galician and Extremeno it is feminine, and in Aragonese, Normand, Catalan, Occitan, Sicilian and Ladino it is masculine. This distribution of masculine versus feminine gender is quite compatible with my guess (masculine at the center, feminine at the periphery) above.

  36. But any such phonological resemblance as the TLFI postulates seems very unlikely, even farfetched. I think the reference was an ordinary metaphorical one
    I agree with JC.

  37. The TLFI says that Petros was a nickname given by Jesus to Simon, perhaps because the name Simon resembled Aramaic (or grecified (?) Aramaic) kêphas “pierre” (sorry, I can’t reproduce all the Greek transcriptions given in the TLFI under pierre). If so, this suggests that initial s:k and medial m:ph or m:f are or were regular correspondences between Hebrew and Aramaic. Does anyone know more about the quotation?
    Simon (Shimmon is a better approximation of Hebrew שמעון) is one of the sons of the biblical patriarch Jacob. The name is based on a root that carries the sense of hearing. He was so named because God heard his mother’s plea (Genesis 29:23 per the traditional Hebrew numbering).
    Hebrew Sh ש is sometimes T ת in Aramaic and Arabic.
    D ד and z ז at least some of the time correspond in Hebrew, Aramaic, Arabic and Maltese. In Hebrew, gold is zahav זהב while in Aramaic and Arabic it’s dahab דהב, and in Maltese it’s deheb.
    Hebrew Ts צ sometimes becomes Ayin ע in Aramaic.
    I’m not aware of an M:P(H) shift, but that doesn’t mean there isn’t one.
    Aramaic kepha כיפא appears only twice in the entire Talmud. The references don’t seem to be to a personal name or to a rock. The word does not appear in the standard Aramaic-Hebrew dictionaries used as Talmud study aids. There are several words in Hebrew based on כפ, but they have nothing to do with hearing or rocks.
    There seems to be an Armenian cemetery near Jerusalem’s Mount of Olives whose name in Hebrew sources is בית כיפא Beit Kepha (House of Kepha).
    Grecified: How about Hellenized?

  38. A check of Onkelos shows that English Simon and Hebrew שמעון is Aramaic שמעון as well.

  39. marie-lucie says

    Thank you both, JC and Paul O. The TLFI is occasionally unreliable.
    Yes, “Hellenized” would be more elegant than my nonce form “grecified”.
    I used to know the Hebrew alphabet, but have forgotten most of it. Nevertheless I think I see a ayin in the middle of the word for Simon, is this why the name in the Bible is usually Simeon?
    Hebrew Ts צ sometimes becomes Ayin ע in Aramaic.
    This is extremely odd from a phonetic point of view. Could it be due to spelling errors, since the two letters resemble each other?

  40. marie-lucie says

    Etienne: So in Latin we find:
    fons, fontis, fontem, etc ‘well, fountain’, f. > Oc font [fun(t)], It fonte, Sp fuente both f.
    mons, montis, montem, etc ‘mountain, peak’, m. > Fr mont, Oc mont [mun(t)], It monte, Sp monte, both m. (No ue in Sp here, because of the initial m ??]
    pons, pontis,pontem, etc ‘bridge’ m. (or f. depending on the time and place?). Fr pont, Oc pont [pun(t)], It ponte, Sp puente, all m. except occasionally Sp f.
    Any more words of this shape, which could have influenced the others?

  41. frons, frontis gives It. fronte and Sp. frente, which are f. when they mean ‘forehead’ and m. when they mean ‘front’ (battle or weather). It was both classically, but without a clean semantic separation, I believe.

  42. Hebrew Ts צ sometimes becomes Ayin ע in Aramaic.
    This is extremely odd from a phonetic point of view. Could it be due to spelling errors, since the two letters resemble each other?
    The letters somewhat resemble each other today, but that was not always so. Compare the early forms at this Wiki entry. (Note also that the form of tsadi צ changes to ץ when it appears at the end of a word.)
    Hebrew for land, country or ground is eretz ארץ. In Biblical Aramaic it’s ara’a ארעא. I’m no specialist in these matters, but it seems very unlikely that a spelling or copying error would arise with respect to such a basic word. Further, the Aramaic appears in a very well known segment of the Passover service:
    הא לחמא עניא די אכלו אבהתנא בארעא דמצרים. כל דכפין ייתי ויכול, כל דצריך ייתי ויפסח. השתא הכא, לשנה הבאה בארעא דישראל, השתא עבדי, לשנה הבאה בני חורין.
    Which in Modern Hebrew is:
    זהו לחם העוני שאכלו אבותינו בארץ מצרים. כל הרעב יבוא ויאכל, כל הצריך יבוא וייפסח. השנה (אנו) כאן, לשנה הבאה בארץ ישראל, השנה (אנו) עבדים, לשנה הבאה בני חורין.
    And in English:
    This is the bread of affliction which our ancestors ate in the land of Egypt. Let all who are hungry come and eat. Let all who are in need come and celebrate Passover. This year we are here: Next year, in the land of Israel. This year we are slaves: Next year may we all be free.
    Curiously – or perhaps not to the well-informed – tsadi צ and ayin ע can also alternate with kuph ק. Modern Hebrew for (electrical) ground was for some reason taken from Biblical Aramaic ארק arak, land or country (not to be confused with the ouzo-like beverage of the region, known as raki in Turkish, and of which even small quantities have been known to cause big confusion).
    Klein’s entry on ארץ gives Ugaritic qrs, Arabic arḍ, Akkadian erṣetu and Tigre ’ard. He also notes that Arakiel is the name of the earth in the Book of Enoch at 8:3.
    Nostraticists take note: Arabic ard is suspiciously close to English earth and German Erde.
    Curiouser and curiouser: The Wiki entry on Arakiel, referring to a certain fallen angel, notes an alternative name Aretstikapha for Arakiel: “(meaning ‘world of distortion’ [the combination of arets + kaphah])” in Enoch 69. Kaphah in this instance is spelled ארסטיקיפה, though Kepha(s), the Aramaic name for Simon Peter noted higher up in this thread, is spelled כיפא.

  43. marie-lucie says

    MMcM: frons, frontis gives It. fronte and Sp. frente, which are f. when they mean ‘forehead’ and m. when they mean ‘front’ (battle or weather). It was both classically, but without a clean semantic separation, I believe.
    Very interesting. French front also means both, but is always m. There is also le front de mer, the part of a seaside city which faces the sea.

  44. marie-lucie says

    Paul O, forgive my display of ignorance. My acquaintance with the Hebrew language, let alone texts, is extremely limited.
    However, speaking as a historical linguist and dealing with generalities which may or may not apply to the situation we are discussing: Sometimes when it looks superficially like “X in language A becomes Y in language B”, this is not because of the evolution of X to Y (or vice-versa) but of a correspondence of the two elements with respect to another, older element Z from which both derive, each of X and Y having evolved separately but still haring some features with Z. An example recently discussed is English wh and Swedish etc v, neither of them “becoming” the other but both deriving from the evolution of PIE *kw, all three consonants sharing the property of being labial but differing in other properties.
    In a large number of languages having the complex consonant /ts/ (functioning as a sound unit or “phoneme”), older forms (if known) and/or at least some related languages have /k/ or /ky/ instead. In the two names of the angel Arakiel, one beginning with arak and the other with arets-, the first form has the word before a vowel, the second one before a consonant, and that could be (as a hypothesis since I don’t know all the facts) the reason for the difference. In any case it is not at all unexpected that Aramaic arak ‘land, country’ should be cognate with Hebrew erets which means about the same thing. On the other hand, it is not possible to link Hebrew ts directly with Aramaic ayin, but Aramaic k and ayin may both derive from another sound such as q (a uvular) or oher “throaty” consonant. Ugaritic qrs also belongs in this group of words through its final consonant: s often comes from the simplification of ts. (As for the initial consonants, there is probably another correwpondence between Ug initial q and the aleph of Hebrew and Aramaic).
    These hypotheses should be checked with accepted Proto-Semitic reconstructions since I am not familiar with them: adding other languages to the mix might show whether they are correct or not for the languages in question.

  45. marie-lucie says

    Sorry, I should have double-checked “who” Arakiel was, and I assumed the name was that of an angel since there are several others ending in -(i)el.
    Nostraticists take note: Arabic ard is suspiciously close to English earth and German Erde.
    Yes, but if you add words from other Semitic languages and reconstruct the Proto-Semitic ancestor the phonological differences are much greater (see Ug qrs). That is why comparing words and languages in isolation rather than in groups can be very misleading. That said, it is possible that there is a link (PIE and PS share some roots, whether by relationship or borrowing), but to my knowledge it has not been convincingly demonstrated.

  46. These hypotheses should be checked with accepted Proto-Semitic reconstructions since I am not familiar with them
    marie-lucie: My credentials are limited to strength in Hebrew as a second language and unsated curiosity; I have none as a linguist.
    Your comments about evolution vs. correspondence make sense.
    K vs Q: I should have been more careful. The terminal sound in ara(kq) is more properly designated by Q. Had I done so the relationships would have been clearer.
    Incidentally, compare Latin Q and Hebrew ק: If you close up the larger element in ק you arrive at something very much like a Q. (Qv. archaic Greek koppa and the visuals here.)
    I’ll ask a Hebrew linguist and report further.
    My note to the Nostraticists was firmly tongue-in-cheek. I’m well aware that superficial relationships like this are usually meaningless.

  47. marie-lucie says

    Paul O: aha! What difference a letter makes. Indeed the spelling araq would have been clearer about “ayin”, although the relationship between araq and erets would have been less clear (since [q] and [ts] are more distant from each other than [k] and [ts]). Both [q] and [k] are liable to change, especially depending on the neighbouring vowels. Although all language sounds may change in the course of history, some sounds are more liable to change than others, and [q] is one of them. A correspondence [q : ts] implies an intermediate stage [k] on the way to [ts].
    Thanks for the links to “koppa” (which I did not know about) and “qoph” with its visuals. I had not noticed the resemblance between Latin Q and Hebrew ק but that makes sense because of their common origin in the Eastern Mediterranean. The continuation of the “qoph” article with a discussion of Arabic “qaf” (with long vowel) gives the various modern phonetic end-points of the original [q] in different Arabic dialects – to [k], [g] and even more.

  48. Yes, hence the many spellings of Q/K/Gaddafi.

  49. I had not noticed the resemblance
    There are many resemblances between Hebrew and Latin characters; they’re quite easy to see by examining a table containing the Phoenician, Hebrew, Greek and Latins alphabets and noting which letter corresponds to what sound. Wiki has a pretty good one and I know there are others on the web. Look especially at A, M, O and T, and through a mirror at P and (lower case) R. (The final-position form of the Hebrew letters is older than the initial- and medial-position form.)
    A-dventure-Z The Story of the Alphabet, a colorful book by Ada Yardeni, an expert on Hebrew script, will bring a smile to the face of anyone who enjoys this sort of thing.
    (Hat: You likely know that Hebrew shin ש was scooped up to represent the SH sound in the Cyrillic alphabet.)
    Proto-Semitic reconstructions
    The introductory essay to the AHD’s section on Semitic roots contains many such reconstructions. Some fit with what I’ve observed; others make little sense to me. Lots of larnin’ to do!

  50. marie-lucie says

    Paul O, thank you for the references. I have seen several charts showing Phoenicien, etc correspondences of ancient and modern letters, but not lately, and I have not studied them in detail. I should take another look.
    I read the AHD essay which precedes the Semitic forms (the list is not given in the link quoted). It is quite informative, but my problem was that many letters which would have needed special symbols were simplly omitted with no indication that they were missing, so that some CCC roots would appear simply as CC (thus SHALOM given as from the Hebrew root “lom”, even though triconsonantal s-l-m was quoted correctly elsewhere), or a word given in three stages would be missing a consonant in the middle stage (as would happen with a sequence s-l-m, lom, shalom). Perhaps this is why some reconstructions made little sense! The printed version in the AHD volume probably uses the correct symbols and transcriptions.
    In addition, when letters were cited individually, those for which the necessary symbol was not available were simply replaced by a comma, so that the alphabetical list at the end included a large number of commas interspersed with the regular letters (and the regular commas!).
    I wonder if others have had these problems, and if so, if someone could suggest what to do to get the proper transcription. In past years I have occasionally received emails in French where all the accented letters were omitted, so that for instance été appeared only as t. This occurred at work and probably had to do with the software then used by the institutino.

  51. thus SHALOM given as from the Hebrew root “lom”, even though triconsonantal s-l-m was quoted correctly elsewhere
    Precisely what got me lost in there. The dropped character was probably ṣ (s with an inferior dot).
    I wonder if others have had these problems, and if so, if someone could suggest what to do to get the proper transcription.
    I had terrible problems like these in the bad old days when I used Eudora for email. Hebrew text caused special grief, letters often appearing left-to-right word-by-word, though the words would be in the correct order. Hebrew letters sometimes replaced accented letters in French or German text too. That was the good part; the bad part was when an entire Hebrew text showed up as gibberish. (Don’t get me started on what happened to attachment filenames.) The cure is to make sure your operating system and all relevant software support UTF-8 encoding.
    In the specific case of the AHD essay I linked to above, I suspect it was copied long enough ago that much of the special character encoding got lost along the way. BTW, neither the essay nor the list of Semitic roots appears in my 1981 hard-copy; the roots alone appear in my 2006 electronic version of the fourth edition.

  52. I have the fifth ed. as a physical book, and it shows the root as šlm.

  53. Trond Engen says

    In the distant past when we all had online access to the AHD’s section of IE and Semitic roots, a couple of years ago or so, it used to replace all non-standard characters with slightly-out-of-position images. I seem to remember that in cutting-and-pasting they appeared as spaces and I had to insert the intended character manually.

  54. Took a little searching, but I was able to find a scan of the hard-copy pages from AHD5. All the special characters and diacritics seem to be there.

  55. . . . and posted by no less than the author of the essay, John Huehnergard, and containing the complete index of Semitic roots.

  56. A great find—thanks for sharing it!

  57. marie-lucie says

    Thank you, Paul O, now the reconstructions and words make sense!

  58. On Hebrew tsade and Aramaic ayin: I think this is usually attributed to deapicalization of an affricate. Something similar happens in Catalan, cf. palau ‘palace’.

  59. David Marjanović says

    Perhaps the shift from masculine to feminine for ‘bridge’ (Lat pons, pontis) in some areas was favoured, if not triggered, by the resemblance with the word for ‘well, fountain’ (Lat fons, fontis).

    Dens, fons, mons, pons are the four masculine exceptions to their otherwise feminine declension class in Classical/textbook Latin. It’s not surprising that some of them have joined the majority in various Romance languages.

    The masculine gender of dens, lost in French, goes all the way back: the German cognate Zahn retains it still.

  60. John Cowan says

    Spanish monte

    ObDigression:

    Why is it monte and not muente? Both the RAE and Wikt simply say “from Latin mons, montem, but /ŏ/, like /ĕ/, normally breaks in Spanish to /ue/, /ie/. There is a recognized exception when the next segment is palatal: hodie becomes hoy and not *huey (cf. bove(m) > buey). So far I have three ideas:

    1) Somehow mŏnte(m) irregularly lengthened to mōnte(m) before vowel breaking.

    2) It is a semicultismo, or minimally adapted borrowing, like fondo ‘back, bottom’ (alongside popular hondo ‘deep, profound and borrowing fundo ‘estate, farm’), fin, or libro. But most of these relate to the learned professions, including religion.

    3) It is borrowed from Italian or another source that has no breaking of Latin mid short vowels.

    Anybody know? I asked my Spanish/Catalan consultant, and he pointed to hombre/*huembre, escombre/*escuembre, rostro/*ruestro. The first two suggest that nasals have something to do with it, but not the third. Observers are if not worried at least puzzled.

  61. Never thought of that. Odd.

  62. I tried to look as some more descendants of MONS, MONTIS and there do seem to be a few differences in vowels from the reflexes of PONS, PONTIS. Sicilian munti would normally suggest mōnte(m) > mónte as the immediate ancestor. Not sure about expected reflexes for Catalan munt (contrast pont from PONS), though I notice that the English Wiktionary entry for it claims mōns, mōntem as the origin with an explicit long vowel.
    Do we actually have any direct evidence for the Latin quality in the oblique cases? There couldn’t be any information from metre.

  63. PlasticPaddy says

    montare had long o and resulted in montar, so could the noun have been adjusted by analogy?

  64. David Marjanović says

    Vowel + nasal sequences within syllables were apparently pronounced as long nasal vowels in Latin, preserving the length of the syllable.

    But that doesn’t explain hombre.

  65. For some reason, seeing “hombre” in italics made me think of this.

    During the first year or two of Friends‘s television run (before I had happened to see any of the show), I sat next to Matt LeBlanc’s mother on an airplane. After a few bloody marys, she told me how excited she was about the show’s success, and encouraged me to watch, especially for LeBlanc. “He’s the funny one,” she informed me.

  66. The /b/ in hombre is intrusive, like the /p/ in Thompson ‘Tom’s son’. But none of this explains rostro, which also underwent a semantic shift ‘nose, beak’ > ‘face’.

  67. rōstrum has a long ō.

  68. Surely the p is intrusive in Thomas’ son, and wouldn’t have been allowed into Tom’s son.

  69. Seems that many descendants of mons have a higher vowel than expected; Occitan also has mont without breaking (though I am unsure how general breaking of *ɔ is there to begin with), Catalan has outright munt.

    The Vulgar Latin derivative *montānia has a palatal that could in principle condition some sort of raising, but it comes a syllable too late.

  70. PlasticPaddy says

    @dm
    http://etimologias.dechile.net/?hombre
    homine (m) with stressed o > *homne > *homre > hombre

  71. David Marjanović says

    Of course; I just can’t answer the original question of why this o didn’t become ue.

  72. There is a discussion of some evidence at http://alatius.com/latin/bennetthidden.html
    (section beginning 41. Pontem, fontem, montem, frontem, frondem.)
    One potential indication of an original short vowel is Greek Μοντανός.

  73. PlasticPaddy says

    @dm
    I don’t think you can have Latin om > Spanish uem. What happens is like dominus/a > dueño/a (although here you also have appellations Don/Doña). So your question is : why not *hueño > US Wayne? Maybe some Spanish had *cueño for coño and the Form *hueño died with its shortlived users????. Etienne would know.

  74. Roberto Batisti says

    In Tuscan (and hence Standard Italian), the raising Lat. /ŏNC/ > Vulg. Lat. /ɔNC/ > /ọNC/ is completely regular: mónte pónte fónte frónte cónte nascónde cómpra rispónde sónno sógno cóntro gróngo ógni (the examples are from Rohlfs, vol. 1, pp. 88 and 138 of the 1966 Italian edition). Only /nn/ is excluded (dònna, gònna).
    The structure /ɔNC/ is found only in learned borrowings from Latin, Greek, or other languages (It. pònto < Lat. pontus < Gk. πόντος) – actually, in such borrowings *only* /ɛ/ and /ɔ/ are ever used. This stems from the Italian pronunciation of Latin, which uses /ɛ/ and /ɔ/ respectively for any , regardless of etymological length.

    Castilian Spanish, on the other hand, apparently presupposes raising in some words but not in others of identical structure. Dialect mixture seems to be the most likely answer.

    @ David Marjanović: nasalization plus compensatory lengthening in Latin only affects vowel+nasal sequences before a voiceless fricative (mainly /s/ and /f/, plus /k/ > [x] before an obstruent at some prehistorical stage). Thus, e.g., consul [‘kõːsʊɫ], often spelled COSVL. The nasal was later restored in Classical orthography (and possibly in learned pronunciation – but Quintilian says that the N here was not pronounced). In words of this shape, Romance languages have no trace of the nasal: Lat. mensis > It. mese, French mois, etc. But there was no such process before stops (as in the examples above).

  75. Lars (not the original one) says

    @Roberto Batisti: But Spanish has only five phonemic vowel qualities. That’s one of the problems I have with other Romance languages; I simply don’t have that intuition about the vowel height of e, o because I learned Spanish first.

  76. Roberto Batisti says

    @ Lars: generally speaking, Spanish /o/, while pronounced closer to cardinal [ɔ], goes back to Proto-Western-Romance *ọ (IPA /o/), while Spanish /ue/ goes back to *ǫ (IPA /ɔ/).

    That said, the ‘correct’ distribution of /e/ vs. /ɛ/ and /o/ vs. /ɔ/ is widely disregarded in actual spoken (non-Central) Italian, to the point that I find myself thinking of the Latin etymon more often than not when I want to be sure about which vowel to use.

    (It doesn’t help that the regional Italian of Bologna and Emilia-Romagna more generally, where I grew up, has an especially confusing distribution of these vowel qualities, since it differs from the standard in mostly non-predictable ways. Compare the regional Italian of Western Lombardy, which more or less follows the rule “open vowel in closed syllable, close vowel in open syllable”, except vowels are also close before a tautosyllabic nasal. Or Sicily and other places where the local dialect has a 5-vowel system, meaning that no distinction is made between open-mid and close-mid vowels when speaking Italian).

  77. John Cowan says

    It’s precisely because of vowel breaking that Spanish has only five vowel qualities: any Latin /ĕ/ or /ŏ/ that didn’t break merges with the reflexes of /ē/ and /ō/ respectively.

  78. David Marjanović says

    nasalization plus compensatory lengthening in Latin only affects vowel+nasal sequences before a voiceless fricative

    Correct, and that is where Greek transcriptions often lack the n, says the source anhweol just linked to; before /t d/, the vowel is short (except in various contractions) and spelled as such, plus n, in Greek transcriptions. Before /t d/ is probably where nasal vowels most easily sprout a nasal consonant, as perhaps seen when comparing French to Polish today.

  79. David Marjanović says

    an especially confusing distribution of these vowel qualities

    Reminds me of German, except there it’s the Western-Lombardy-like variant that happens to be standard!

Speak Your Mind

*