Negus, Śawt, Ḍäppa.

I got curious about the word Negus, “Used formerly as a title for emperors of Ethiopia” (AHD) or, as the more expansive OED entry (updated September 2003) puts it, “(The title of) a king of Ethiopia or of a province or kingdom within Ethiopia; spec. (the title of) the supreme ruler of Ethiopia; the Ethiopian emperor.” The OED has:

Etymology: < Amharic nəgus king < näggäs- to become a king (as the title of the Ethiopian emperor also as nəgusä nägäst, lit. ‘king of kings’). Compare French négus (1556 in Middle French as negus; rare before 18th cent.).

Well, if the Amharic word is nəgus, how come the English pronunciation is /ˈniɡəs/? Happily, the AHD also gives /nɪˈɡuːs/ (in their own idiosyncratic rendering), which suits my sense of things much better, so I am adopting it.

But the AHD etymology says Amharic nəgus is “from Ge’ez nəguś, king, ruler, verbal adjective of nagśa, to rule, become king; see ngś in the Appendix of Semitic roots.” What did this ś represent? Wikipedia, s.v. Śawt (the name of the corresponding Ge’ez letter), says the Proto-Semitic sound was a “voiceless lateral fricative *ś [ɬ], like the Welsh pronunciation of the ll in llwyd,” and Rick Aschmann’s Reflexes of Proto-Semitic sounds in daughter languages says:

2. Proto-Semitic */ś/ was still pronounced as [ɬ] in Biblical Hebrew, but no letter was available in the Phoenician alphabet, so the letter ש did double duty, representing both [ʃ] and [ɬ]. Later on, however, [ɬ] merged with [s], but the old spelling was largely retained, and the two pronunciations of ש were distinguished graphically in Tiberian Hebrew as שׁ [ʃ] vs. שׂ [s] < [ɬ].

But how do they know? Via loan words in other languages?

Furthermore, the Śawt article says “See also Ḍäppa ṣ́ ፀ,” but that Ḍäppa link redirects to Ḍād, and a Google search on Ḍäppa essentially turns up only the Śawt article. What’s going on? Is Ḍäppa a thing? All information gratefully received.

Comments

  1. David Eddyshaw says

    There’s no problem with the mere fact that Hebrew originally preserved a distinct reflex of Proto-Semitic /ś/, of course, as that is clear from the history of the sibilants within Hebrew itself; I suppose it’s simply most parsimonious to suppose that the way it was distinct from /s/ and /ʃ/ was that it still preserved the Proto-Semitic sound. That, in turn, is still there in all its pristine glory in Modern South Arabian.

  2. David Eddyshaw says

    Lambdin’s Ethiopic grammar transliterates ፀ as ḍ, but says that ḍ was confused so early with ṣ that the change has contaminated nearly all manuscripts, to the degree where the choice of one or the other in roots is sometimes arbitrary.

  3. David Eddyshaw says

    Weird.

    My comments are showing up in a sort of Heisenberg fashion. The sidebar lists them, and then when I click on them, they disappear as if they had never been …

    Time to sacrifice a black cock to Akismet again …

  4. J.W. Brewer says

    You say “the” pronunciation, but your AHD link gives two different ones, as does Merriam-Webster. https://www.merriam-webster.com/dictionary/negus It’s obviously possible that one is “authentic” in some historical sense and the other is a spelling pronunciation that became current because more people read the word than chanced to hear it uttered aloud by someone comfortable with it. There’s also the possibility of interference from the unrelated noun “negus,” meaning (per one source) “a hot drink of port and lemon juice, usually spiced and sweetened”* and named for its supposed inventor https://en.wikipedia.org/wiki/Francis_Negus Different sources offer different theories as to the pronunciation of the beverage, and of course surnames can have looser orthographic/phonemic connections than many other sorts of word do.

    *Can’t say I ever encountered this word before – it was apparently drunk by characters in novels by various prominent 19th-century writers but may not have survived into the 20th century, at least under that name.

  5. J.W. Brewer says

    Turns out here’s a now-retired Australian tv journalist named George Negus whose surname is apparently pronounced (in AustEng) with the FLEECE vowel. Back in 1979 he interviewed Bob Marley, who would have been familiar with the (Jamaican pronunciation of) the Ethiopian title. Negus’ Aussie pronunciation of phrases like “marijuana haze” is something to behold.

  6. You say “the” pronunciation

    I meant the OED’s.

    There’s also the possibility of interference from the unrelated noun “negus,” meaning (per one source) “a hot drink of port and lemon juice, usually spiced and sweetened”

    Yes, I’ve encountered it (the word, not the drink), which is another reason I want a distinctive pronunciation for the Ethiopian ruler. Distinguo!

    Turns out here’s a now-retired Australian tv journalist named George Negus

    Memorably lambasted here in 2006 as “one of the most vacuous TV presenters ever to clutter the (already mediocre) Australian airwaves.”
    https://languagehat.com/grr/#comment-23710

  7. Why is Akismet not taking my comments? I run this joint!

  8. test comment

  9. January First-of-May says

    Why is Akismet not taking my comments? I run this joint!

    …yeah that’s actually starting to look weird.

    (Comments 4459430 and 4459432 appear to be substrings of comment 4459431, so you’d probably want to keep that one.)

  10. J.W. Brewer says

    One could survey reggae songs where the Ethopian-derived “negus” turns up in the lyrics to see if they all have the same pronunciation or exhibit variation. In this one (Toots & the Maytals) the word first turns up around 3 mins 11 secs in: https://www.youtube.com/watch?v=UrFbwScTxN0.

  11. January First-of-May says

    Comment test. Quadraticity drinks imagination. Akismet is a ch*rping m*stard.

  12. January First-of-May says

    Проверка связи. Русский язык самый русский язык в мире. Акисмет зелёный ягуар.

  13. David Marjanović says

    But how do they know? Via loan words in other languages?

    Yes, e.g. balsam < bśm. I don’t know how much that says about the timing, though. I also doubt that [ɬ] > [s] can be done in one step; I suspect [ɬ] > [ç] > [ɕ] (what ś is meant to suggest) and then > [s] in Hebrew, [ʃ] in Arabic.

    ḍ was confused so early with ṣ

    i.e. [tɬʼ] with [tsʼ]

  14. David Marjanović says

    I think it’s not taking anyone’s comments in this thread.

  15. January First-of-May says

    Now it’s worse, the main page of the site shows the missing comments (in the sidebar), but the article page ignores them. A case of really bad caching, maybe, but even then it’s weird as ch*rp.

  16. Yes, e.g. balsam < bśm.

    Thanks!

  17. January First-of-May says

    i.e. [tɬʼ] with [tsʼ]

    …no offence but a [tɬʼ] phone(me) feels very Klingon, or maybe Salishan.

    I think it’s not taking anyone’s comments in this thread.

    This is my impression, but at least I’ve gotten around the apparent caching issue by opening the website in another browser. (It’s probably about to start caching there too.)

    …in retrospect it could be because the URL is languagehat.com/negus-sawt-%E1%B8%8Dappa (where the %E1%B8%8D bit represents ḍ) and it expects its normalizing function to only produce ASCII but apparently Ḍ was weird enough that it didn’t get normalized.

  18. another test comment

  19. David Marjanović says

    …no offence but a [tɬʼ] phone(me) feels very Klingon, or maybe Salishan.

    It’s also scattered over the Caucasus, and at least some Modern South Arabian languages have a Tlingit-style [ɬʼ].

    The Klingon [tɬ] is not ejective and not even aspirated; but Marc Okrand almost turns it into a trill.

  20. What’s going on? Is Ḍäppa a thing?

    The entry for ሠ śawt says “see also ḍäppa” because the original Ge’ez letter ፀ ḍäppa was used to represent the sound descended from PS *ṣ́ (< Arabic , Hebrew צ , in Aramaic eventually falling together with ע /ʕ/, etc.), and this is the emphatic (that is, ejective) counterpart to *ś (> Ge’ez ሠ ś, Arabic š, Hebrew שׂ ś, Aramaic /s/ written ס orשׂ, Syriac ܣ s). And Semitists often discuss the two together (as Alex Bellem, page 153f here, references numbers removed, with information relevant to the rest of the thread):

    The arguments for a lateral series… in Semitic are now fairly well accepted, post-Cantineau’s (1951) study of the evidence that had accumulated over the previous decades. Firstly, it is well known that Modern South Arabian has, as realisations of Semitic ‘ḍ’ and ‘ś’, the laterals ɮ and ɬ, respectively, with no evidence that these were the result of innovation or borrowing. Moreover, Arabic was clearly produced as a voiced lateral fricative at least until the eighth century CE, as per the description given by the medieval grammarian Sībawayh. Additionally, in the Ḥaḍramawt dialects of southern Yemen, Arabic is produced as the emphatic lateral .

    On top of this, there is much evidence that in loanwords from Arabic in the early years of the Arabo-Islamic conquests was interpreted in other languages as ‘ld’. There is also the evidence of the Akkadian merger of š (< etymological ‘ś’, i.e. ɬ) and l before apical stops and spirants, which makes sense if Akkadian š was (at least historically) a lateral ɬ. Additional evidence, parallel with ‘ld’ for Arabic in loans, is drawn from examples such as Greek balsamon ‘Commiphora opobalsamum (tree)’ borrowed from Semitic *b-ś-m, also from reported variation in Arabic of certain forms, e.g. qišda ~ qilda ‘clarified butter sediment’, and in Hebrew between certain roots, e.g. ṣḥq ~ śḥq (‘laugh’), and also inter-Semitic variation in certain roots, as in (again) that relating to ‘laugh’, which is Arabic ḍḥk and Hebrew ṣḥq ~ śḥq.

    August Dillmann has explained the Ge’ez name däppa as being related to Arabic ضبّة ḍabba ‘door bolt’ (see here. The original form of the letter in the Old South Arabian script was 𐩳 (if this character doesn’t display for you, it looks rather like the Chinese character 日).

  21. What’s going on? Is Ḍäppa a thing?

    The entry for ሠ śawt says “see also ḍäppa” because the original Ge’ez letter ፀ ḍäppa was used to represent the sound descended from PS *ṣ́ (< Arabic , Hebrew צ , in Aramaic eventually falling together with ע /ʕ/, etc.), and this is the emphatic (that is, ejective) counterpart to *ś (> Ge’ez ሠ ś, Arabic š, Hebrew שׂ ś, Aramaic /s/ written ס or שׂ, Syriac ܣ s). And Semitists often discuss the two together (as Alex Bellem, page 153f here, references numbers removed):

    The arguments for a lateral series… in Semitic are now fairly well accepted, post-Cantineau’s (1951) study of the evidence that had accumulated over the previous decades. Firstly, it is well known that Modern South Arabian has, as realisations of Semitic ‘ḍ’ and ‘ś’, the laterals ɮ and ɬ, respectively, with no evidence that these were the result of innovation or borrowing. Moreover, Arabic was clearly produced as a voiced lateral fricative at least until the eighth century CE, as per the description given by the medieval grammarian Sībawayh. Additionally, in the Ḥaḍramawt dialects of southern Yemen, Arabic f is produced as the emphatic lateral .

    On top of this, there is much evidence that in loanwords from Arabic in the early years of the Arabo-Islamic conquests was interpreted in other languages as ‘ld’. There is also the evidence of the Akkadian merger of š (< etymological ‘ś’, i.e. ɬ) and l before apical stops and spirants, which makes sense if Akkadian š was (at least historically) a lateral ɬ. Additional evidence, parallel with ‘ld’ for Arabic in loans, is drawn from examples such as Greek balsamon ‘Commiphora opobalsamum (tree)’ borrowed from Semitic *b-ś-m, also from reported variation in Arabic of certain forms, e.g. qišda ~ qilda ‘clarified butter sediment’, and in Hebrew between certain roots, e.g. ṣḥq ~ śḥq (‘laugh’), and also inter-Semitic variation in certain roots, as in (again) that relating to ‘laugh’, which is Arabic ḍḥk and Hebrew ṣḥq ~ śḥq.

    August Dillmann has explained the Ge’ez name däppa as being related to Arabic ضبّة ḍabba ‘door bolt’ (see here. The original form of the letter in the Old South Arabian script was 𐩳 (if this character doesn’t display, it looks rather like the Chinese character 日).

  22. balsam < bśm: the vocalism is obscure to me. In the (Tiberian) OT there’s beśem and bośem but no bāśām. The plural bǝśāmīm is much commoner but it would be the plural of any of the three forms. The Aramaic is buśam as far as I can tell.

  23. …in retrospect it could be because the URL is languagehat.com/negus-sawt-%E1%B8%8Dappa (where the %E1%B8%8D bit represents ḍ) and it expects its normalizing function to only produce ASCII but apparently Ḍ was weird enough that it didn’t get normalized.

    I have removed the last, weirdest word from the URL; let’s see if that helps.

  24. tv journalist named George Negus

    For me, the most famus Negus is Arthur, beloved of BBC TV “Going for a Song”/”Antiques Roadshow”.

    The wiki says the Negus’s trace their ancestors back to 1700’s. Fortunately no diacritics. The name seems to be thoroughly Anglo.

  25. David Eddyshaw says

    Ah me! Yes. Arthur Negus …

  26. “[…] there’s no […]bāśām.”

    Could not the form underlying בְּשָׂמִי ‘my ….’ in Song of Songs 5:1 be *bāśām?

  27. Could not the form underlying בְּשָׂמִי ‘my ….’ in Song of Songs 5:1 be *bāśām?

    It would indeed, and good catch. But then why would the attestation of the singular in the SoS be bośem?

  28. Also Chaldean < כַּשְׂדִּים (Kasdim in modern Hebrew).

  29. John Cowan says

    I have always supposed that the Grand Nagus (pronounced with FACE) of the Ferengi Alliance was named Doylistically after the emperors of Ethiopia.

  30. In Modern Hebrew, נוֹגֵשׂ ~ נֹגֵשׂ noges is the guy in cheesy cartoons of Ancient Egypt, standing on top of a huge stone on rollers and cracking a whip at the slaves pulling it.

  31. We have a king serving in Congress. If I say that Eritrea and Ethiopia are close enough would it be very wrong or only a little wrong?

  32. David Eddyshaw says

    Well, Tigre and Tigrinya are actually more closely related to Geʽez than Amharic is, so close enough.

  33. if the Amharic word is nəgus, how come the English pronunciation is /ˈniɡəs/?

    This went unremarked, but has a very simple answer: in Ethiopicist transcription, ə is /ɨ/ (and ä is /ɐ~ə/, for tripping up anyone expecting /æ/ or for that matter Tocharologists expecting /ɨ/). Also Ge’ez is BTW similar retranscription from Gəʽəz.

    Tigre and Tigrinya are actually more closely related to Geʽez than…

    Last I heard, any genealogical notion of North Ethiosemitic was now considered dubious / mostly characterized by archaisms. Bulakh & Kogan 2014 even consider a few family trees like {Tigre, {Ge’ez, {Tigrinya, South ES}}}.

  34. David Marjanović says

    I have removed the last, weirdest word from the URL; let’s see if that helps.

    I just came here following a link on the “recently commented” site, and the URL I see is https://languagehat.com/negus-sawt-%e1%b8%8dappa/#comment-4459592 (except I actually see the ḍ displayed).

  35. Here is the entry in the online SED for *ngś “to push, to press, to drive to work”. They say:

    Probably related to Akk. nagāšu ‘to leave, to go away’, in which case the transitive meaning “to drive” may be regarded as a common innovation of CS/EthS. Conversely, there is hardly any connection between this root and [Mehri] nəgūŝ ‘to shake milk for butter’, Jib[bali] ngɔŝ id.

    Some idle speculation with no particular conclusion… In regard to the (non)-relationship between ‘to press, oppress, drive’ and ‘churn, shake milk for butter’, however, I am reminded in some ways of the curious case of Indo-European *neih₁/₃- ‘turn (something) in a certain direction, lead’ and *neiH- ‘churn butter’. PIE *neih₁/₃- ‘to turn in a certain direction, lead’ can be seen in Indo-Iranian and Anatolian: Sanskrit náyati ‘he leads, directs, governs’, Avestan naiieiti ‘leads’; Hittite nāi ‘he turns (something/someone) in a certain direction, sends, wraps around or ties (as a cord)’, reduplicated intensive nanna/i- ‘drive (an animal, a vehicle)’; Luwian nana- ‘to lead’. (Perhaps also Old Irish nia ‘warrior’?)

    Beside this, there exists another *neiH- ‘to churn butter’, with reflexes in Latvian and Indo-Iranian: Latvian sviestu nīt ‘to make butter’, paniņas ‘buttermilk’; Sanskrit nīta- ‘fresh butter’, netra- ‘cord for moving the churning stick’; Khotanese nīyaka-, Kurmanji nivîšk, Balochi nēmaġ ‘butter’; finite verbals forms in Roshani nay-/nid, Sariqoli nεy-/nůd, nïd ‘to churn’; Shugni nīm-δōrg ‘churnstaff’; Persian panīr ‘cheese (originally as made by reprocessing buttermilk for curds?)’. It has often been suspected that these two Indo-European roots are somehow one. Could the same be true in Semitic?

    But if the two are ultimately one, in either Indo-European or Semitic or in both, does the semantic connection have to do with the actual motion, the turning of the churn-staff in the case of Indo-European, and the pushing of a goatskin churn in the case of Semitic? (From what I could find out about traditional Omani butter making, a suspended skin bag was shaken (illustration here from Kurdistan), agreeing with T.M. Johnstone’s definitions for the Modern South Arabian words.) Or could the development in Modern South Arabian have been from *‘demand, collect together’ (as seen in Hebrew and Arabic and Old South Arabian) further to “shake milk to collect the butter”?

  36. January First-of-May says

    Judging by my most recent attempt to view a new comment, there might be comment-related issues again.

    (I guess I’ll know for sure if this doesn’t send.)

  37. January First-of-May says

    Now my comment is invisible but the one I was looking for is visible. Weird.

  38. Here is the entry in the online SED for *ngś “to push, to press, to drive to work”.

    Thanks very much for that and your informed commentary/speculation!

  39. Thanks very much for that and your informed commentary/speculation!

    Thank you for entertaining my speculations… I hope that a LH commenter can provide a piece of data from the vocabulary of milk-processing in Africa or Siberia or the Balkans that will clinch it.

  40. David Eddyshaw says

    Fulfulde for “butter” is n(y)ebbam

    Sadly, the bit of Africa I know about is not dairy country, apart from the cattle-nomad Fulɓe …
    Actually even this Fulfulde word seems less specific than “butter”; it’s “oil, grease” as well; moreover, with a different class suffix, nyebbere means “bean.”

    The pan-Oti-Volta etymon represented by Kusaal kpaam means “oil, grease, (shea) butter”, though you can use it for butter-as-in-dairy. You can specify that explicitly by making a compound: bin’iskpaam, where the first element is bin’sim “milk”, which is itself derived from bin’isir “breast, udder” (both pan-Oti-Volta, again, but primary referring to human breasts and human milk.) For “making butter” it seems you just say exactly that in Kusaal, with maal “make, create.”

    None of my Oti-Volta dictionaries seems to have a word for “churn.” Most languages (even Nawdm, way over in Togo) have borrowed the word for “butter” from English.

  41. I think caution is advised with the Mehri ‘churn’ word. That is because there is such an extensive butter vocabulary, judging from Johnstone’s vocabulary. There are the following (I’m listing the root only, plus the gloss for the first item derived from that root):
    ḥmź ‘to make butter, shake milk for butter (which takes 15-30 minutes).’
    ḳṭmm ‘unclarified, freshly made butter.’
    ngś ‘to mend; to shake milk for butter.’
    xzr ‘fresh (unclarified) butter.’
    xźź ‘to wade; to shake (milk for butter).’
    Plus terms for the preparation of dried buttermilk, etc. To my taste, it’s better to look at the semantic map as a whole first before etymologizing the individual terms.

    That said, I wonder if the first term is related to the Hebrew ḥbṣ ‘to churn’.

  42. The SED thinks it’s rather related to Hebrew ḥmṣ “be leavened; be sour”.

  43. I was wondering about that, then decided the semantics were too disparate. But if the phonology matches, so be it. Maybe Hebrew ḥbṣ is kin to ḥbṭ ‘beat’.

  44. Owlmirror says

    Y:

    ḥmź ‘to make butter, shake milk for butter (which takes 15-30 minutes).’

    TR:

    The SED thinks it’s rather related to Hebrew ḥmṣ “be leavened; be sour”.

    You’ve reminded me about something I remember about paneer:

    Paneer is prepared by adding food acid, such as lemon juice, vinegar, citric acid or dahi (yogurt), to hot milk to separate the curds from the whey.

    I have to wonder if the Mehri made such acid-started cheese, either as an independent invention, or transmitted from somewhere else.

  45. Owlmirror says

    Akismet mislikes me…

  46. David Eddyshaw says

    Nah, it’s just got a thing about dappae. Or possibly neguseses. Probably sees them as rivals.

  47. Owlmirror says

    ++?????++ Out of Cheese Error. Redo From Start.

  48. A bit of searching leads to some details on Arabian dairy products. It is my great pleasure to bring the word oggtt to the attention of those present here.

  49. Oggtt is one of the best words I’ve ever seen, but is it real?

  50. Maybe instead of cheese, I should have thought of buttermilk (sour due to lactic acid).

    Originally, buttermilk referred to the liquid left over from churning butter from cultured or fermented cream. Traditionally, before the advent of homogenization, the milk was left to sit for a period of time to allow the cream and milk to separate. During this time, naturally occurring lactic acid-producing bacteria in the milk fermented it. This facilitates the butter churning process, since fat from cream with a lower pH coalesces more readily than that of fresh cream. The acidic environment also helps prevent potentially harmful microorganisms from growing, increasing shelf life.

    Traditional buttermilk is still common in many Arabic, Indian, Nepalese, Pakistani, Finnish, Polish, and Dutch households, but rarely found in other Western countries. […] In the Arab world, buttermilk is a common beverage to be sold ice cold with other dairy products. It is popular during Ramadan, where it is consumed during iftar and suhur.

  51. Indeed, أُقْطٌ (I think). Here are related forms, with an etymology. The <gg> seems to stand for q, the <tt> for .

  52. It’s probably phonologically and grammatically naive of me to wonder if ʾaqiṭ is related to ḳṭmm above.

  53. @Y, thank you, it is beautiful. “Studies on Oggtt” (from the bibliography to the article).

  54. @Y, Steiner 1977 The Case for Fricative Laterals in Proto-Semitic, 123-129 (XVI. Evidence from Balsamon – Bśm for Lateral Ṥ) comments on it this way:

    The Greek and Roman sources, then, lead us to Palestine. But when we turn to Hebrew sources, we find that both the meaning of bśm (it denotes any perfume or pleasant-smelling substance) and its usual vocalization (bośɛm) do not match those of the Greek word balsamon. This problem is easily solved, however, by identifying the latter with the hapax båśåm, which, in context (Cant. 5:1 ʔåriṯi mori ʕim běśåmi “I plucked my myrrh with my båśåm“), seems to refer to a specific plant (Encyclopedia Miqrait s.v.).

    Presumably Encyclopedia Miqrait has more…

  55. (i forgot the link. And a small addendum is here)

  56. I like the first addendum: “I now prefer the term lateral fricative to fricative-lateral.”

  57. So there are attestations of three different vocalizations, one of which (indirectly) fits the one sought here. That’s not bad. It’s curious, though, that so many forms exist. I wonder if bośem and beśem are not original, but backformed from the plural bǝśāmīm, which is attested far more than the singular.

  58. I’m trying to read something about the history of alpha-beta-gamma (the names). So I am reading this (sci-hub).

    “and even a bēt, whose pictographic character is not usually questioned, [29] looks like a ‘house’ only if we posit a surprisingly abstract drawing of a house front or a ground plan. [30]”

    Well, it looks like 9 (Or more exactly) but does anyone know how children in generalized Timbuktu (that is: where roofs are not like ^) draw a house?

  59. …but that Ḍäppa link redirects to Ḍād,…

    Which in turn contains this line: “the corresponding letter in the South Arabian alphabet is [himjar za2.PNG, text when mouse over: ḍ] ṣ́, and in Ge’ez alphabet Ṣ́appa ፀ

    Sigh. Meanwhile the article quoted above has this note:

    The wave-like shape seems to favour naḥaš ‘snake’ (Driver [n. 1], 154 and 165; cf. however Ruijgh [n. 20], 541), but this name is attested, like the other Ethiopian letter names, only in 1548 in a translation of the New Testament printed in Rome; Ullendorff (n. 31), 211–14, suspects that the entire Ethiopian name series was invented in the sixteenth century by European missionaries or scholars (cf. Sass [1991] [n. 1], 92, but differently Nöldeke [n. 20], 131–3, and Lundin [n. 29], 21). One may also ask why a name naḥaš ‘snake’, which fits the shape rather nicely, should have been replaced in the Northwest Semitic tradition by a less plausible nūn ‘fish’. On further name changes see Demsky (1997) (n. 20), 364 (hehin and pepin in Rabbinic literature); Ruijgh (n. 20), 542 (dalt renamed from dag ‘fish’; cf. Cross and Lambdin [n. 24], 25, Sass [1988] [n. 1], 113–14), and § VIII on *ṯannšin; the excessive reliance on such changes in Lidzbarski (n. 30), 126–38, is already criticized by Gardiner (n. 1), 7–8.

  60. Stu Clayton says

    I suspect that powerful linguist lobbies, wanting to ensure job opportunities for themselves, heavily influenced the decision to confuse the language of all the earth. Unfortunately, market prices collapsed about 50 years ago.

    The second wave of language confusion was created by computer science lobbies, to ensure that programmers will always be in demand.

  61. @LH, it seems this Ḍäppa-Ṣ́appa (take it as a reduplication…) comes from (archive):

               T E S T A M E N T U M

            ɴᴏᴠᴠᴍ ᴄᴠᴍ ᴇᴘɪꜱᴛᴏʟᴀ ᴘᴀᴠʟɪ ᴀᴅ

    Hebreos tantum, cum concordantijs Euangelistarum Eu-
     sebij & numeratione omnium verborum eorundem.
      Miſſale cum benedictione incenſi ceræ et c. Alpha=
       betum in lingua ግእዝ፡gheez, id eſt libera quia a
        nulla alia originem duxit, & vulgo dicitur
         Chaldea, Quæ omnia Fr̃ Petrus Ethyops
          auxilio piorum ſedente Paulo.III.Pont.
           Max.& Claudio illius regni Im=
             peratore imprimi curauit.

           ᴀɴɴᴏ ꜱᴀʟᴠᴛɪꜱ ᴍ. ᴅ. xʟɪɪɪᴠ

    (I typed it as printed partly for fun, partly because ᴍ. ᴅ. xʟɪɪɪᴠ is sooooooo much better than
    ᴍᴅxʟɪɪɪᴠ. And I wonder what this mutation ግእዝ / ግዕዝ (Gəʼəz / Gəʽəz) is about:-/ At least it is comforting that not only apostrophes are confused….)

    It has a table (or more conveniently as jpeg), where the letter is called

    ፀጳ

    and values are defined as

    .........ጳ pa
    ፀ zɑ ........

    In other words, it is what

    Ullendorff (n. 31), 211–14, suspects that the entire Ethiopian name series was invented in the sixteenth century by European missionaries or scholars (cf. Sass [1991] [n. 1], 92, but differently Nöldeke [n. 20], 131–3, and Lundin [n. 29], 21).

    was written about. A later Historia Æthiopica calls it Zappa – again, i don’t know what this gemination is about.

  62. = was meant to be ⸗

  63. PS: I understood the question “is Ḍäppa a thing?” as referring to the name Ḍäppa, not the character ፀ.

    Zappa in

    IOBI LVDOLFI
    aliàs 𝕷𝖊𝖚𝖙⸗𝖍𝖔𝖑𝖋 dïcti
    HISTORIA ÆTHIOPICA, sive brevis & ſuccincta deſcriptio REGNI HABESSINORUM,
    Quod vulgò malè Pʀᴇꜱʙʏᴛᴇʀɪ Iᴏʜᴀɴɴɪꜱ vocatur.

    Letter names:
    Hoi Lawi Haut Mai Saut Rεεs Sat Kaf Bet Tawi Harm Nahas Alph Qaf Wawe Ain Zai Jaman Dent Geml Tait Pait Zadai Zappa Af Pſa

    Another table with comparison with Samaritan letters. Without Zappa.
    Alf Bet Gεml Dεnt Haut Wavv Zaj Hharm Tait Jaman Caf Lavvi Maj Nahas Saat Ain Af Tzadai Kof Rεεs Saut Tavvi

    (English translation)

  64. PS: I understood the question “is Ḍäppa a thing?” as referring to the name Ḍäppa, not the character ፀ.

    Exactly right, and thanks for the information.

  65. Well, apparently it is different people insert their transliteration of choice in _appa, where _appa follows a tradition that originated in 16th century.

    But then there is another question of where Zappa comes from.

    Ullendorff:

    ___________________________

    Names of the letters of the alphabet
    ……………………………….

    Furthermore, as we have seen above, there is no Abyssinian source, recent or ancient, which discloses any knowledge in this respect. On the contrary, there is a tradition-prevalent in all the three major modern Abyssinian languages—which can be traced back at least a hundred years (and which may, conceivably, be very old) in which an altogether different system of nomenclature is adopted.

    Praetorius, who in his monumental Amharische Sprache (I879) had already pointed out that the old names of the letters were nowadays completely ‘in Vergessenheit geraten ‘, drew attention to the fact (op. cit., p. 8) that all letters are called by their normal sound in the first order,¹ e.g. Iä, rä, fä, &c. The only exceptions to this rule are those letters which are no longer distinguished in pronunciation. They alone possess names known to every Abyssinian, whether Tigrinya-, Tigre- or Amharicspeaking, as follows:

    (a) h- halʸeta; ḥ- ḥamär; ḫ- bəzuḫan; (b) š- nəguš; s- ʾəsat; (c) ʾ- ʾalʸef; ʿ- ʿayn; (d) sˡ- sˡälot
    (also sˡəlmat and sˡədəkˡ); (e) dˡ- dˡäḥay.

    ˡ
    These Geʽez names are not only used when Geʽez spelling is discussed, but they are universally applied to all modern Abyssinian languages that are written in the Ethiopic alphabet, even if these languages themselves have abandoned those distinctions.

    If my view that the ‘old’ letter-names in Ethiopic are of comparatively recent and non-indigenous origin is correct, then it will be readily understood that some sort of nomenclature must have been introduced at a fairly early date (certainly before the end of the first millennium A.D.), in order to distinguish such letters as had lost their characteristic pronunciation. This assumption concerning the age of the above nine names appears to find confirmation in their acceptance over the entire Abyssinian language field. Indeed, the fact that these names are common Abyssinian property might conceivably point to a period before the modern languages had succeeded Geʽez, although it has to be admitted that the universality of this nomenclature could also be explained through the traditional methods of Geʽez instruction.

    The principle underlying these names is not necessarily that of acrophony, but rather the choice of a representative and frequently occurring word in which the desired sound has a prominent position: ḫ- bəzuḫan; s- ʾəsat opposite ḥ- ḥamär or dˡ- dˡäḥay.

    Ḥamär is exclusively a Geʽez word; Tigrinya, Tigre and Amharic use märkäbʾEsat occurs in all but Tigrinya where ḥawwi is used. In Amharic bəzu the operative consonant is missing. ʾAlʸef and ʿayn are no doubt derived from the Psalter (the alphabetically ordered psalms), and sˡəlmat and sˡədəkˡ take in some schools the place of the usual sˡälot. The tradition regarding these names must thus have been old and powerful enough to counterbalance the facts that (a) most of these distinctions have become meaningless in the pronunciation of the modern languages, and (b) some of the names are not even part of the ordinary vocabulary of those languages.

    Abba Yaʽqob Gebreyesus in his grammar of Geʽez (in Amharic) and his grammar of Tigrinya (written in Tigrinya) does not give names to any of the symbols of the

    ¹ Cf. also Mittwoch (Trad. Ausspr., p. 8): ‘Der Abessinier kennt den Begriff des Konsonanten für sich überhaupt nicht, sondern kann ihn sich immer nur in Verbindung mit einem Vokal vorstellen. Das ist auch bei sadəs der Fall, das neben der Vokallosigkeit auch den unbestimmten Vokal ə bezeichnet.’
    ² In other words: the names of the letters have universal currency, although the objects incidentally represented by these words (e.g.’ ship ‘,’ fire’, &c.) are, in some cases, called differently in the various languages.

    ____________________________

    Ethiopic syllabary except to those nine sounds mentioned above: ‘There are three h sounds:

    h- halʸeta; ḥ- ḥamär; ḫ- bəzuḫan are their names.’¹ or: ‘š-This s sound is called nəguš;’ ² &c., &c. but: ‘r- is the 6th letter²’ &c., &c.

    The names of the Ethiopic letters thus provide one instance where indigenous tradition and external evidence allow us to arrive at the same conclusions.

    Innovations in the script—as compared with Geʽez
    …………………………..

    ¹ Mäsˡḥaf Säwasəw zägəʼ’əz (sic!), Asmara 1920 (i.e. 1927), p. 7.
    ² Nay təgrəňňa säwasəw, Asmara 1926 (i.e. 1933), p. 13.

    __________________________________________________

    And https://en.wikipedia.org/wiki/Geʽez :

    … Geʽez ś ሠ Sawt (in Amharic, also called śe-nigūś, i.e. the se letter used for spelling the word nigūś “king”) is reconstructed as descended from a Proto-Semitic voiceless lateral fricative [ɬ]. Like Arabic, Geʽez merged Proto-Semitic š and s in ሰ (also called se-isat: the se letter used for spelling the word isāt “fire”).

  66. But then there is another question of where Zappa comes from.

    I know the answer to that: Baltimore.

  67. Yes, they (17th century sources) should have retained the spelling Zɑppa (from the preface to the NT) to avoid confusion.

  68. ḍ vs. ṣ́ : I guess

    ḍ means “a Semitic correspondence to Arabic ḍad”
    ṣ́ means “an emphatic counterpart to *ś in PS” (see Steiner above: intro, conclusions (p 155) and the PS word for laugh 111).

    ś itself likely meant some sort of palatal fricative which is not the current (lateral..) party line. It coudl be just s² but I can see why they don’t like ṣ².

    Not sure if ṣ́ is any better than ḍ. It points on the (actual) connetion with another sound in the same language. ḍ points at the (actual) connection with another sound in related language. But “this diacritic means that once upon a time we thought that a related sound is palatal (based on Arabic)” is a poor motivation to adding a diacritic… Could just write… ɬʼ
    —-
    ….Or Cyrillic Ц with an extra dent for emphasis. …

  69. I know the answer to that: Baltimore.
    I was thinking about answering along similar lines, but then I thought, no, Languagehat is a serious blog, that kind of silliness is not appropriate 😉

  70. Quite right, we don’t tolerate that kind of thing around here!

  71. Well, I just calculated that if you appear naked in public, the effect will be even more striking if you behave as if you were dressed.

    Also I thought “*appa-цараппа”.

  72. zägəʼ’əz (sic!)

    Oh

    It was zägəʼəz (sic!)🙁

  73. WP “Many of the letter names are cognate with those of Phoenician, and may thus be assumed for the Proto-Sinaitic script.”

    Ha-ha. “Phoenician” names are attested in the form “alpha, beta, gamma….”.

  74. Michael Jursa Die Kralle des Meeres und andere Aromata / The Claw of the Sea and other Aromatics p 156 (academia, GB):

    c. baltām(mu) (BM 77429: 22)
    Dieses Wort, in unserem Text ˢ̌ᶦᵐ⸢ba-al-ta⸣-am geschrieben, findet sich auch als ˢ̌ᶦᵐba-al-tam-[(mu)] in BM 73126⁴² und als ˢ̌ᶦᵐba-al-tam-mu in BM 75774. Beide Tafeln sind Aromata-Listen aus dem Ebabbar-Archiv und stehen im Zusammenhangmit der Herstellung von Salbölen. Dies ist ein westsemitisches Lehnwort; die Kombination steht für /ś/. Das Wort ist also als *baśām „‘Balsam’“ anzusetzen –es, oder zumindest die Wurzel, ist im Hebräischen, Aramäischen und Arabischen (bašām ) gut bezeugt. Als bálsamon bzw. balsamum findet man es in Theophrast,Dioskurides, Strabo und Plinius.⁴³ Besonders letzterer ist von diesem Aroma eingenommen: sed omnibus odoribus praefertur balsamum, uni terrarum Iudaeae concessum.⁴⁴ Der Baum, der der Ursprung des Harzes ist, wird in der Regel mit dem Mekkabalsam, also Commiphora opobalsamum,⁴⁵ identifiziert, wenn auch mit Vorbehalten. Der Mekkabalsam wächst heute in Südarabien und Somalia, wohl auch in Ägypten; im Altertum war aber, wie das Plinius-Zitat zeigt, vor allem der in Judäakultivierte Balsam berühmt.⁴⁶ Unsere baltām-Belege sind die frühesten außer-biblischen Bezeugungen dieses Wortes. Beachtenswert ist ferner, daß das Wort auchals (zweiter) Name einer Sklavin bezeugt ist (Wunsch 1995/96: 55b unten)

    ⁴² Hier wird ein Preis angegeben, der leider abgebrochen ist: eine Mine des Harzes kostet [x] Schekel Silber. Daraus kann man wenigstens entnehmen, daß „Balsam“ in Babylon nicht das zweifache seines Gewichts in Silber kostete, wie Theophrast, Historia plantarum IX vi 4 berichtet.

  75. die Kombination steht für /ś/.

    Ouch. It is “die Kombination <-lt-> steht für /ś/.”. Sadly, when the comment appeared, I couldn’t edit it.

    If anyone is curious about die Sklavin, Wunsch 1995 Die Frauen der Familie Egibi only has this note “Z. 7f.:Beide Zeilen sind über eine Rasur geschrieben; lies gegen Bertin ᶠBal-ta (statt DA)-am-mu; wie in 33933:4 (Nr.12)”, but she appears here, p. 30-31, ᶠbal-ṭa-am-mu,  p. 33 ᶠba-al-ta-am-mu (line 4 and note 45) and especially p. 34 note 49.

  76. John Cowan says

    Well, I just calculated that if you appear naked in public, the effect will be even more striking if you behave as if you were dressed.

    Quite right. I once answered the door with no clothes on, so I just carried on normally. Fortunately, the person knew me and my oddities.

  77. Drasvi, that article by Jursa is excellent! I was unaware of this Akkadian evidence for the interpretation of onycha (p. 154f). Thanks for calling attention to that.

  78. Yes. https://en.wiktionary.org/wiki/ܛܦܪܐ has “The senses “onycha” and “onyx” are semantic loans of Ancient Greek ὄνυξ (ónux).” which does not make much sense.

  79. “There’s no problem with the mere fact that Hebrew originally preserved a distinct reflex of Proto-Semitic /ś/, of course, as that is clear from the history of the sibilants within Hebrew itself; ”

    To me it is less clear. Look at English. It was spoken on Earth during much of 21st century.
    There were all sort of vowel mergers, and yet they used the same literary langauge and the same writing system…

  80. I mean, imagine:
    community A has two sounds: s1-2, s3.
    community B has two sounds: s1, s2-3.
    They still share literary langauge (orthography, grammar, vocabulary) because of shared ancestry or convergence.

    What is this if not English?

  81. David Eddyshaw says

    This is surely the only way to account for the fact that written shin in some (not all) lexical items later coincided in pronunciation with samech, and that those three lexical groups show distinct reflexes for the relevant consonants elsewhere in Semitic. Dialect mixture can’t explain that.

  82. @DE, I am only beginning to expore Hebrew philology so anything I say is just my own precautions, not a learned opinion.

    My motivation here as it is over and over again is that I hate the perception that diversity of ancient Mediterrinean has been surpassed by modern Dagestan (or even Brittany) by orders of magnitude. They spoke Latin, Greek (with dialects), Hebrew, Phoenician, Aramaic…. Somewhere to the south there were people with camels and pre-Arabic. Surely it is simpler to assume that if we know about only N lnaguages, then there were only N languages. But it is the same sort of parsimony as in “we don’t know what is there at 55° N, 38° E, so let’s believe nothing is there: no space-time, nothing”.

    Now, no, it is not the only way. There could be a language or dialect whose speakers (1) used literary Hebrew and (2) have been merging these two all the way from proto–West Semitic.

    I am not saying that they did not have a distinct phoneme, I am saying that could have lost it. Maybe they were not Canaanites (linguistically) at all, but the Jews who came in the land of Canaan with Abraham or Moses.

    I could have taken any modern language as an example of a language where speakers systematically maintain in writing distinctions that they don’t make in pronunciation. Russian or anything. It is normal. The question is only why such community would not use a phonemic writing system.

    But in this case the explanation is that there was another community for whom this writing was more phonemical. It is accepted as a fact: this system did not represent Hebrew phonemes. It merged two sounds that Masoretes distinguish. The existence of this community is also a fact.

    I only propose that it also distinguished between two sounds that conincide for some other groups of speakers (the source of Masoretic spelling), and such a situation is so common that it is universal.

    Now I do not mean that Hebrew “did not distinguish” between these sounds. I only don’t find the argument “but how else Masoretes could know?” strong.

  83. David Eddyshaw says

    There’s nothing at all controversial in the statement that the Hebrew alphabet did not represent all the consonant distinctions of actual spoken Hebrew (regardless of dialect.) Clearcut examples are cheth and ayin, each of which represented two distinct consonants; we know this for certain, because the transcriptions in Greek letters differ.

    I hate the perception that diversity of ancient Mediterrinean has been surpassed by modern Dagestan (or even Brittany) by orders of magnitude

    I don’t know about “orders of magnitude”, but different regions plain do differ very substantially in linguistic diversity. It’s just a fact.

    Even within Oti-Volta, Mooré, Gulmancéma, Dagbani and Dagaare between them account for well over half the land area these languages are spoken in: the other twenty or so divide the rest between them. And the Atakora département of Bénin by itself is as diverse internally (just considering OV languages) as all the rest of the area, extending over Ghana, Togo, Niger, Côte d’Ivoire and Burkina Faso.

  84. “There’s nothing at all controversial in the statement that the Hebrew alphabet did not represent”

    Yes, of course. And there is also nothing controversial in the statement that the alphabet distinguished between sounds that merged in some pronunciations.

    The question is when Hebrew speakers began to mix sibilants up, and the evidence of Masoretes tells me nothing about it. Either they wrote “שׂ” because it once was same as ‎‎ס‎ or they wrote שׂ because it once was a lateral. God knows.

    Steiner (link) presents arguments in favour of 3-sibilant system and I don’t see how any date for שׂ – ‎‎ס‎ merger follows from his arguments.

  85. “orders of magnitude” – Maybe one decimal order. I associate “orders” with powers of e, and even though I meant decimal orders, the association affected me. I am not saying that anyone would actually claim: “there were only ten languages in the ancient Mediterranean” or something like that.

    But people often assume that what was spoken is what we know. Depending on context it can be just a few major languages – which gives us decimal orders for Dagestan (Brittany is rather an example of complicated dialectology) – or all attested languages. And all attested langauges is still not much. For super-Saharn Africa it “Egyptian, and a couple of obscure epigraphic languages, presumably related to proro-Berber”.

    People are more generous when something is attested later (or ideally is spoken today): then they are ready to extrapolate it back in time for the same place (as with Berber).

    The quality of attestation: consider African Romance.
    A large language with many speakers in a wast area. How well is it attested? So why things should be better for languages with comparable number of speakers in the Middle East 2000 years ago?

Speak Your Mind

*