Arabic is the best-preserved model of the Semitic languages

Discussion in 'Etymology, History of languages and Linguistics (EHL)' started by Marsario, Nov 29, 2012.

  1. Marsario Senior Member

    Helsinki, Finland
    Italia, italiano
  2. berndf Moderator

    Geneva
    German (Germany)
    Such descriptions usually refers to phonology: Arabic preserves most of the Proto-Semitic consonant distinctions while many other living and attested dead Semitic languages have merged many consonants. See here.
     
  3. rayloom Senior Member

    Paris, France
    Arabic (Hijazi Arabic)
    The article probably refers to Classical Arabic (CA).
    When it comes to CA, CA has preserved several archaic features attributed to Proto-Semitic PS, which have otherwise been lost in many living and dead Semitic languages. Some other features in addition to what berndf has mentioned:
    - Preservation of the PS word order: VSO, NG, NA. (The VSO has become SVO in Colloquial Arabic, Modern Standard Arabic still maintains the VSO order)
    - Preservation of the inflections and case endings of PS (some vestiges remain in Colloquial Arabic)
    - A functional dual (still preserved in Colloquial Arabic). A distinction of the dual in verb conjugation.
    - The PS 8 vowel system: 3 short, 3 long, 2 diphthongs. Many colloquial dialects have witnessed a monophthogization of diphthongs.

    (The wikipedia article is a good place to start)
     
  4. fdb Senior Member

    Cambridge, UK
    French (France)
    This is a very dubious statement. Arabic does retain a lot of (apparently) archaic features that are lost in many other Semitic languages, but the same can be said of Akkadian, which has a good number of (apparently) archaic features lost in Arabic and all other languages (e.g. multiple present classes in the derived stems). But since proto-Semitic is in any case a hypothetical reconstruction from the attested Semitic languages, the question of which features of the daughter languages are retentions and which are innovations is to a large extent a petitio principii. The situation is totally different from that in (say) Romance or Indo-Aryan, where the mother language (Latin, Old Indian) is actually attested, rather than reconstructed.

    To stay with phonology: Ancient South Arabian (curiously missing from the table in the Wikipedia article), Modern South Arabian, and (Masoretic pointed) Hebrew retain three voiceless non-emphatic sibilants (s1, s2, s3), while Arabic merges s1 and s3 as س and has ش for s2.
     
    Last edited: Dec 2, 2012
  5. rayloom Senior Member

    Paris, France
    Arabic (Hijazi Arabic)
    It's true that if you take just one aspect, some languages are more conservative than Arabic, as in the phonology of South Arabian. However, if you do compare Arabic with other languages in the group, Arabic ends up being more conservative.
    Akkadian for example, preserves the older verbal system of PS, but when it comes to phonology, it shows the most phonemic loss from PS. Hebrew also merges nearly 6 others, while still guarding the distinction (in writing) of the sibilants.

    Proto-Semitic is a hypothetical reconstruction, true. However, Comparative Semitists can have a clear idea whether a feature is a retention or an innovation (I have once linked to an article by Huehnergard). Let's look for example at the identical case endings, would you say it was an innovation on the part of Classical Arabic, Akkadian and Ugaritic? 3 languages from different sub-families, diachronically remote from each other!
    Unlikely...thus it can be safely assumed to be a retention from PS.
     
  6. tFighterPilot Senior Member

    Israel - Hebrew
    The distinction between the different phonemes has been almost completely preserved in Arabic (except the the equivalent of the Hebrew sin) but their phonetic value is quite different in many cases. For example, it is assumed the the original realization of the emphatic consonants was ejective, as in the southern Semitic languages, rather than pharyngealized.
     
  7. fdb Senior Member

    Cambridge, UK
    French (France)
    This is actually an excellent example of the pitfalls of extrapolating a proto-language from an arbitrary selection of daughter languages. According to the most recent experimental-phonetic research (Janet Watson) the emphatic consonants in Modern South Arabian (Mehri etc.) are not ejectives but pharyngealised (very much as in Arabic). Consonants (emphatic and non-emphatic) are realised as ejectives in pausal position only.

    This leaves us with (phonologically) ejective emphatics only in Ethiopic (for the moment at least).
     
  8. fdb Senior Member

    Cambridge, UK
    French (France)
    As I mentioned before, Arabic DOES preserve the reflex of s2 (=Heb. śin) as ش. . But s1 (=Heb. šin) and s3 (=Heb. samek) merge as س .
     
  9. Abu Rashid

    Abu Rashid Senior Member

    Melbourne, Australia
    Australian English
    A few other features Arabic has preserved that most other Semitic languages have lost are:


    • Retention of noon when it has no vowel, which is lost in other Central Semtiic languages and in Old South Arabian and probably others.
    • Retention of distinctiveness of 28 out of 29 of the original PS phonemes (not clearly mentioned above)
    • Retention of broken plurals (this is debated, but there's evidence that it existed in different sub-families, indicating it was probably an original feature)
    • Retention of a massive amount of vocabulary lost or unrecorded in other Semitic languages. Arabic has by far the largest repertoire of primitive 3-radical roots.
    • Retention of waw in w-initial roots, all other Central Semitic languages shifted these to become y-initial.

    As well as these features which are pretty much distinctly preserved in Arabic, there are many other features, as mentioned above, preserved in Arabic and a few other languages.

    Also in regards to the merger of s1/s3, some dialects of Old North Arabian still retained this distinction, and around the same time Old South Arabian had also begun undergoing the same merger, so both lanaguages retained all 29 PS phonemes for roughly the same time.

    Also what is distinct about Arabic's retentions is that they've survived right up until the present day, when most of these features were lost thousands of years ago in all other languages.
     
  10. fdb Senior Member

    Cambridge, UK
    French (France)
    I think we agree about most of this. Just let me make a few remarks on your five bulleted points:


    1. I think you are referring to the non-assimilation of n to the following consonant. Yes, Arabic does not have this innovation, but neither do many other languages (including the Mesopotamian dialects of Aramaic such as Mandaic).
    2. You mean 29 CONSONANT phonemes, I think. If so, I agree. But don't forget about Ugaritic. (I think Old Akkadian probably also had the full battery of Semitic consonants, but could not write them in the Sumerian-based script. Similarly the oldest stage of Aramaic.)
    3. Broken plurals are only in Arabic, South Arabian and Ethiopic. Are they an innovation or a retention? How are we to know?
    4. “Massive vocabulary”. Yes, but bear this in mind: Arabic has by far the largest text corpus of any Semitic language, and also a very rich lexicographical tradition.
    5. The shift of initial w to y is restricted to Aramaic and Canaanaic.

    May I add that the subdivision of Semitic into “Central”, “Southern” etc. etc. is a matter of intense disagreement among Semitists. I personally think that the arguments for grouping Arabic together with Aramaic and Canaanaic in a “Central Semitic” group are extremely weak.
     
  11. tFighterPilot Senior Member

    Israel - Hebrew
    Broken plurals are most likely a retention since there are traces of it in Hebrew as well. In Akkadian, however, there are no traces of it at all, so it might be a west Semitic innovation.
     
  12. Abu Rashid

    Abu Rashid Senior Member

    Melbourne, Australia
    Australian English
    Yes, sorry my description was not quite right. The noon is geminated to the next consonant, but in languages like Hebrew that gemination then disappeared altogether.

    Ugaritic retained 27 of them (3200+ years ago).

    As tFighter mentioned some other languages have vestiges of them.

    True. But Aramaic had a similar tradition during its time as the Lingua Franca of the Middle East, yet its arsenal of roots is nowhere near as broad as Arabic's.

    And Ugaritic... ie. pretty much all Central Semitic languages except Arabic.

    Well some of the most recognised scholars in the field group them this way. What alternative nomenclature do you suggest?
     
  13. fdb Senior Member

    Cambridge, UK
    French (France)
    The number of living “recognised” scholars of comparative Semitic can probably be counted on the fingers of one hand. I think you will have trouble finding two who share the same view on the subgrouping of Semitic.
     
  14. tFighterPilot Senior Member

    Israel - Hebrew
    That is true, but most do agree on the division between east and west Semitic. The disagreements are on smaller things. For example, my professor for Semitic linguistics (Shlomo Yizre'el) said that the old south Arabian languages were central Semitic rather than south Semitic.
     
  15. origumi Senior Member

    Hebrew
    The gemination is lost in Modern Hebrew but there are other traces of the noon: b-g-d-k-p-t are pronounced in such situation as if they were geminated. At least b-k-p.
    Hebrew of that time seems to have retained 25 (the 22 letters plus heth, ghain, sin). Some say 26, adding the ancient th based on the Shiboleth incident for example.
    Traces of broken plural in Hebrew? I am not sure there's such reliable phenomenon.
    The shift of w -> y in Hebrew is only partial. Many conjugations preserve the "w", either as "w" or "o".
     
  16. tFighterPilot Senior Member

    Israel - Hebrew
    Shiboleth indicated the difference between Shin and Samekh which still exists today.

    The traces of broken plurals I speak of are that the difference between singular and plural in Hebrew is not only the added suffix but also a change in form. For example, nouns of the form KeTeL are changed in plural to KTaLim.
     
  17. origumi Senior Member

    Hebrew
    Not agreed by everyone. I saw the other hypothesis in an article by Prof. Haim Rabin (as far as I remember).
    This is a different thing, not broken plural. Maybe traces of a secondary change during/after the shift from proto-Semitic qatl to Hebrew qetel.
     
    Last edited: Dec 5, 2012
  18. fdb Senior Member

    Cambridge, UK
    French (France)
    The expansion of *qatl (Heb. qεtεl) to *qatal in the plural occurs also in Aramaic, e.g. in ʾalp-ā ‘thousand’ (Syr. alpā, with hard p), pl. abs. ʾalap-īn (Syr. alpīn, with soft p). But I do not really think this can be called a remnant of broken plurals. The system of broken plurals in Arabic, South Arabian and Ethiopic is much more complicated than this.
     
    Last edited: Dec 5, 2012
  19. berndf Moderator

    Geneva
    German (Germany)
    Not really. The spellings with Samaekh and Shin you find in the Tanach is just an approximation in the attempt to represent a phonetic difference that cannot be expressed with the Phoenician 22-letter Abjad.

    EDIT: Crossed with #17
     
  20. Ben Jamin Senior Member

    Norway
    Polish
    I thought that Akkadian was extinct about 2000 years ago.
     
  21. berndf Moderator

    Geneva
    German (Germany)
    Yes, and?

    (We are taking about mergers that must have happened between 3500 and 2500 years ago.)
     
  22. Ben Jamin Senior Member

    Norway
    Polish
    But you use present tense.
     
  23. tFighterPilot Senior Member

    Israel - Hebrew
    The language still preserves it, it's just that nobody has spoken it for over 2500 years. Languages themselves are immortal.
     
  24. Abu Rashid

    Abu Rashid Senior Member

    Melbourne, Australia
    Australian English
    This is not really true.

    If this were the case then we could say Arabic preserves all 29 PS phonemes, because in some dialects of Old North Arabian s1, s2 & s3 were still distinct.

    We cannot speak about Akkadian in the present tense, as it no longer preserves anything. All we can say is that at such and such a point in Akkadian's history it preserved such and such feature like this. Just because that history stopped progressing 2000 years ago, doesn't change anything.
     
  25. fdb Senior Member

    Cambridge, UK
    French (France)
    I do understand Abu Rashid’s point: if we are asking about a “well-preserved” language we are in fact talking about a modern spoken language and asking how well it has “preserved” the phonological and morphological systems of proto-Semitic. From the morphological point of view, modern Arabic is of a Middle Semitic type (e.g. total loss of the case and mood endings, as was, by the way, the case already in Neo-Babylonian and Neo-Assyrian, judging by the erratic spelling of the final vowels of nouns and verbs). At a phonological level, many Arabic dialects of the bedouin and rural types still distinguish all the Classical Arabic consonant phonemes apart from ض (which merges with ظ except in a small number of Arabic dialects in Southern Arabia), while the dialects of the urban type have reduced the repertoire of consonants to an extent comparable to Middle Aramaic. The full range of Semitic contrasting consonants is, however, retained in some of the Modern South Arabian languages, e.g. Mehri.

    Arabic, as it was spoken at the beginning of the Islamic era, was probably “better preserved” than Akkadian in the Neo-Babylonian period, more than a thousand years earlier. But it was less “well preserved” than Old Akkadian two millennia before that.
     
    Last edited: Dec 20, 2012
  26. tFighterPilot Senior Member

    Israel - Hebrew
    From my understanding, Akkadian never preserved all the different phonemes due to the influence of the non-Semitic Sumerian.
     
  27. Ihsiin

    Ihsiin Senior Member

    England
    English
    Really?
     
  28. fdb Senior Member

    Cambridge, UK
    French (France)
    Yes, the emphatic lateral represented by ض merged with the emphatic interdental ظ in Middle Arabic. The realisation of ض as [ḍ] and of ظ as [ẓ] is a modern reading pronunciation. Lateral ض survives only in the Modern South Arabian languages and a few Arabic dialects adjacent to the South Arabian speech zone.
     
  29. fdb Senior Member

    Cambridge, UK
    French (France)
    This is debated. There is no doubt that the cuneiform script does not adequately reflect the sound system of Old Akkadian. But the fact that in Old Akkadian and Old Assyrian the vowels before and after a Semitic laryngeal are not contracted into single vowel (as they are in Old Babylonian) suggests that the laryngeals were still pronounced.
     
  30. Ihsiin

    Ihsiin Senior Member

    England
    English
    Yes, I'm aware of the merger, but I didn't think there were any Arabic dialects that did not merge the two.
     
  31. fdb Senior Member

    Cambridge, UK
    French (France)
    Lateral ض has been reported in the dialect of Dathina in southern Yemen.
     
  32. fdb Senior Member

    Cambridge, UK
    French (France)
    ..... but I am sorry that I did not realise which part of my convoluted sentence you were asking about.
     
  33. rayloom Senior Member

    Paris, France
    Arabic (Hijazi Arabic)
    The simple present can be used to that extent :)
     
  34. berndf Moderator

    Geneva
    German (Germany)
    Arabic is Arabic and "some Old North Arabian dialects" are "some Old North Arabian dialects". Those are different things.
     
  35. Ihsiin

    Ihsiin Senior Member

    England
    English
    I wonder whether this lateral dhad was preserved, or re-imported from South Arabian. I suspect the latter?
     
  36. fdb Senior Member

    Cambridge, UK
    French (France)
    Or preserved under the influence of a South Arabian substratum.
     
  37. Abu Rashid

    Abu Rashid Senior Member

    Melbourne, Australia
    Australian English
    Modern Arabic differs no more from Old North Arabian than Modern Hebrew differs from Biblical Hebrew, in fact probably less. I don't see why you would consider one case the same language and the other not.
     
    Last edited: Dec 31, 2012
  38. berndf Moderator

    Geneva
    German (Germany)
    I would, if we were discussing features where Biblical and Modern Hebrew differed. Using labels like Arabic, Hebrew, English or German as an umbrella term for a multitude of dialects and development stages is always a(n over-)simplification which useful where the differences don't matter or when we explicitly mean is as a collection of dialects and development stages.
     
  39. Abu Rashid

    Abu Rashid Senior Member

    Melbourne, Australia
    Australian English
    Agreed, but the context was of Biblical Hebrew having retained all 3 sibilants. Modern Hebrew, like modern Arabic, does not, and therefore that point is no more relevant than the fact that Old North Arabian retained them as well.

    I still do find it strange though that Biblical Hebrew & Modern Hebrew are considered part of the same language continuum, as are the oldest attestations of Aramaic with the neo-Aramaic languages, whilst Modern Arabic and Old North Arabian are not.

    In the case of Aramaic stages/dialects they are as far apart as most Semitic languages are from each otrher.
     
  40. Youngfun

    Youngfun Senior Member

    Pekino, Ĉinujo
    Chinese/Italian - bilingual
    I could argue that Modern Standard Arabic is a sort of artificial language made intentionally more similar with Classical Arabic, and thus more similar with the ancient Proto-Semitic language.
    While the "natural" Arabics that people speak - Arabian Arabic, Gulf Arabic, Lebanese Arabic, Egyptian Arabic, Algerian Arabic, Moroccan Arabic, etc. - have all diverged much more from Classical Arabic, and thus from Proto-Semitic language.
     
  41. Abu Rashid

    Abu Rashid Senior Member

    Melbourne, Australia
    Australian English
    You could, but even if we look at all other Semitic languages at the time Classical Arabic was frozen, it was still worlds apart from them in terms of archaic retentions, except for maybe Old South Arabian.
     
  42. berndf Moderator

    Geneva
    German (Germany)
    That is just a matter of naming convention, not of substance.:)
     
  43. origumi Senior Member

    Hebrew
    I cannot compare to Arabic, yet convinced that Biblical Hebrew and Modern Hebrew are the same language with some different preferences in regard to vocabulary and grammar. Biblical-style Written texts are perhaps >95% intelligible by most educated people. Even sister Canaanite languages can be understood (in writing), not fully though, by Modern Hebrew speakers provided that punctuation (nikkud) or matres lectionis are added. I guess that the sounds are different enough to break the mutual understanding. Even listening to Modern Jewish Yemenite old guys speaking Hebrew - is a great challenge.

    Look at the first sentences of the Mesha Stele:

    אנכי מישע בן כמוש מלך מואב הדיבוני. אבי מלך על מואב שלשין שת ואנכי מלכתי אחר אבי ואעש הבמות זאת לכמוש בקרחה ... כי הושיעני מכל המלכין וכי הראני בכל שונאי עומרי מלך ישראל ויענו את מואב ימים רבים כי ינאף כמוש בארצו ויחליפה.


    Pick a Hebrew speaker in the street and he'll understand, with some doubt on the bold letters.
     
    Last edited: Dec 31, 2012
  44. berndf Moderator

    Geneva
    German (Germany)
    I also understand Bavarian and in most cases I wouldn't differentiate between German and Bavarian. But when I am analyzing characteristics where the two differ (e.g. Bavarian doesn't have a simple past and the present perfect has absorbed all of its uses while Standard German differentiates between the two), I would of course regard German and Bavarian as different languages. If you call Biblical Hebrew and Modern Israeli Hebrew as two development states of one language or whether you call the one a predecessor language or the other (you could call modern Hebrew "Israeli", like you call Italian "Italian" and not "Modern Latin") is a factually and conceptually insignificant terminological convention; much like it is linguistically completely irrelevant, whether you talk of a Serb-Croatian language with a Serbian and a Croatian dialect of of a Serbian and a Croatian language. The difference between language and dialect is political and/or cultural and not linguistic.
     
  45. origumi Senior Member

    Hebrew
    The situation here is different. Almost every component of good Biblical Hebrew is regarded as good Modern Hebrew, high-register though. In this sense, the attested Biblical language is a subset of the modern language. Many of the exceptions were most likely regarded as archaisms even in Biblical times. I guess it would be considered as "incorrect language" to imitate Bavarian while speaking German. It's absolutely "correct" and usually beautiful to take words, sentences, idioms, grammatical structures from the Bible when writing or speaking Hebrew to audience. The old language is very much alive for the Modern Hebrew speaker.
     
  46. berndf Moderator

    Geneva
    German (Germany)
    Nevertheless they are different languages. In the Vatican, e.g. both Italian and Latin are "much alive"; that is no reason to call them the same language.
    We also use Latin quotes in modern European languages. That doesn't make them the same as Latin. And your comment underlines what I said: The difference between dialect and language is a cultural and political distinction.
     
  47. Hulalessar Senior Member

    Andalucía
    English - England
    Is the point not so much that Modern Hebrew is a continuation of Ancient Hebrew (as Italian can be said to be a continuation of Latin) but rather that it is Ancient Hebrew* revived some two thousand years after it stopped being spoken? In other words is it not more like the Latin used in the Vatican, that is essentially the language used in Rome two thousand years ago but modified so as to be useful in the modern world, rather than Italian?

    *or some variety of Ancient Hebrew as it must have changed over the centuries in antiquity
     
  48. Abu Rashid

    Abu Rashid Senior Member

    Melbourne, Australia
    Australian English
    Anyway my point simply was that if we can accept Hebrew had 3 sibilants in a recorded stage of the language, then we can accept the same for Arabic. Undoubtedly all Semitic languages descend from a form which had 3 sibilants anyway, so the only point to really make is whether or not such a form was ever attested. For both Hebrew & Arabic, yes there are older forms of these languages which are attested with 3 separate sibilants. Whether we want to consider those forms of those languages as different languages or earlier stages of the same language is really a moot point.

    I think also it's worth noting that Arabic & Aramaic are almost like sub-families themselves anyway. There are enough varied "variaties" of both languages to warrant considering them to be sub-families IMHO. There's about as much variance between dialects/stages of Arabic as there is amongst Canaanite languages.
     
  49. tFighterPilot Senior Member

    Israel - Hebrew
    Indeed it has. I think one can even claim that Mishnaic Hebrew (early AD) is closer to modern Hebrew than to Biblical Hebrew.
     
  50. origumi Senior Member

    Hebrew
    But then Modern Hebrew is not closer to Mishnaic Hebrew than it is to Biblical Hebrew. Also, some claim that the term Mishnaic Hebrew is misleading, this is a dialect that existed already in Biblical times, in parallel to the "official" Biblical Hebrew (Prof. Haim Rabin et al., see Wikipedia for עברית משנאית).

    We are far off-topic. Maybe the moderators would consider splitting the Hebrew discussion to a new thread.
     

Share This Page