Talk:Proto-Semitic language

From Wikipedia, the free encyclopedia

This article is within the scope of the WikiProject Languages, an attempt at creating a standardized, informative, and easy-to-use resource about languages. If you would like to participate, please visit the project page, where you can join the project and see a list of open tasks.
??? This article has not yet received a rating on the quality scale.

Contents

[edit] Egyptian and Semitic

The article states that Egyptian is the closest of the Afroasiatic languages to Semitic, but then proposes that Proto-Semitic entered the Middle East paradoxically from the Horn of Africa!!! This seems to be a very paradoxical arrangement, as Afro-Asiatic languages probably only appeared in the horn of Africa after an emigration of Cushitic peoples from the Sudan, some 3,000 years BCE!!

Given the fact that Egyptian is the closest of the Afro-Asiatic languages to Semitic, isn't it more likely that ProtoSemitic speakers travelled across from Egypt to Sinai, possibly at the end of PPNB stage, with the Harifian hunter gatherers, to form what Jaris Yurins calls the Circum Arabian Nomadic Pastoral Complex in the period from 6,000-5,600 BCE? It was these peoples who, as a part of the Arabian bifacial complex, introduced the neolithic into Southern Arabia.

John D. Croft 20:22, 26 December 2006 (UTC)

Relations within Afro-Asiatic are not known. Some researchers say Egyptian is closer to Semitic than other branches, but others say that all branches are equally disparate/related. Note that those who relate Egyptian to Semitic also relate it to Beja; this raises the possibility of Proto-semitic having arisen in the Red Sea region of Sudan/Eritrea. Don't forget that there could have been multiple dispersal points for PS, too. Both a Sinai and Bab al-Mandeb pathway, possibly even with more dispersal points along the western Red sea (trade across it is very ancient) coast. Also, Cushitic is much older than 3000 BCE in Ethiopia. Ethiopia is (currently) the most likely home for Proto-AA, and Omotic split off from PAA about 11kya, and Cushitic about 9kya. Southern Sudan is the home of Nilo-Saharan speakers, not Cushitic, so a migration from there is very unlikely. — ዮም | (Yom) | TalkcontribsEthiopia 21:07, 26 December 2006 (UTC)
Thank you Yom, for your posting. Regarding the Beja language, it is true that there is a close connection with Egypt. Coming from the Medjay, Beja tribesmen served as a policeforce in Egypt at least from the Middle Kingdom if not before and there may have been borrowing of Egyptian words and grammar into Bedja from that time.
Regarding the Southern Sudan as a homeland of the Proto-Afro-Asiatic, I have done significant work on the neolithic archaeology of this region and it appears that there were a number of waves of cultural influence travelling from the Southern Sudan into Northern Kenya (from 6,000 years BCE) and then into the Ethiopian Highlands, reaching the Eritraean coast about 3,000 years BCE. Prior to this, on genetic and other grounds, it appears that this area was inhabited by a Capoid people, related to the South African Bushmen, and probably speaking a Khoisanid language, perhaps related to the Jazda or Sandawe. These waves allow us to relate them to the various sub-groups of Cushitic fairly closely, which makes Southern Sudan a prime candidate for being the homeland of Proto-Cushitic. It was the arrival of Nilo-Saharan speaking Nuba and Dinka people into the region, during the 2nd Intermediate period in Egypt (they are portrayed in New Kingdom reliefs) that may have led to the eclipse and disappearance of Cushitic languages from this region. Thus Southern Sudan is unlikely to be the Urheimat of Nilo-Saharan people. Generically they carry the L2 marker which shows a West African origin, showing that they moved into the Southern Sudan probably from the Darfur region, with the increasing dessication of the Sahel.
Thus while on Linguistic grounds you may argue that Ethiopia was the Urheimat of the Proto-Afro-Asiatic, it doesn't add up archaeologically. Like Highland areas elsewhere, they tend to be places of linguistic refugaria, where languages that have been displaced from lower, more favourable climes survive in isolation. This is clearly shown archaeologically, as the archaeology shows movements into Ethiopia (generally either up the Blue Nile and Atabara to the Highlands or across the Bab El Mandib through Afar and Eritrea into the Highlands.
Regarding splits in Proto-Afro-Asiatic, I find Chris Ehret's work seems to me to accord most closely with what is known archaeologically. Thus what we find seems to be that Omotic split first from PAA, possibly between 10,000 and 8,000 years BCE. Archaeologically this seems associated with the movemnt of East African microlithic Capsian culture from the Southern Sudan into Northern Kenya. The second split from PAA seems to have been Cushitic. They share the SOV structure of Omotic and a number of other early features. But archaeologically they seem related to the waves of neolithic people who entered the Ethiopian Highlands from the area of the Southern Sudan from about 6,000-3,000 years BCE. Chadic seems to have been the third split from PAA, moving into the region around Lake Chad from the Tibesti and Hoggar region with the beginning of the dessication of the Sahara from about 6,000 BCE. This leaves the VSO group of Afro-Asiatic people, the Berber, Egyptian and Semitic people as the last group of AA's to split. The first of these to separate is the Berber, and their movements seem to have been to the west from out of the Fezzan region, from about 4,000 BCE onwards, moving into the Atlas Mountains with the spread of Neolithic into this region. Their split with the Proto-Egypto-Semitic group seems to have occurred between 8,000-6,000 BCE. This seems archaeologically related to the spread of Domestic cattle in the Wet-Sahara phase, and the wavy line pottery, which some claim was Nilo-Saharan, but on genetic grounds this hypothesis must be rejected. The split between Proto-Egyptian and Proto-Semitic then seems to have occurred between 7,000-6,000 BCE and seems associated with the rise of the Nile community and the shift archaeologically of the Harifa culture, with an African Outacha retouch technology of microliths across the Sinai from Fayyum. Ancient Egyptian, it is suggested recently, developed out of a pidgin trade language that united the various scattered communities along the Nile from Aswan to Asyut, and which was later imposed upon the north of the country with the unification under Narmer and Aha of the 1st Dynasty.
Putting all of this evidence together, we have a PAA Urheimat that extends from Fezzan to the Southern Sudan, and seems centred in the Eastern Sahara region east of the Tibesti and Hoggar region, but largely west of the Nile. What we can then say is that the spread and dispersal of PAA seems to be associated with the dessication of the Sahara, from the Wet Pluvial to the present desert conditions of today. This is clearly shown with the Nabta culture of the Sahara (which currently seems the best origin for Egyptians) and the subsequent drying out of Wadi Hammamat during the Naqada III phase.
Hope this helps
Regards John D. Croft 08:53, 27 December 2006 (UTC)

[edit] Proto-Semites?

the question is raised: if there was a language called proto-semitic, then was there a specific group of semites that spoke that language known as proto-semites or the proto-semitic people?

The term Semitic is used here in a very specific sense. It is applied to a broad group of languages that show sufficient similarity for it to be postulated that they are genetically related. Note that the languages are genetically related, not the people. The language any one person speaks is not dependent on their biology, but on a number of social and geographical factors. Yes, if Proto-Semitic ever existed (as it is only an academic reconstruction, and has never been discovered), a group of people must have spoken it. To suggest that this postulated group are the ancestors of every person who now speaks a Semitic language is highly improbable. Sometimes I read that we should do away with the term Semitic in language contexts because of its links with the biblical character Shem: it is a biblical word used as a scientific term, and the two are best not confused. Gareth Hughes 12:11, 1 Mar 2005 (UTC)
i've heard that after proto-semitic, there have been reconstructed proto-west semitic, proto-central semitic, and proto-east semitic languages. does anyone on here know anything about this?Gringo300 12:59, 15 Jun 2005 (UTC)
i've done more research on the semitic language family... the more research i do the more confused i get. Gringo300 10:08, 12 August 2006 (UTC)

[edit] phonology

we cannot just say "generally reconstructed" here. There is no consensus on the exact PS sound system. You need to say who reconstructs this particular system. Afaik, the main division is between people basing PS on Akkadian vs. on Arabic. The present version seems to just "fill up" the system with lots of phonemes, so that all descended systems can be comfortably explained. dab () 08:24, 15 Jun 2005 (UTC)

Here is a replication of the consonantal inventory presented in:
Sáenz-Badillos, Angel [1988] (1993). "Hebrew in the context of the Semitic Languages", A History of the Hebrew Language (Historia de la Lengua Hebrea), trans. John Elwolde, Cambridge, U.K.: Cambridge University Press, 18-19. ISBN 0-521-55634-1. 
Point of articulation Manner of articulation Degree of aperture
Voiceless Voiced Emphatic
Bilabial p b Plosive
m Nasal
Interdental t̠̣ Fricative
d̠̣ Lateralized
Dental t d Plosive
s z Fricative
l Lateral
ś Lateralized
r Vibrant
n Nasal
Prepalatal š Fricative
Dorsopalatal k g q Plosive
Velar ġ Fricative
Pharyngeal Fricative
Laryngeal Plosive
h Fricative
I'm not at all clear on what Sáenz-Badillos means by his use of <d̠̣> - the notation would seem to indicate an emphatic interdental voiced fricative (which shouldn't be possible - emphatics cannot be voiced), but is also noted as "lateralized". Perhaps [ɬˁ], but then why doesn't he place it in a single voiceless/voiced/emphatic series with <ś> / [l] and l / [l]?
Note also that while he mentions the existance of semivowels [j] and [w], he omits them from this chart.
Sáenz-Badillos cites his sources as follows:
On the consonant inventory in general:
  • Cantineau, J., 1951: 'Le consonantisme du sémitiqueé', Semitica 4, 79-94.
  • Martinet, A., 1953: 'Remarques sur le consonantisme sémitique', Bulletin de la Société de Linguistique [de Paris] 49, 67-78.
  • Moscati, S., 1954: Il sistema consonantico delle lingue semitiche. Rome.
On the affricate realization of in Semitic:
  • Steiner, R.C., 1982: Affricated ṣade in the Semitic languages. New York.
On a proposed sixth sibilant:
  • Goetze, A., 1958: 'The sibilants of Old Babylonian', Revue d'Assyriologie et d'Archéologie Orientale 52, 137-49.
On the sibilants generally:
  • Faber, A., 1984: 'Semitic sibilants in Afro-Asiatic context', Journal of Semitic Studies 29, 189-224.
On the nature of the emphatics (ejectives vs. pharyngeals):
  • Moscati, S. 1969 (ed.): An introduction to the comparative grammar of the Semitic languages. Phonology and morphology. 2nd ed. Wiesbaden. (23f.)

[edit] South Arabian

Why no mention of South Arabian alphabet?--JWB 21:45, 24 July 2005 (UTC)

where, and why? dab () 11:12, 20 January 2006 (UTC)

[edit] Canaanite phonemes

the Canaanite table entries seem to just echo the Proto-Semitic ones. I suppose they should be changed to transliterations of the Phoenician letters in the same column? dab () 11:09, 20 January 2006 (UTC)

also, since Hebrew is also Canaanite, the "Canaanite" column should probably be titled "Phoenician", or the Hebrew column should be dropped. dab () 11:13, 20 January 2006 (UTC)

[edit] How about using IPA?

I found this article extremely hard to follow, mostly because there is no use of IPA. I know these symbols are widely used to represent afro-asiatic languages, but to a person like me whose previous knowledge of language pronunciation is in IPA, this is really hard to follow.--ikiroid | (talk) 04:49, 22 January 2006 (UTC)

well, when we manage to invite a speaker of Proto-Semitic to our phonetics lab, we'll let you know. Of course we can discuss surmised IPA values of the phonemes, but it should be borne in mind that the notation symbolizes reconstructed phonemes. I suggest we add conjectured IPA values to the phoneme table, much like on Sanskrit (we don't write Sanskrit in IPA either, even though that would be possible). dab () 12:28, 22 January 2006 (UTC)

Either the reconstructed phonemes are just abstract labels for correspondences among the various attested languages or they are intended to have some phonetic value. If they're just abstract labels, it doesn't matter what symbols are used; they could be P1, P2, ... (for "phoneme-1" etc.), or slightly more mnemonic symbols, but it makes no difference in principle. I doubt that anyone intends that. The alternative is that they express some idea as to what they sounded like, no matter how tentative, rough, or approximate, just as descriptive terms like "interdental fricative" do. In that case, there is simply no excuse for using any symbols other than IPA. It's much easier for a Semitist to guess what is meant by IPA symbols, even if they are not what s/he is accustomed to using, than for someone who is new to Semitic linguistics to understand the ideas being expressed while at the same time also trying to keep track of a whole host of idiosyncratic symbols. (I'm a Semitic linguist myself.) Maybe in a book for Semitic specialists the traditional symbols are fine (though as far as I'm concerned they ought to be abandoned) but certainly not in a work for the general public.

I beg to differ. We use standard transliterations for all sorts of languages all over Wikipedia. Nobody would use IPA to transliterate Arabic or Sanskrit for example. The transliteration has to be explained somewhere, but our guiding principle has to be academic convention. These conventions exist for a reason, they are time-tested and useful. The "general public" won't know what to do with an IPA symbol for "pharyngealized voiceless dental occlusive" any moder than with a t with an underdot. You have to provide an IPA key to the convention somewhere (and we do), but you'll just use the convention after that. dab () 07:40, 23 March 2006 (UTC)
I agree with you, if you're talking about transliteration, which is a representation of the original spelling. But I was talking about how to represent the sounds that we figure were spoken in ancient times. The general public doesn't know how to deal with any symbols for unfamiliar sounds. It's especially when we're writing for the general public that we ought to use a reasonably uniform set of symbols for all languages; why should a general reader, who doesn't intend to become a specialist, have to learn one set of symbols to find out some basic information about Semitic and a different set of symbols to find out about some other language family? So there's no need to use IPA to transliterate Arabic or Sanskrit, but if you want to tell how Arabic خ is pronounced, well ch is ambiguous, kh means something different when you're transliterating Sanskrit, h with a breve under it or k with a line under it require explanation; we should the IPA symbols x or χ, which will refer to the same sound(s) in articles on Russian, German, Brazilian Portuguese, etc. An excellent example is the ś for a lateral fricative. The same symbol ś is used for the sound [ʃ] in Sanskrit. The fact that ś was chosen to represent a certain Hebrew letter, simply to show the particular way Hebrew spells the sound [s] in certain words, long before anybody had an idea that it came from a lateral, is a poor reason to use it to represent a lateral sound when we're not even talking about Hebrew but about Proto-Semitic. It is indeed a matter of efficiency for the general reader.
(Incidentally, I'm not opposed to deviating from the IPA when the IPA symbol is really inadequate. For example, the IPA symbols for pharyngealization are typographically poor. The tilde through the letter is hard to see, and the raised ʕ diacritic is bewildering if a whole syllable or word is pharyngealized, so I use the traditional Semiti(ci?)st underdot.)

[edit] Ge'ez

what is the basis of your change ([1](, Yhever? Any reference? Or how will we tell if this was an improvement? dab () 17:24, 30 January 2006 (UTC)

ሠ ፡ ንጉሥ and ሰ ፡ እሳት were interchanged in several articles on Ethiopian languages. I am at home right now, so the only sources I have right now are:
Wolf Leslau's "Concise Dictionary of Ge`ez" (Otto Harrassowitz, Wiesbaden 1989), p. ix: "Note that in the present-day pronunciation of Geez some consonants have lost their original pronunciation. Thus, አ ʔ and ዐ ʕ are pronounced as zero; ሀ h, ሐ and ኀ are merged in the pronunciation as ሀ h; ሥ ś and ስ s; ጸ and ፀ as ጸ ."
and: Thomas L. Kane's "Amharic-English Dictionary" (Otto Harrassowitz, Wiesbaden 1990), vol. 1, p. 437: "ሰ or ሠ to distinguish between the two letters which are now identical in pronunciation, ሰ is called ሰ ፡ እሳት (the used in spelling əsat) while ሠ is called ሠ ፡ ንጉሥ the used in spelling nəgus."
Now the words ንጉሥ and እሳት which are used to distinguish these two ancient sounds, both have good Semitic etymologies: ንጉሥ nəgus corresponds to וְהַנֹּגְשִׂים wə-han-nōgəśīm (Ex. 5:13) in Hebrew and النجاشي an-nağāšī in Arabic; whereas እሳት əsat corresponds to אֵשׁ ʔēš in Hebrew, išātum in Akkadian.
You can check the correspondances in the table. In fact, שׂ in Hebrew and in Aramaic should also be transcribed as ś. yhever 00:33, 1 February 2006 (UTC)
Oh, bugger, it seems I created Śat under mistaken assumptions. I don't know about transcribing שׂ as ś though; see Hebrew transliteration. For Hebrew, it is enough to distinguish š and s, no need for a ś. But why are you transcribing ፀ as ṣ́? Is that so in new literature? My Gesenius has , I think. I really have to check Kienast on this, but I don't have it handy right now. dab () 08:04, 1 February 2006 (UTC)
Hebrew שׂ was written in the originally unpointed text with a ש and then, the Masoretes added a small ס above the letter in words in which the pronunciation was (at the time of the Masoretes, not the time of the original composition) s; These words that have שׂ happen to correspond to words with š in Arabic (as in: Heb. עֶשֶׂר ʕéśer ≈ Arab. عَشَرَة ʕašara(tun); Heb. שָׂנֵא śānē ≈ Arab. شَنَعَ šanaʔa), whereas ס and שׁ correspond to words with s in Arabic (Heb. שָׁמַע šāmaʕ ≈ Arab. سَمِعَ samiʕa; Heb. סַמִּים samm-īm ≈ Arab. سَمّ samm(un)).
The whole point of transcribing שׂ as ś is to differentiate the two different s sounds according to their etymology -- and by assuming that originally this שׂ had a special sound, different from both שׁ and ס (remember that originally the letter that was chosen to spell this phoneme was ש and not ס).
The same goes for Ethiopic ፀ; in modern pronunciation there's no difference between ጸ and ፀ (both are pronounced: ), but ፀ corresponds to Arabic ض (e.g. Ge'ez ፅብዕ ṣ́əbʕ ≈ Arab. ضَبْع ḍabʕ). Now, Leslau transcribed this letter according to the corresponding Arabic sound, but ejective phones cannot be voiced, so that couldn't have been the original pronunciation. On the other hand, a voiceless alveolar lateral fricative (as in Welsh "ll") can explain both Proto-Semitic ś and ṣ́ -- thus forming a triangle of voiced:voiceless:emphatic (l:ś:ṣ́) on the analogy of d:t:, g:k:, etc.
This theory has been shown in several studies, but I don't have any of them here, so I cannot cite from them. The only reference I can give for the meantime is the article on voiceless alveolar lateral fricative mentioned earlier in which it is said: "The sound is assumed as a phoneme for Proto-Semitic, usually transcribed as ś, since it evolved into Arabic [ʃ], Hebrew, [s]".
Even if this theory is not true, it is better to transcribe ፀ according to its traditional pronunciation () and add a distinctive mark to differentiate it from ጸ -- both etymologically, and probably pronunciation-wise. yhever 19:38, 1 February 2006 (UTC)

[edit] Sound System, editors gone crazy, hehe

Okay, there seems to be a major problem here. Certain recent edits on the subject of the sound inventory are making me crack ;)

You could say that all Proto-Semitic sounds are theoretical and tentative so it seems meaningless to say that some are. Actually, I know there could be a big long paragraph on the controversies of some Proto-Semitic sounds, so maybe that would be better than this arcane asterisking that nobody outside of our tiny Proto-Semitic world understands. Wouldn't that be swell? --Glengordon01 22:11, 12 July 2006 (UTC)


That would be nice, provided we can come up with citations. The book I quoted above in #phonology has some nice citations on the ejective vs. pharyngeal debate, as well as the affricates. I'll try to dig them up on inter-library loan, if I can. --Peter Farago 22:21, 12 July 2006 (UTC)
A paragraph discussing the possible sounds would be great and infinitely better than an unexplained list. The article right now is barely above a stub, while there's actually plenty to discuss regarding its homeland, phonology, Grammar, etc. — ዮም | (Yom) | TalkcontribsEthiopia 23:09, 12 July 2006 (UTC)
I agree, remove the asterisks from the table and explain the intricacies of 'emphatic' sounds in prose. Since it is made clear that the IPA values are "tentative" anyway, it is poinless to single the emphatic ones out as "extra tentative". Do add material explaining the debate, though! dab () 09:39, 13 July 2006 (UTC)

[edit] Underlines

In the sound table, the link underlines make it difficult to distinguish the alveolar stops from the interdental fricatives, since the latter are usually indicated with a macron below. Is there anything we can do about this? --Peter Farago 19:10, 13 July 2006 (UTC)


I think the links should be taken out because they are out of context. This table is showing us the sounds of Proto-Semitic, but the links are taking us to letters common in the Semitic writing systems. We shouldn't be mixing up language with writing. Proto-Semitic was spoken long before the alphabet came into existence! Perhaps to the ire of others, I think I'm going to take them out and if someone has a good reason to revert, let us know. --Glengordon01 19:40, 15 July 2006 (UTC)


[edit] let's use IPA *please*!!

i just made a bunch of updates to this page, including adding IPA symbols where the traditional semitic ones are non-IPA, and adding phonology for modern hebrew and modern south arabian. for the latter two, i unavoidably had to use IPA, and now the mixture of IPA and traditional semitic looks really stupid. let's please just use IPA everywhere, with the traditional semitic representation in parens in the proto-semitic column, rather than the other way around.

Benwing 05:13, 26 July 2006 (UTC)


This sounds sensible. Keeping IPA the default throughout Wikipedia respects modern linguistic standards afterall! I think if the masses are too stupid to learn IPA, a pox on their houses. Science shouldn't be dumbed-down for the sake of mental troglodytes, hehe :) --Glengordon01 06:28, 26 July 2006 (UTC)

Actually, maybe I should clarify that this should be the standard in tables and such. Of course, we still need symbols to transliterate Semitic languages in paragraphs and the like, because IPA is so ugly when used for transliteration of languages with non-Roman alphabets. So IPA in tables and charts, good; IPA in paragraphs for use in transliteration, not so good. --Glengordon01 06:36, 26 July 2006 (UTC)

OK with me. Benwing 05:26, 31 July 2006 (UTC)

OK, I rewrote the whole table to use IPA, with Semiticist symbols in parens. Benwing 06:42, 2 August 2006 (UTC)

I am sorry, but why would you do such a thing? Use IPA when discussing phonetics (it's called the International Phonetic Alphabet), use standard symbols when discussing phonology. Most of this article by far discusses phonology, no IPA needed. THe statement "Keeping IPA the default throughout Wikipedia respects modern linguistic standards afterall" is completely mistaken, failing to recognize the difference between phonetics and phonology. dab () 06:56, 2 August 2006 (UTC)
Pointlessly argumentative. You've quoted me out of context (as usual), with a spin that I didn't intend (as usual). --Glengordon01 13:59, 12 August 2006 (UTC)

sound changes fall squarely within phonology, and IPA is used everywhere for sound changes. maybe you forgot the /f/ vs. [ɸ] distinction? If you look in e.g. "The World's Major Languages", you see use of IPA whenever possible, since the various systems used in each historical linguistic tradition appear as random line noise to those outside the tradition. Especially for a comparison across a number of languages, we really need to know what actually happened to the sounds; the traditional notation has two problems here: [1] it's opaque except to a small number of people in the Semiticist camp; [2] by using cover terms that cover huge sound changes, it disguises the depth to which changes have taken place. By Using IPA symbols, we can see directly what happened and has led to the differences.

Benwing 08:14, 2 August 2006 (UTC)

Dab, I object to your single-handedly reverting my change. There seems to be a consensus in favor of IPA, and you're the only one objecting. I put my change back. Do not just revert again; get a consensus in favor of this if you really want to. BTW I made a number of fixes to the Ge'ez data in the process of my change (based on the Ge'ez page and [2]), which you also erased. Benwing 00:47, 3 August 2006 (UTC)

I have also reverted your changes for the reasons given by Dab, Benwing. Furthermore, your changes to the Ge'ez phonology were not correct. ሰ represented historical ʃ, s, and θ, but these merged into "s," at least for the common era (see Wolf Leslau (1987), Comparative Dictionary of Ge`ez). Regarding the source given, its determination that ዘ represented IPA ts' is incorrect. Though ṱ merged with ṣ, it was represented with ጸ. Moreover, ፀ is usually transcribed as ṣ́, rather than ḍ (as it is in Arabic). — ዮም | (Yom) | TalkcontribsEthiopia 01:40, 3 August 2006 (UTC)
look, consensus doesn't trump policy. Policy says we are bound to academic consensus, which may not be overruled by roll-your-own wikiconsensus. The field relevant to Proto-Semitic is Semitology (qua historical lingustics). No expert uses IPA notation for Proto-Semitic, and hence we shouldn't, either. Even if for some reason people came up with the consensus to use the Gothic alphabet in this article (I know I am exaggerating), we still shouldn't do that. dab () 17:05, 12 August 2006 (UTC)
Dab, you're arguing against your own muddled confusion. I agree with you of course. Who wouldn't. But speaking for myself at least, I never ever stated, nor ever meant, that Proto-Semitic should be written with IPA symbols! You completely misunderstood. Semitologists have their own symbols as do IEists or Uralicists, of course. However in the specific case of the phonology table (which is what I was referring to), it's perfectly valid to offer suggested values based on what would be expected if Proto-Semitic were a living human language. IPA is the standard in linguistics as I previously stated before you started twisting things out of context. --Glengordon01 17:18, 12 August 2006 (UTC)
how am I arguing against my own "muddled confusion", good Sir? I didn't even talk to you. Thank you kindly for agreeing with me, try to do it without your habitual condescen..sion some time. I was replying to Benwing above who argued that I was the only one opposing IPA. Now with your shining intellect siding with my humble troglodite mind, I am convinced Benwing will not dare pursue the point. In case you mean to say the phonology table should supply hypothetical phonetic values for Semitologist symbols, as it has done for ages, up to Benwig's change, btw, we have no quarrel. dab () 17:40, 12 August 2006 (UTC)


I suppose it is a subtle point, but the sound represented by <*ś> is not fully known, we can suggest, and mainstream semiticists do, what phonetic value it may have had, but we also know that its phonetic value changed as the languages developed. Thus, retention of the symbol <*ś> is not done because semiticists can't be bothered to learn IPA (I'm a semiticist who uses IPA regularly). This symbol is used because it represents an entity in the proto-Semitic language that developed into different entities in different languages, whose phonetic value as unknown shouldn't be tied too closely to it. Another example: in some Semitic languages the entity <*ḥ> has merged with <*ḫ> and softened <*k> so that they all represent the same phoneme. However, it's important to represent them in transliteration as different entities when it is important that the historical roots of words are compared. I'm not sure if that's clear, but that's the reason why we use them. — Gareth Hughes 17:45, 12 August 2006 (UTC)

[edit] citation needed

"The most probable Proto-Semitic Urheimat is the Arabian peninsula. This hypothesis is based on fact that the Canaanite, Aramaic, and Arab nomadic tribes are recorded to have emerged from there. The same area of origin is likely for the Akkadians."
Every single one of these statements requires citation.


I agree. Many think Proto-Semitic was in Syria or Palestine actually. So to say "most probable" seems suspect. There has to be given very secure reasons as to why the Arabian peninsula should be chosen besides being, maybe, a cool fad ;)
In fact, I can think of one reason Semitic must be further north. That reason is Proto-Indo-European strangely enough. It's been well known by IEists that numerals like *sweḱs "six" and *septm̥ "seven" can only be borrowings from Semitic *sidθu "six" and *sab`itum "seven" respectively. In IE, *septm̥ happens to be an unanalysable disyllabic stem with a curious shape (ie: the stem ends in a zero-grade syllabic resonant *- that can't be explained by IE morphology), whereas by realizing that it's a loanword, we can see that it's specifically derived from the Semitic counterpart *sab`itum, the mimated masculine form of the stem *sab`-. Since IE is dated to around 4000 BCE and yet demonstrates conclusively fossilized Semitic morphemes in these loanwords, this is one consideration suggesting that Proto-Semitic must surely have been further north than the Arabian peninsula at this time. --Glengordon01 04:32, 29 July 2006 (UTC)

Sorry, that should be *šidθu, not **sidθu --Glengordon01 04:35, 29 July 2006 (UTC)

I think I managed to trace the "source" of the Arabian peninsula myth/theory: Outline of a Comparative Grammar of the Semitic Languages, C. Brockelmann, 1908-1913. I hope you all realize how long ago 1913 was. On the other hand, we have The American Heritage® Dictionary of the English Language, Fourth Edition. 2000. Proto-Semitic Language and Culture: "The level of technology that the reconstructed Proto-Semitic vocabulary points to is that of the late Neolithic or early Chalcolithic." and "The words for many other agricultural products may provide clues as to the original homeland of the Semites [...]" such as for "figs", "pistachios", "garlic", "palm trees", "date honey", "almonds", "cumin", "cultivated grapes", "vines", "vineyards" and "wine". So between the IE loanwords and these quotes from the American Heritage Dictionary, the areas of Syria and Palestine have a better shot of being right. --Glengordon01 05:02, 29 July 2006 (UTC)

I'm not sure where you're getting your info from, but the idea that PIE numerals are Semitic borrowings is not at all mainstream. Benwing 05:29, 31 July 2006 (UTC)

the referene requested is Kienast (2001), passim. That's 5 years ago. Just because Brockelmann already suggested similar things in 1913 doesn't make the idea outdated. In fact, I believe it is so standard that it would be misleading to attribute it to Kienast in particular, and I would be interested in a reference for "Syria and Palestine" having "a better shot of being right". The numerals connection is completely spurious. dab () 07:01, 2 August 2006 (UTC)

It's not that much of a shocker. Just need to put things in context. Here's another book on IE-Semitic numeral connections and the neolithic: Semitic and Indo-European.
Regarding IE *septm̥, the fossilization of Semitic morphology is so absolutely clear, right down to the preservation of the numeral's masculine extension, *-at- (I mistakingly write *sab`itum instead of *sab`atum above, sorry). The word is unanalysable in IE despite few such roots, let alone disyllabic ones. Most linguists simply ignore the larger implications of this straight-forward borrowing, but questioning something so obvious is an impossible task here.
Regarding IE *sweḱs, that's a different matter. I empathize with those who think it's a hoax, but the connection is often dismissed based on linguistic ignorance. If Semitic *š- is labialized (as in English), IE speakers would naturally perceive it as *sw-. Perfectly natural. IE * can no longer be called a "palatal" phoneme due to simple phonological markedness (therefore IE * = /k/ but *k = /q/). Knowing this, we are merely comparing Semitic *-- with IE *-ḱs- /-ks-/ on the phonetic level.
Since IE lacks alveolar stops, velar * (pronounced as plain /k/) makes for the only, albeit crude, approximation of a Semitic alveolar stop *-d-. In this way, IE *-ks- causes automatic retraction of dental *s to an alveolar position. Close enough for an IE speaker. (Note this is hardly anywhere near as strange as the way Hawaiians pronounce "Merry Christmas" as "Mele Kalikimaka", replacing the foreign English sound /s/ with their nearest approximation /k/!). And naturally since IE had no sound like theta as far as we are aware, *s is a natural replacement. Hence *sweḱs is actually a quite expected reflex of the corresponding Semitic numeral in its feminine form.
I always find the gender opposition fossilized in IE's "6" and "7" to be seductively fascinating. Most likely, the numeral borrowings entered IE circa 5000 BCE or so during the crux of the neolithic, due to economic trade. These numerals may have had a deeper religious/numerological significance than we now appreciate, thus explaining the widespread borrowing of the same two numerals from Semitic into other language groups (Kartvelian, Vasconic (Basque)). The pattern is too curious to explain in any other way than this. --Glengordon01 05:52, 4 August 2006 (UTC)
Hm.
The case for "7" looks pretty convincing, but "6" is IMHO off.
· If PIE *k was [q], why didn't anyone fricate it away or turn it into a glottal stop or something? Isn't the AFAIK more-or-less standard interpretation – that *ḱ, *k, *kʷ formed a West-Caucasian-style triad /kʲ/, /k/, /kʷ/ – much, much easier?
· What makes you think you can tell if PIE *t, *d, *dʰ were alveolar or dental? And what makes you think it matters? Within German, the southern and central dialects have laminal denti-alveolar /d/ and /t/ like French or Russian, while the northern ones have apico-alveolar ones like English. It took me (an Austrian) a long time to figure out what the minute difference between the [southern] German and the English ones was. If I don't pay attention, I don't hear the difference at all. So, even if the PIE ones were apico-dental (like the Indic non-retroflexes are), these sound still much closer to laminal or apical alveolar stops than to velar (or palatalized velar) ones. In sum, why [ks] and not [ts]?
· The reason for why Hawaiian represents [s] with /k/ is actually pretty simple – that language has one phoneme that is pronounced [t] or [k] depending on the island. The step from [s] to [t] is not that big…
Even the case for "7" needs a bit more explanation. The grave accent you put into the Proto-Semitic form is a voiced pharyngeal fricative, [ʕ], right? What if PIE *h3 was a voiced velar, uvular or pharyngeal fricative (which seems to be the most common interpretation nowadays)? Why is it *septm̥ and not *sebh3tm̥?
(Incidentally, a *b would greatly bolster your hypothesis of a loan from something else into PIE, as opposed to, say, a loan in the other direction.)
Lastly, I'd like a few more comparisons. Is it really just Semitic, or is it Afro-Asiatic as a whole? Because if Afro-Asiatic, Indo-European, and Kartvelian all have those words, they should be added to the evidence for Nostratic – not loan, but inheritance. I need to check (…in the extremely meager literature I have…) if any Nostraticists have already tried to make that connection.
Incidentally, based on comparisons to other Nostratic languages, PIE *h3 is most easily interpreted as a voiced uvular fricative, and the palatalized velar stops are derived from velar stops followed by unrounded front vowels, while the labialized velar stops are derived from velar stops followed by rounded vowels; and *t and *ḱ, *k, *kʷ correspond to Proto-Afro-Asiatic and Kartvelian ejectives. This should help decide that question.
David Marjanović | david.marjanovic_at_gmx.at | 16:30 CEST | 2006/10/28

Oh, PS, about a "Syria/Palestine" location being likely. It's simply borne out from the numeral connection above. If we finally accept that these numerals have been borrowed, it logically follows by necessity that these borrowings were exceedingly early (neolithic). However, Semitic is dated to this very time. So in order to get Semitic numerals into a language spoken so far north, it's inevitable that Semitic must have been much further north than the Arabian peninsula at that time. Some try to place IE in Anatolia (Ivanov, Gamrelidze) but this just isn't accepted and eastern Europe is accepted likeliest. So what other conclusion can there be that takes stock of all facts? --Glengordon01 06:14, 4 August 2006 (UTC)

Proto-Afro-Asiatic: "Since the 1960s, however, it has been shown that PPNB developed in an unbroken sequence from the Natufian cultures north of Damascus, forging a link between Palestine, Mesopotamia and the Anatolian cultures of Catal Huyuk, and Halicar, with which it shares some similarity. It has been suggested that this northern part of the range was developing as Proto-Semitic." Comments? --Glengordon01 06:28, 4 August 2006 (UTC)

Personally I don't feel comfortable with associating such old cultures with languages. We have an unbroken line from Sumer to Akkad, and another from the Etruscans to the Romans, but the languages involved are very different in both cases. The present distribution of languages argues for the origin of both Afro-Asiatic in general and Semitic in particular somewhere around Ethiopia, and that's all evidence I know of. (Of course this doesn't rule out later Semitic-IE contacts.)
Also personally, I can't see why "this just isn't accepted". What are tree names doing in the Don basin? Why can't "wheel" and "horse" be early Wanderwörter? Some place in or east of Anatolia fits very well, not just the lexical evidence but also the consistent result that the Anatolian languages split off first and the loans from Caucasian (e. g. *abel "apple tree").
David Marjanović | david.marjanovic_at_gmx.at | 16:40 CEST | 2006/10/28

[edit] proto-semitic affricates

User User:Gilgamesh claims that Proto-Semitic fricatives were actually affricates.

This does not agree with e.g. [3]. Also, the change was made only to the bottom half, not the top half, so there is internal inconsistency. If we are to maintain this, we need references proving that this is the accepted view among Semiticists.

BTW the affricates may well be the accepted idea for Proto-Hebrew, but this is not the same as Proto-Semitic.

Benwing 05:38, 31 July 2006 (UTC)

User:Gilgamesh also changed the reconstructed Koranic equivalent of ʃ to a lateral, which is almost certainly wrong, so i reverted it. Benwing 05:31, 2 August 2006 (UTC)

[edit] use gifs pngs for ge'ez symbols?

the ge'ez symbols don't show up at all on my system (standard Windows XP), so i suspect most others don't see them, either. Benwing 07:21, 2 August 2006 (UTC)

MS Internet Explorer is simply not able to cope very well with multilingual texts: it doesn't change the font to deal with changes in orthography in the way that most other browsers do. The workaround we use is to plave Template:Unicode around everything that MSIE is likely to have a headache with, so that it forces it to change font. If that doesn't work, it means that you don't have Ge'ez-capable Unicode font installed on your machine. — Gareth Hughes 12:35, 2 August 2006 (UTC)
it's better to use a specific template, I guess {{lang}} in this case. {{lang|gez|ግዕዝ}} displays as ግዕዝ. If you still can't see it, you should install some Unicode font. The Ge'ez abugida is widely supported in fonts these days, and while I can see the use of using graphics for rare things like the Gothic or Italic alphabets or Linear B, uploading pngs for all Ge'ez syllabograms would be a terrible waste of effort. dab () 13:06, 2 August 2006 (UTC)

Does anyone know if Ge'ez "b" was always b or if it was ever β? I know that some words with b are pronounced β in modern Amharic (e.g. "Abeba" in Addis Abeba; not sure about Tigrinya). — ዮም | (Yom) | TalkcontribsEthiopia 13:51, 2 August 2006 (UTC)

ok thanks, i got the font from ftp.ethiopic.org and it works fine now. Benwing 00:38, 3 August 2006 (UTC)

[edit] Uvular Fricatives

Benwing - re: your edit "someone forgot the uvular fricatives":

The omission of uvulars was not an oversight - an older version of this page included them, but did not provide any citation. I found a reconstruction of the phonology of Proto-Semitic in a general-interest book, which I reproduced above verbatim, and included with minor modifications in the text. This source made no mention of a series of uvular fricatives; nor does the other link included in the references section. As such, if you wish to restore the uvular fricative series, I must insist that you provide a credible citation to substantiate their inclusion. --Peter Farago 14:37, 10 August 2006 (UTC)

If that is the case, did someone not forget the velar fricatives? — Gareth Hughes 14:50, 10 August 2006 (UTC)
Someone did. How embarrassing. --Peter Farago 16:16, 10 August 2006 (UTC)

[edit] Emphatic Labial

Some reconstructions posit the existance of a (probably marginal) phoneme 'emphatic P' which accounts for cognates in different languages where /b/ corresponds to /p/. I haven't added it in, but if anyone knows more about that they might want to lay out the evidence and/or anti-evidence.

[edit] "Semitic abjads"

I changed the see-also link text "Semitic abjads" to "History of the alphabet" because abjad is not a generally accepted term in this context, and because most people know what an alphabet is, but have never heard of an abjad. In any case, Semitic abjads redirects to History of the alphabet so there is nothing to be gained by using obscure terminology. The reason I have written this note is because the term "abjad" seems to be cropping up around wikipedia in ways that suggest it is the generally accepted term for an alphabet without vowels, which it is not. Ireneshusband 01:03, 5 November 2006 (UTC)

[edit] Sound changes section notes

"Biblical Hebrew as of the 3rd century BCE apparently still distinguished ġ and ḫ (based on transcriptions in the Septuagint)." -- Should this say that it distinguished ġ and ʻ? Or x and ḫ? (Or χ -- does that phoneme vary between velar and uvular in Modern Hebrew?) 24.159.255.29 03:30, 12 December 2006 (UTC)

[edit] Phoneme Chart and Title of Second Section

Is there a reason the phonemic inventory of PS is displayed the way it is? I'm not familiar with the orthographic traditions of the field, so I hope I'm not stepping on anybody's toes when I say this, but it seems to me that a more conventional layout such as this might be more effective:

   p        t        k         ʔ
            t'       k'
   b        d        g
       θ    s    ʃ   x    ħ    h
       θ'   s'
       ð    z        ɣ    ʕ
   m        n
   w        r    j
          l,ɬ,ɬ'

Of course, we'd have to label the rows and columns; but it seems to me, as an outsider to this field, that this layout makes the peculiarities of the phonemic inventory more apparent (for example, the lack of labial fricatives, the myriad alveolar liquids) than the current way it's set up. Moreover, [n] and [r] are not postalveolar fricatives, but the current chart (unintentionally) claims they are.

And on an unrelated note: Does anybody else think that a better title for the section "Sound changes between..." would be "Sound correspondences between..."? Again, I'm not an expert, but it seems that "sound changes" implies a diachronicity that the chart doesn't convey, in that there is very little information of the type, "PS phoneme X became Y in the East Semitic languages, as attested in Akkadian Z." (I'm guessing here that there is a PS branch called East Semitic, and that Akkadian is a member of that branch; I am, more than likely, wrong on at least one count, but I hope you get my point anyway.) Anyway, that's the sort of information I would expect to find in a section entitled "Sound changes," and I think renaming the section "Sound correspondences" resolves that inconsistency.

Froboyd 18:09, 9 February 2007 (UTC)