Talk:Proto-Semitic language

From Wikipedia, the free encyclopedia
Jump to: navigation, search
edit·history·watch·refresh Stock post message.svg To-do list for Proto-Semitic language:

Here are some tasks awaiting attention:
  • Article requests : standardize Semiticist orthography (especially between Semitic reflexes and Afro-Asiatic correspondences); emphatic p?

Emphatic Labial[edit]

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.—Preceding unsigned comment added by (talkcontribs) 25 August 2006


Dating proto-Semitic to "between 4,800 BP and 4,500 BP" can't be right. Akkadian is attested as early as that, and Akkadian is considerably changed from proto-Semitic. —Preceding unsigned comment added by (talk) 23:11, 27 May 2009 (UTC)

Assuming that "BP"="BC", I would think that you are mistaken about Akkadian being attested so early. For example, this site states that "Akkadian is first attested to in proper names in Sumerian texts (ca. 2800 BCE)". Mo-Al (talk) 01:07, 28 May 2009 (UTC)
Assuming that "BP" = "Before Present" that makes a lot of sense tho. --Trɔpʏliʊmblah 09:02, 5 June 2009 (UTC)
@Mo-Al: Assuming BP = BC? Why would you think that? I'm not looking at the version that was current when this section was added, but... 4800 BP ~= 2800 BCE, which is the date you quoted for Akkadian in Sumerian texts. Browsing the article as it is right now, it doesn't give a BP date, but neither does it give a hypothetical date at present (just the earliest attestations of 3000 BC and ~2400 BC, but neither of these are for PS). Is there a scholarly convergence on a hypothetical date window for PS? Ad hoc, 4800 BCE seems a bit early, and, as the author of this section points out, 2800 BP seems a little late. As a side-note post scriptum: appreciation for your contributions to the Hebrew phonology pages (I don't even bother double checking them when made, at this point, I assume they are good from past experience of your work). — al-Shimoni (talk) 13:38, 16 May 2011 (UTC)
Hmm, to be honest I don't know what I was thinking two years ago. It seems clear that BP = "before present", as you say, and I definitely would not believe a PS date later than 3000 BCE. If I have time I may look for some sources to clarify this issue. (And thank you for the approbation -- it's nice to know that my additions help :) )Mo-Al (talk) 22:16, 16 May 2011 (UTC)

Old Aramaic *ś[edit]

The article states that Old Aramaic *ś>s was written originally with < š > and then with < s >. Would it be more accurate to say that it was originally written with < ś >, given the niqqud? Mo-Al (talk) 01:51, 28 May 2009 (UTC)

The niqqud (meaning shin and sin dots) was only invented in the 6th-century A.D. or later, and was applied retroactively to Biblical Aramaic as part of the Hebrew Bible at that time. Old Aramaic inscriptions used no dots... AnonMoos (talk) 22:13, 20 December 2010 (UTC)


The homeland section has been expanding with opinions and hypotheses which are -I believe- beyond the scope of the article or wikipedia. I think a more compact section would be more friendly to the readers (compare the current section with the same section a year ago [1]). Besides, the current section is a bit messy and lengthy...maybe a new Homeland article would better contain the different hypotheses and present them in more detail and a more decent categorization (according to region for example). I would've gone through the effort hadn't it been a sensitive subject where even the strongest of opinions were merely speculative at best.--Xevorim (talk) 05:43, 5 October 2009 (UTC)

Please define your understanding of "the scope of the article or wikipedia" and how it relates here. Otherwise this is a very vague remark. Are you saying this article is now too long and needs to be split, or you saying the article now contains material which is un-encyclopedic?--Andrew Lancaster (talk) 18:33, 5 October 2009 (UTC)

I'm saying both actually. The section (not article) is too long and contains materials which are un-encyclopedic.--Xevorim (talk) 20:41, 5 October 2009 (UTC)

The article is not particularly long by any standard I am aware of in Wikipedia. Have a look around. Indeed splitting it might create two "stubs" which would certainly not be good form. And you do not say what is unencyclopedic. The basis of your remarks is still not clear. I am tempted to think that you are avoiding saying that the section in question simply needs to be improved? If this were the case then there would be work to be done, and simply appealing to rules and regulations as an excuse to hide the stuff under the rug would not be the most constructive approach. The question I am asking now is whether you honestly think that discussion of a proto semitic urheimat is something which should not appear in Wikipedia at all, after any amount of work, which you have kind of said, and which would be a pretty controversial opinion, or whether you just think this section is just poorly done as it currently stands.--Andrew Lancaster (talk) 20:51, 5 October 2009 (UTC)
The article claims Proto-Semites were responsible for the collpase of the Ghassulian culture 3,300 BCE. This is unlikely on a number of grounds. Firstly the appearance of Akkadian East Semitic language from about 3,300-3,100 BCE at Kish suggests a much longer period of development. Secondly the Ghassulian chalcoolithic seems to have been responsible for the system of Mediterranean agriculture associated with West Semitic, and there is no clear ethic change after the end of the Ghassulian to the Early Bronze Age. This view has been championed by Christopher Ehret and others.
Secondly Ghassulian itself developed out of what Juris Zahrins calls the the Circum Arabian Nomadic Pastoral Complex, which has alll of the characteristics of the Proto-Semitic. This culture itself seems to have appeared with the developments of the Minhata of Israel, itself which seems to have been a fusion between Sinaitic elements of the PPNB Neolithic, and Harifian hunter-gatherers, whose Outacha retouch microlithic technology shows clear connection with the Egyptian Fayyum. Check out World Archaeology, Volume 15, Issue 1, 1983 Special Issue: Transhumance and pastoralism, "The secondary exploitation of animals in the Old World" by Andrew Sherrat, pages 90-104. It suggests that early farming populations used livestock mainly for meat, and that other applications were explored as agriculturalists adapted to new conditions, especially in the semi‐arid zone. John D. Croft (talk) 17:04, 4 October 2013 (UTC)

The chapter about "Cradinal numbers" is wrong[edit]

Every "s" should be replaced by "š". (talk) 22:27, 17 October 2009 (UTC)

Chiastic concord?[edit]

Can anybody explain what "chiastic concord" might be? I am referring to the main article where, sometimes, nouns enumerated are masculine, are followed with numbers that are feminine? I don't exactly understand this. I think the main article would be improved if someone added some examples.

And in proto-semitic, there are only two genders? Are there any common gender nouns, like in Latin? Dexter Nextnumber (talk) 02:04, 15 January 2010 (UTC)

Latin has neuter gender; "common gender" as the name of a grammatical category is generally reserved for languages like Dutch and some of the Scandinavian languages, which have a neuter vs. non-neuter distinction, but no distinctively masculine or feminine grammatical gender categories.
And Semitic languages have only masculine and feminine (no neuter). "Chiastic concord" is a traditional 19th century way of saying that seemingly masculine-appearing numerals go with grammatically feminine nouns and seemingly feminine-appearing numerals go with grammatically masculine nouns. (See Chiasmus, Chiastic structure, etc.) -- AnonMoos (talk) 05:24, 15 January 2010 (UTC)

Sound system update[edit]

All recent literature admits to Faber's reconstruction. This article is outdated. I recommend the reconstruction by Huenergard in The Cambridge Encyclopedia of the World's Ancient Languages. This reconstruction is thorough and reflects the up-to-date consensus. —Preceding unsigned comment added by (talk) 21:37, 6 February 2010 (UTC)

The asterisked symbols have really acquired independent status, so that "*š" is very commonly understood as referring to a certain specific cross-Semitic cognate correspondence set. Even if it turns out that in actual proto-Semitic what is commonly called *š was pronounced phonetically as [s], I would not advise replacing "*š" by "*s" in the article, since that would have a potential for creating great confusion. It would be better to add a "phonetic reconstruction" column, or something similar... AnonMoos (talk) 23:28, 6 February 2010 (UTC)

I don't agree since many major works have already used the revised symbols, e.g. Hasselbach's study of Sargonic Akkadian. One would fear confusion most in Assyriology, since the situation of sibilants there is messed up, yet the new symbols have been used in Assyriology without a problem (actually the new symbolism does solve many problems). The revised symbols are also used in the reconstruction of PS in The Cambridge Encyclopedia of the World's Ancient Languages, which is the best reconstruction known to me so far. A good article should mention the traditional sound reconstruction and then go on to the new one and use it. Remember that the new sound system was introduced to clear confusions not to create them. —Preceding unsigned comment added by (talk) 12:34, 7 February 2010 (UTC)

The second table needs to be fixed, and the section needs some reorganization. I laid the path and somebody should continue. —Preceding unsigned comment added by (talk) 23:42, 9 February 2010 (UTC)
Is there any structural difference involved, or is it all phonetics? It seems we could just list the phonemes once (ie. take phonetics out of the table) and then discuss the realization in the dedicated sibilant section. --Trɔpʏliʊmblah 18:41, 11 February 2010 (UTC)
It's all phonetic differences. I think it would be better to remove that section, and discuss Faber's reconstruction in the paragraph below the original table. --Xevorim (talk) 17:14, 12 February 2010 (UTC)
Done. I hope didn't mess it up too bad. :) In particular I'm kinda winging it about the laterals; and does everyone still leave the interdentals as they are? I seem to recall once seeing something about reconstructing those as postalveolars.
Also, I couldn't find a good place to fit this in: "Because the reconstruction of Proto-Semitic was originally based primarily on the Arabic language,[1] the phonemic inventory of reconstructed Proto-Semitic is very similar to that of Arabic, with only one phoneme less in Arabic than in reconstructed Proto-Semitic."
Thing is, I'm not sure if this is making a point that PS needs more consonants than it currently has, or something about the phonetics. --Trɔpʏliʊmblah 21:25, 12 February 2010 (UTC)

Thanks. This looks better than the previous version. I'll try to fit that statement probably before the table.--Xevorim (talk) 20:14, 16 February 2010 (UTC)


Ugaritic is mentioned in the first table, but not in the next two. —Preceding unsigned comment added by (talk) 15:36, 18 August 2010 (UTC) Something is known of Ugaritic pronouns and vowels. —Preceding unsigned comment added by (talk) 10:15, 19 August 2010 (UTC)


The intro sentence about a constant pattern of migration out of the Arabian peninsula to the fertile crescent is an old early 20th century view. Mainstream view amonst archaelogists[citation needed] is that small groups moved into Arabia from the fertile crescent and indeed to this day Arabia remains sparesly populated. Kuratowski's Ghost (talk) 18:47, 11 September 2010 (UTC)

We are talking here only about the last 5000 thousand years; namely we're talking about such peoples as the Amorites, Arameans, Arabs etc. I have been reading a lot in history and archeology, and as far as I can tell it is still quiet a prevalent view that such peoples as the Amorties, Arameans, and Arabs came from the desert. What you say does not make sense. Arabic is spoken today from the Atlantic to the Indean Oceans, even though the Arabs were (and are still) few in number. I need to see citations supporting your claims. We know for sure that the Arabian nomads have been migrating from Arabia into the fertile crescent from at least 500 BC until now. This is 2500 years of proven migration, and there is no reason to assume that this wasn't happening before. I need to see a good reference saying that e.g. the AmorItes weren't Arabian nomads. As far as I know this is almost undisputable fact in archaelogy.HD86 (talk) 19:42, 11 September 2010 (UTC)
There is a difference between "Arabian peninsula" and "Arabia" in the classical sense of the Syrian Desert - Amorites, Arameans, Ancient Arabs are understood to have come from "Arabia" meaning the Syrian Desert, not the Arabian peninsula. And the nomadic people of the Syrian Desert and later the Arabian peninsula arose from the fertile crescent in the first place, offshoots of the Yarmukian, Halaf and related cultures. Kuratowski's Ghost (talk) 20:30, 11 September 2010 (UTC)
Doubt it -- you're leaving out the whole matter of south Arabia... AnonMoos (talk) 20:41, 11 September 2010 (UTC)
No I am not, ok we have evidence of Mousterian (Neanderthal) culture in the peninsula going way back, but when we get to modern human culture we have evidence of offshoots of the Levantine cultures entering the west and north of the peninsula and Mesopotamian plus Indus Valley cultures entering on the east of the peninsula. Kuratowski's Ghost (talk) 20:53, 11 September 2010 (UTC)
Arabia = Arabian peninsula. There is no difference. This is your first big mistake; and by admitting that the Semitic speakers came from the northern Arabian desert you admit that the section is not wrong. The question of where did the Semitic nomads arrive in the desert from is another question that is not related to the spread of the Semitic languages. We know for sure that Mesopotamia and the Levant were not Arabic until the Arab nomads started infiltrating them in the second half of the 1st millenium BCE (most of those Arabs are recorded to have came from Yemen). We also know for sure that Mesopotamia and the Levant were not Aramaic until the Aramean nomads started infiltrating them in the second half of the 2nd millenium BCE. We also know for sure that Mesopotamia and the Levant did not have any Amorite cities such as Babylon, Yamhad, Qatana, and Ugarit until the Amorite nomads started infiltrating them in the late 3rd millenium BCE. We also know for sure that Mesopotamia was not Akkadian until the first half of the 3rd millenium BCE. These are commonly accepted facts. Secondly, what you say is already mentioned in the section. The section starts by mentioning a prevalent view and then moves on to newer suggestions such as yours. The section is well written and provides a full understanding of how the Semites expanded in successive waves. You shouldn't disrupt that just because you don't agree with it. You can elaborate on your view down in the section. HD86 (talk) 06:07, 12 September 2010 (UTC)
Arabia is not always = Arabian peninsula, unless one includes the Syrian Desert as part of the Arabian peninsula. The section does basically cover what I am saying in the last few paragraphs. My objection is to the opening sentences which create the impression that there were possible cultures in the Arabian peninsula not having common origin with those of the fertile crescent that gave rise to the Semitic languages - that was the hypothesis in the early 20th century but was not supported by the archaelogical evidence. The evidence was that people spread into the peninsula and the deserts form the villages of the fertile crescent and then as their populations grew too big, moved back into the fertile crsecent. The southern part of the eastern arm of the crescent and the extension of people from it into the peninsula is not regarded as having produced the Semitic languages, the current views favour the rest of the crescent and the extension of people from it into the the north and west of the peninsula as the general area in which the Semitic languages started.
I would prefer the opening to be less dogmatic, the actual pattern of migration is into Arabia and then out again, not simply out of Arabia. I am not convinced that the simplistic view of Arabia being the Urheimat is still the majority view and would prefer a more neutral statement. I would move the mention of the later spread into Southern France etc etc to the very end of the section. I would move the recent views reviving Syria/Mesopotamia to immediately after the discussion of how the earlier idea of Mesopotamia was rejected after the dicovery of the Sumerian civilization. Kuratowski's Ghost (talk) 14:55, 12 September 2010 (UTC)

In order to do what you're saying, you need to prove to us with reliable citations that what you're saying is really a majority view, and that the earlier view is no longer so. You have not provided any sources supporting your claim. I have been reading and I know that many recent and reliable sources still adopt the traditional theory, which is by far the strongest.

We all seem to agree that migration from the Arabian desert into the Fertile Crescent did happen for at least the past 5000 years in Mesoptamia and the past 4000 years in the Levant; so the opening statement can't be called 'false' or 'dogmatic' because it simply states a well-known fact of history to build on it. Even if your theory were true, the opening statment would still be true; there is no conflict here. The point of disagreement between your theory and the traditional thoery regards what happened before 4000 years ago in the Levant. Your theory is that before that time the direction of migartion in the Levant was the opposite of the later attested direction. This is too far back in time; you can't consider the opening statement 'wrong' because you believe it is only true for the past 4000 years but not before.--HD86 (talk) 16:13, 13 September 2010 (UTC)

This is a good link on the origins of the Amorites and the Arameans. They both came from the middle of the desert. And this is an even better link that proves you wrong when you say that the desert in the pre-Bronze-Age could not support culture. These findings prove that the north Arabian desert did have flourishing cultures. I believe that similar discoveries were made in southeastern and eastern Arabia. Did you ever read about that?--HD86 (talk) 16:22, 13 September 2010 (UTC)

We all know that the Semitic languages weren't invented in Arabia— they reached there from Africa, perhaps through the Levant. But you need to realize that we're talking here about Proto-Semitic, not about pre-Proto-Semitic or Proto-Afro-Asiatic.--HD86 (talk) 16:28, 13 September 2010 (UTC)

The growing view is that Proto-Afro-Asiatic began in the middle east and not north africa. We need to be careful on the language here as either view, at this point, could be said to have equal ground. In re of earlier statements, I concur that "Arabia" does not always equal South Arabia (id est, the peninsula); there is a lot of evidence for a migration from-Northernly direction (even though there is some evidence for a from-Southernly direction). In all events, I stress walking a narrow line on the language to account for both possibilities, at this point. — al-Shimoni (talk) 13:48, 16 May 2011 (UTC)

Ignorant statements[edit]

Somebody added this statement:

Given the fact that Semitic is most closely related to the Ancient Egyptian language of all the Afro-Asiatic languages

This was obviously added by somebody who doesn't know what they're talking about. I read the whole book of Ehret and I know that most of his opinions represent only the opinions of some scholars. From my reading I can say that most scholars consider Semitic to be more closely related to Berber than Egyptian. This is even more logical on historical basis because we know that the Proto-Berbers were also nomadic like the Proto-Semites until fairly recently. I don't know when did this statement became a 'fact.' It can at best be called a hypothesis.--HD86 (talk) 07:13, 15 September 2010 (UTC)

Agreed. I don't know anyone who takes Ehret seriously. However, the idea that linking Semitic and Berber makes sense because they were both nomadic is reminiscent of Hamitic. Lifestyle has little to do with language, and for all we know the speakers of pre-Egyptian were also nomadic. — kwami (talk) 09:03, 15 September 2010 (UTC)

I said "until recently." Everybody were nomadic. Timing is what matters here. The Egyptians stopped being nomandic thousands of years before the Proto-Berbers and Proto-Semites. I also want to say that it is inappropriate to belittle a scholar because you don't agree with him.--HD86 (talk) 10:30, 15 September 2010 (UTC)

The other problem with the removed sentence is the style. It looks like Wikipedia taking a position. The field has different ideas and Wikipedia should just summarize them if it can.--Andrew Lancaster (talk) 12:41, 15 September 2010 (UTC)

š is a PHONE[edit]

I wrote here before that the symbols š, ś, and s in most Semitic literature represent PHONES NOT PHONEMES. I remember that I changed these symbols with the symbols s₁, s₂ and s₃, but somebody apparently restored the older symbols mistakenly believing that these represent phonemes, wich is not the case. I did mention examples from the literature before and I am not repeating.--HD86 (talk) 20:50, 1 October 2010 (UTC)

HD86, just to remind, our aims on WP are in some ways ambitious but in some ways not very ambitious at all. We don't try to be cutting edge, just to report what the mainstream says. So we tend to end up using whatever symbols and terms are "traditional" and then if appropriate we can mention newer ideas as well.--Andrew Lancaster (talk) 20:59, 1 October 2010 (UTC)
This is exactly what I've been trying to do here from the beginning. The symbol š as used by Semitists means the sound sh, not the letter sh. It represents a SOUND. When I see this symbol I think of the sound not the letter. This is the meaning of what I've been saying.--HD86 (talk) 03:08, 2 October 2010 (UTC)
See my comment of "23:28, 6 February 2010" above... AnonMoos (talk) 23:20, 1 October 2010 (UTC)
The comment you refer to is untrue. What is your evidence for this:
"š is very commonly understood as referring to a certain specific cross-Semitic cognate correspondence set"
I can mention several sources by famous scholars who use the symbol differently.--HD86 (talk) 03:08, 2 October 2010 (UTC)
I can hardly remember anybody using the symbol š in the way you say, except probably in the domain of Canaanite. Perhaps your experience is mainly limiterd to Northwest Semitic, but this page is about Common Semitic. It is true, people who work on Northwest Semitic tend to take š as representing a phoneme (which is probably because the value š=sh is so established in Northwest Semitic), but this is a limited usage. The sitaution is different in e.g. Assyriology.--HD86 (talk) 03:20, 2 October 2010 (UTC)
The "revisionist" reconstruction of proto-Semitic sibilants was somewhat obscure before the early 1980's, and only started making its way into standard reference works and grammars after that — while in the preceding century and more of Semitic comparative work it was generally assumed that Hebrew ש etc. would correspond to Proto-Semitic *š. It's nice if there have been recent advances, but this article should not be phrased in such a way as to make most works written more than 25 years ago (and some written within the last 25 years) semi-incomprehensible... AnonMoos (talk) 03:35, 2 October 2010 (UTC)
I think the aim should be to use whatever nomenclature people are most likely to come across off wiki. This does not mean newer theories can not be explained also. This is just a common problem on many WP articles, and normally not that difficult to handle.--Andrew Lancaster (talk) 07:21, 2 October 2010 (UTC)

You both seem to be still missing the point— the symbol š in Semitic literature denotes a sound, not a letter (except for the exception I talked about above). This is not something new but it has been so for at least the current generation. I tried to explain in multiple ways and I don't how else to say it. Please refer to literature dealing with the subject. As an example I can name John Huehnergard who uses the symbols just like I am saying.--HD86 (talk) 08:17, 2 October 2010 (UTC)

I would think that a fair reading of our replies shows that we understand your point, that the symbol is used to denote a sound? But you are exaggerating, and your own words show that you realize it, when you imply that this symbol is never used in the way under question. Concerning the "correspondence set" there are different nomenclatures, as you know, and so we are simply proposing that the most common one found in books people are likely to be looking at should be the primary one, and then it is possible to add in explanations about other nomenclatures, variant interpretations etc. Nomenclature problems are common on Wikipedia, but they are easy to solve as long as people remember that Wikipedia never aims to be cutting edge whenever doing so would create confusion. We are not writing for academic journals and people who already know everything. We are writing for general readers.--Andrew Lancaster (talk) 08:43, 2 October 2010 (UTC)

The notation using s₁, s₂ and s₃ is terribly confusing. What's worse, it's not even used consistently throughout the article. The worst part, however, is that nowhere in the article is it explicitly and clearly explained what those symbols are meant to stand for. It can be inferred if you work your way through the appropriate sections, but it is unnecessarily cumbersome. Why not simply stick to the traditional symbols and contrast them with their modern interpretation in the phonology section? Once, the modern values were given in the table, which I feel is no problem as long as the commentary makes it clear that these phonetic values are only hypothetical (as all the phonetic values given in the table are, eventually) and to a certain extent controversial, although (thanks to the strength and number of the arguments) more and more accepted in the expert community (correct me if that is wrong, I'm no Semiticist).

  • s₁ = *š = [s]
  • s₂ = *ś = [ɬ]
  • s₃ = *s = [ts]

Correct? --Florian Blaschke (talk) 21:15, 8 October 2010 (UTC)

I agree that whatever method is used, the correspondence between the TWO systems should be explained in the article. Consistency would also be nice.--Andrew Lancaster (talk) 06:42, 9 October 2010 (UTC)
I don't think the subscript-number notation is particularly confusing, certainly nowhere near the visual near-identicality of ’ and ʻ. (Perhaps that ought to be changed, too.) I've not seen anyone complain about eg. the entirely analogous way of transcribing IE laryngeals as h₁, h₂ and h₃. However, using subscripts for these but keeping the traditional notation for *ṣ, *ṣ́ is rather inconsistent. And OTOH having *ṣ₁, *ṣ₂ in their place would also be confusing (it suggests there's a *ṣ₃ missing). So I'm narrowly leaning on the traditional *s, *ś, *š. --Trɔpʏliʊmblah 16:56, 10 October 2010 (UTC)
I meant confusing for linguists, even. Linguists are used to the IE laryngeal notation because it is traditional, the notation using numbers for the Semitic sibilants is not traditional at all – it's an invention apparently cooked up by some Wikipedian, which in itself is reason enough to drop it. Traditional notations are always preferrable because that's what you are most likely to encounter in the literature, although common alternative notations should also be explained. What's worst, the table listing the correspondence sets (the prime source for establishing the definition of the symbols) uses the traditional symbols exclusively, not the notation with the numerical indices. --Florian Blaschke (talk) 19:29, 10 October 2010 (UTC)
The "s₁ s₂ s₃" notation is traditional in discussing the writing of ancient South Arabian languages with the South Semitic alphabet, but very untraditional (as far as I'm aware) in labelling proto-Semitic sound correspondences... AnonMoos (talk) 04:09, 11 October 2010 (UTC)
Interesting, thanks, I didn't know that. By the way, I find the indices harder to read than the traditional diacritics, as well, in fact while trying to transcribe the proto-Semitic reconstructions of the numbers I found out afterwards (by reading about Modern South Arabian and noticing a mismatch) that I had misread the s₂ as s₃, that's part of the reason why I said it is confusing. --Florian Blaschke (talk) 17:00, 14 October 2010 (UTC)

You know, however this battle was waged, it ended up in a state where half the article was in *s1 *s2 *s3, and half was in *š *ś *s, and no-one bothered to indicate in the article text what the correspondence between the two was! I don't think it's controversial to say that that state of affairs is leaps and bounds worse than either system used consistently. I've at least put in such an indication, but it's still the sort of thing that's liable to confuse, even readers who understand that phoneme symbols are conventional labels... 4pq1injbok (talk) 23:19, 8 November 2010 (UTC)

Fix the phones and phonemes[edit]

I am not a linguist (just a geek who would like to know how Hebrew was pronounced by those who wrote the Old Testament), but I will put my two cents in anyway. I gather that many experts now replace the notations *š *ś *s by *s₁ *s₂ *s₃. This is confusing because the subscripts give no clue to the pronunciation at all -- is this deliberate on the part of those who changed the notation? But it does emphasize that the original notation may have given the wrong idea.

The way the article is now written, one is constantly pressing PageUp to see how things are thought to have been pronounced. Confusing. Frustrating. Do not expect casual readers to memorize the meaning of the subscripts. An encyclopedic work is first and foremost a reference work -- one should be able to look something up to find out a detail quickly.

The solution to to revise the article, especially all the tables. Make two columns for Protosemitic, one for each notation, showing in the tables the IPA for both the "traditional" and the "new". This will emphasize the unsettled nature of present knowledge. It will make the tables more useful. Solo Owl (talk) 22:25, 14 November 2010 (UTC)

I don't know that it's true that "many experts now replace the notations *š *ś *s by *s₁ *s₂ *s₃" (and would tend to doubt it unless there has been a big change in the past few years). As far as I'm aware, "*s₁ *s₂ *s₃" was one editor's innovative personal attempt to extend the transcription conventions for ancient South Arabian inscriptions to also cover Proto-Semitic, in order to try to be neutral between "old" and "new" ways of representing Proto-Semitic sounds... AnonMoos (talk) 22:51, 14 November 2010 (UTC)
  • s₁ *s₂ *s₃ is one way of representing the letters. Another common way is simply to write them *s *ś *ts. I chose the first way because most people here are Hebrew-oriented and I just wanted to cut short the argument.--HD86 (talk) 23:36, 14 November 2010 (UTC)

Look, HD86. If you want to insist that this article be written entirely in *s₁ *s₂ *s₃ notation and not *š *ś *s notation, then go rewrite it that way. But, while the article is in the scrambled state it's in now, using both systems in the text, do not delete the equivalences between them. It makes the later parts impossible to interpret. The line

PS is reconstructed as containing six "s"-type sounds, for which there are multiple systems of notation in use: voiced *z, two emphatics *ṣ, *ṣ́, and three plain voiceless *s₁ = *š, *s₂ = *ś, *s₃ = *s.

[emphasis added] is about notation; there is no claim about sounds here. 4pq1injbok (talk) 01:17, 15 November 2010 (UTC)

Or, if you like, another sensible resolution of this would be to expand this notational issue to a whole paragraph, discussing the fact that the reinterpretation of the phonetic value of these correspondence sets has caused a multiplicity of notation, and then say (truthfully!) which ones we're using. We could put *s *ś *ts in as well. 4pq1injbok (talk) 01:33, 15 November 2010 (UTC)

I don't know if you have read what I wrote above or not, but this: (*s₁ = *š, *s₂ = *ś, *s₃ = *s) is WRONG. You don't make a page consistent by making it all wrong.
I have explained repeatedly and I feel tired repeating:
š represents a SOUND, NOT A LETTER.
This is a notational issue. When you write something, you must use the notation correctly.
In PS studies, scholars usually write in the following way (I hope this example makes what I am saying clear):
  • s₁ was s in PS.
  • s₁ was s or ś in Old Akkadian.
  • s₁ was s in Old Assyrian.
  • s₁ was š in Old Babylonian.
  • s₁ was š in Aramaic and Biblical Hebrew.
  • s₁ was s in Ethiopic.


Is it clear? š means the SOUND [ʃ], not the letter /ʃ/. It is not me who is saying so but this is how Semitic scholars use it. As for the article, I believe I already wrote 50% or so of the current article, and unfortuantly I don't have time to fix it or add anything to it. I have suggested before that somebody use the PS reconstruction by Huenergard in The Cambridge Encyclopedia of the World's Ancient Languages as a reference to improve this lousy article (I don't like everything in that reconstruction, but it reasonably reflects current majority views) but it looks like no body wants to do it.--HD86 (talk) 04:06, 15 November 2010 (UTC)
Anyway I am tired of arguing, so if you want to use š as a letter in this article go ahead and do it, but I doubt that you will find any source on PS using this notation system. It will be a Wikipedia-special notation system.--HD86 (talk) 04:13, 15 November 2010 (UTC)
I'm only irritated that you have time to revert my well-meaning insertion (and write the above screed) but not to carry your notation through to Proto-Semitic language#Reflexes of Proto-Semitic sounds in daughter languages and the remainder of the discussion in Proto-Semitic language#Fricatives. Right now, someone arriving at this article and looking at that reflexes table will see *š *ś *s in the left column, and the article doesn't say what they mean. That is unuseful; the article is not serving its purpose. And that is my concern. I'm not trying to allege that the values were [ʃ ɕ s] or whatever, far from it.
As for "a sound and not a letter", that's more sloganism than it is helpful. Certainly *š is a "letter" in the sense that it has been used to denote a proto-phoneme (i.e. correspondence set); it's sophistry to deny that. Now, maybe, every author who has ever used *š this way has done so because they posit that its value was [ʃ] -- that's what I assume you're claiming when you say it's a "sound".
I'd be surprised if that's the case. I come at this as an amateur and don't have an eidetic memory for sources, so I unfortunately can't back this up, but I do recall reading about this in an article which accepted [s ɬ ts] as the values of this triplet but still used *š *ś *s as the notations for conformance with earlier work (and, for what it's worth, I've also read writers who support the glottalic theory of PIE but still use *bh *dh *gh *gwh). Now, I can't actually find any article that does this for *š on a quick websearch, though e.g. Huehnergard, Comparative Semitic linguistics in Semitic linguistics: the state of the art at the turn of the twenty-first century is a near miss; he does mention "voiceless *ɬ (a fricative-lateral, also transliterated as *ś [...])" and says that "voiceless *s [...] [was] actually [ts]". So perhaps in this aspect, you're right. In which case, someone forgot to tell whoever wrote that table in section 4... 4pq1injbok (talk) 17:19, 15 November 2010 (UTC)
Aha, I see Tropylium has done exactly what I was after. Many thanks! 4pq1injbok (talk) 17:21, 15 November 2010 (UTC)
Yet it's still a work in progress. As long as this conversion is in progress, I too don't see what's the point of throwing a hissy fit over explaining the other system in use. *s₁ = *š, *s₂ = *ś, *s₃ = *s) explains how the systems used in this article (perhaps not in literature, but actually in this article) correspond.
The point about the symbols being not phonetical can be made more explicit if that's your (HD86's) beef here, but you may wish to note that the Wikipedia standard for phonetics is IPA. Thus, š is not a sound; [ʃ] is. Whether š means [s] or [ʃ] or /ʃ/ or /ʂ/ or whatever is completely context-specific. And this article, at the moment, uses it to mean "the Proto-Semitic phoneme also known as /s₁/". Until the point that it doesn't, this deserves to be noted. Even if this is at odds with existing literature.
If you have any comments on if the emphatics should be also differentiated with subscripts, that would be appreciated…
As for the original point about the subscripts being unintuitiv — I think that should be handled by actual phonetical transcription, like s₁ [ts], as I gather the revamp of the sibilant phonetics is an accepted consensus view. For that matter, I also think ʻ and ʹ are graphically far too similar and could benefit from being replaced by or additionally noted to be [ʕ] and [ʔ] more widely. --Trɔpʏliʊmblah 17:30, 15 November 2010 (UTC)
4pq1injbok and Tropylium, your personal thoughts and opinions mean nothing. Did either of you write a reference book on PS by any chance? This discussion has obviously become useless since you INSIST on using notation that is NOT used in any worthy reference. I can name at least five or six references supporting what I say, you could name nothing and still you are forcing your edits. User:Tropylium, FORCING your false edits is called edit warring.--HD86 (talk) 22:15, 15 November 2010 (UTC)
We aren't insisting on keeping using the other system indefinitely, nor are we asserting that it is some kind of an "official" system; we are only insisting that as long as the system has not been yet edited out from the article, it should be explained for the reader's benefit. Can you grasp this distinction? Until you do, this is indeed a useless argument. I have no objection if you wish to work towards abolishing the notation with š etc.
Also, if you wish to add some references specifically about the transcription, feel free to.
As far as policy goes, I think you'll find a single revert supported by discussion is not "edit warring" in any sense. --Trɔpʏliʊmblah 01:04, 16 November 2010 (UTC)

Emphatic consonants[edit]

On rereading the article I noticed this: Emphatic consonants “are generally reconstructed as glottalized in Proto-Semitic.[nb 1] Thus, *ṭ for example represents [tʼ].” According to the IPA tables in Wikipedia, [tʼ] is the symbol for ejective t, not glottalized t. Are they the same thing? If not, and [tʼ] is used due to the absence of a symbol for glottalized t, shouldn't the article say so? (I would recommend [tʔ] for clarity, perhaps noting that it isn't standard IPA.)

The footnote [nb 1] says, “This explains why there is no voicing distinction in the emphatic series (which wouldn't be necessary if the emphatics were pharyngealized).” This is unclear. Are you saying that it is biologically impossible to distinguish voiced and voiceless glottalized consonants? (If so, maybe this should be noted in the text rather than a footnote.) Are you contrasting this situation with pharyngealized consonants? Or is the same thing true for pharyngealized? In languages where emphatics are realized in ways other than glottalization, can the voiced/voiceless distinction occur? — Solo Owl (talk) 00:55, 17 December 2010 (UTC)

The modern reconstruction (when phonetically precise) is in fact as ejectives, and "glottalized" is something of a loose and possibly inexact synonym. Voiced ejectives are presumably not "biologically impossible", but they are typologically a lot less common than voiceless ejectives. And there's no problem with having both voiced and voiceless velarized/pharyngealized consonants -- many modern Arabic dialects have both. AnonMoos (talk) 03:00, 17 December 2010 (UTC)

Thanks for clearing this up. The footnote should be entirely rewritten to reflect your comments. — Solo Owl (talk) 02:23, 18 December 2010 (UTC)

The Sibilants in the Fricative Section[edit]

A number of issues.
• The section title I think should be renamed "Sibilants" as the the discussion here is entirely about the sibilant sounds rather than about the fricatives in general.
• The subscript notation system is rather confusing, even for someone (like myself) who is well versed in the semitic sibilant discussion. Additionally, where subscript (and superscript) notation is used in scholarly work, the systems are not consistent between scholars (particularly when their language focus is different from the next scholar). For example, in Ugaritic studies, for a while, it was common to use s1 for S, s2 for ś, and š for š (this has generally changed, now). Generally, it has been adopted to use š, ś, and s in Proto-Semitic discussion, even when the discussion is about how these were originally pronounced.
• The statement that the current view is that s and š are affricative [ts] and [s] looks faulty as of the last time I checked (fairly recently) this theory had been generally regarded as weak, although still being held by a minority of scholars. There is no citation for this statement, and the remainder of the section is mostly devoted to supporting this view (so much so that a minority of the majority have often been downright condescending to the proponents of the affricative proponents). A number of the supporting statements have, in the past, been shown to be weak, such as the statement that in Akkadian a T sound immediately followed by a sibilant becomes a geminate of that sibilant would be better explained is the sibilant was affricative. However, in linguistics, the T sound is known as a particularly weak phone, primarily in word terminal position (anyone who has seen French should know this already, also look at the feminine suffixes in the Semitic languages), secondarily in syllable terminal position followed by a sibilant (it has a tendency to geminate the sibilant, an English example is saying "cat city" quickly which then begins to sound like "Cassidy", this is the very action being weakly used to support the affricative hypothesis), tertiarily in other positions where it does not begin a word (Cockney "buʔʔer" for "butter" is a famous English example). Several of the other support statements of being affricative have been shown to be rather weak (or faulty), which is why most folks in this field don't generally support it.
I would also like to note that the personal pronoun table and the cardinal number table later in the article should have š rather than s in many of those words ("šuʔa"/"šu", et cætera). — al-Shimoni (talk) 19:20, 20 December 2010 (UTC)

"*s₁ *s₂ *s₃" does have a standard meaning in transcribing ancient South Arabian inscriptions, but you're right that it's not very usual for Proto-Semitic. And the hypothesis that the sibilants *s and *z were affricates is actually fairly respectable (a kind of extension of the emphatic-ejective hypothesis which dictates that if *ş was ejective, then it was almost certainly an affricate) — but I agree with you in doubting whether it's so dominant that the traditional reconstruction notation should be discarded (see my comments of October 2 and February 6 above)... AnonMoos (talk) 22:06, 20 December 2010 (UTC)
We're using the Old South Arabian notation? Eek! Does this mean they have finally settled on the sibilant correspondences? I once referred to OSA while researching something with PS, but discovered that (at that time) they had no real clue which of their sibilant letters corresponded to which letters of any other Semitic language (let alone how they were pronounced in OSA). They were saying that one letter could be either of 2 out of 3 sibilants, while the next was also possibly either of 2 out of 3 (which included one of the possibilities of the first), while the final letter they thought could be any of the 3 sibilants. (In simpler words: they had no F'ing clue). If they have figured it out by now, this would be great as I could continue along that line of investigation (instead of having given up on that option soon after discovering the above situation). — al-Shimoni (talk) 09:50, 21 December 2010 (UTC)
who uses this stupid notation? what books? WPians are told to cite refs (Personal attack removed) (talk) 16:50, 21 December 2010 (UTC)
Which system are you referring to? If you are talking about the s1 type notation currently used in this article, then it is used by a few books on Old South Arabian (there aren't many as there isn't a whole lot known about the language). A related system which differentiated two of the letters in Ugaritic language (which used a superscript number for the ś equivalent letter as an exception to the diacritical system) was used in older books on the language, but newer books and studies now use ś instead. If you are referring to the diacritical system that uses š, ś, et cætera, this is used in the overwhelming majority of books and journals in reference to Akkadian, Sumerian (which is not a Semitic language, just for clarification), Ugaritic, Mari, Amorite, North Arabian (which is extremely limited), et alli. For books on the Hebrew and Arabic languages, it is rather common for these to either mix that system with the the Hebrew/Arabic letters, or to just use Arabic and Hebrew (it rather depends on what exactly is being discussed). Frequently, whenever any of these papers/books which use the š, ś (etc.) are discussing exactly how a particular letter may have been pronounced, most modern works use IPA for the pronunciation to explain the diacritical notation. In effect, at least in these works, it makes the diacritical system (ie: š, ś, ḏ etc.) act as more of a reference to the lineage of letters from PS, not necessarily as an exact reference to pronunciation (ie, both Arabic shin and Hebrew sin are of the same PS lineage, referenced as ś in the diacritical system, however, in practice, in Arabic ś is modernly pronounced [ʃ] while Hebrew ś is pronounced [s]). An Arabist reading a work on PS typically knows the ś is his š and it is pronounced [ʃ]. This might not be clear to a newbie, though.
    Actual references that use the diacritical system is huge... A sampling (old and new) include the BDB (old enough that I think it no longer falls under copyright), and works such as "A Grammar of Akkadian" by John Huehnergard, "A Grammar of Ugaritic Language" by Daniel Sivan, "Amorite Personal Names in the Mari Texts" by Herbert Bardwell Huffmon, "Comparitive Lexicon of Ugaritic and Canaanite" by Issam K. H. Halayqa. Plus, just about any journal on Semitic linguistic or Semitic studies in general will use the diacritical system. IPA is rarely ever used for general transcription, pretty much, as said above, only used when specifically speaking about the pronunciation of a letter or a group of letters in certain contexts. Likewise, most Semitics works (at least for the past few decades) rarely ever use a subscript system like the one currently used by this Wiki article (Old South Arabian centered works may be the only exception). Any Semitics student/scholar/amateur would find the use of the subscript system in this particular wiki article a bit confusing. Additionally, most of the other wiki articles on Semitic languages use the diacritic system, which may be minimally augmented by IPA). Apologies for the excessive length in this reply.
al-Shimoni (talk) 06:20, 22 December 2010 (UTC) original vers: 06:14, 22 December 2010 (UTC)
it's still hilarious that WP is stuck in analysis paralysis like geeks. decide on an action. don't yap on. seriously. lame. (talk) 07:28, 22 December 2010 (UTC)
Well, since there seems to be a consensus including AnonMoos, al-Shimoni and myself, I've gone ahead and edited this, restoring the conventional symbols while noting their variant interpretations. I haven't added any sources, though; I hope someone more laborious will take care of that.--Anonymous44 (talk) 16:45, 11 April 2011 (UTC)
Speaking of hilarious: OMG! Is the IP suggesting that Wikipedia (especially articles about obscure, academic and highly technical topics such as language reconstruction) is edited by cool people?! --Florian Blaschke (talk) 18:26, 25 January 2014 (UTC)

Link to Wiktionary at the end points to brokenness[edit]

Can someone mend it?CecilWard (talk) 23:28, 4 April 2011 (UTC)

Use of ɬ[edit]

I think I understand that the PS sibilant reconstructions are in the process of being revised, but why list /ɬ/ and /ɬ'/ as the pronunciations for Ge'ez ሠ ፀ? This makes this page inconsistent with the Ge'ez language page. G.broadwell (talk) 09:38, 15 May 2011 (UTC)

It was, in fact, originally pronounced that way in Ge'ez. Mo-Al (talk) 16:28, 15 May 2011 (UTC)

I'm happy to accept that this is true, but it's not consistent with the sources I have for Ge'ez. Could you provide a reference? G.broadwell (talk) 15:29, 17 May 2011 (UTC)

Sure. Take a look at Gragg (1997) in The Semitic Languages, edited by Robert Hetzron. On p. 243-244 he says that ሠ ፀ were originally pronounced as different than ሰ ጸ, and later merged. So although there isn't a good way to be sure of their exact phonetic values in Ge'ez, it's known that they were distinct. (I was a little hasty in saying that they "were" pronounced that way -- I think the evidence is conjectural.) I'm also not sure how this differs from the Ge'ez language article, which also uses /ɬ ɬʼ/. Mo-Al (talk) 20:28, 17 May 2011 (UTC)

Thanks! I added the grammatical discussion at Ge'ez language, but now I see that the transcriptions in my source are not consistent with the chart of consonant phonemes on the same page. Will revise! G.broadwell (talk) 23:25, 30 May 2011 (UTC)

nouns in proto Semitic that seemingly make an African origin for the language impossible – ice, oak,[edit]

Oak are found in north Africa, and I presume that is not through planting by humans? OK, if a great source mentions this maybe we keep it but this sentence is not sourced.--Andrew Lancaster (talk) 13:03, 22 September 2011 (UTC)

There's cold weather in Ethiopia (which is more relevant to the Proto-Semitic Urheimat problem than Algeria or whatever). AnonMoos (talk) 22:50, 22 September 2011 (UTC)
OTOH, it might be true that experts have really argued this way. Would be much more comfortable with this sentence if we had a clear source, and we could say "according to...".--Andrew Lancaster (talk) 09:15, 23 September 2011 (UTC)

Urheimat speculations in lead paragraph[edit]

It's nice to try to bulk up the lead, but the detailed and inconclusive Urheimat speculations really should go somewhere else in the article... AnonMoos (talk) 22:53, 22 September 2011 (UTC)

In fact there is a sub-section for it I think? OTOH, mentioning something about it might be ok, just not too much detail?--Andrew Lancaster (talk) 09:15, 23 September 2011 (UTC)
Well, the lede should have a basic summary of what the topic is, or is about, so mention of it would be ok (one should not get into the "nitty-gritty" in the lede, though). However, the amount of attention to its Urheimat was a little out of proportion to the rest, so I reëdited it, condensing the details more, and taking out some others. When I replaced the lede yesterday, it was in response to your (Andrew's) next-earlier edit; I took a paragraph from the Semitic article and made a few quick edits including trying to incorporate the sentence that comprised the previous lede. I didn't look too closely at the balance(proportions) of the information, though. :/ — al-Shimoni (talk) 20:05, 23 September 2011 (UTC)

Arabian Urheimat revisited[edit]

The common view currently is not that Arabia was the homeland of the Semites. The sources given go back to the early 20th century are incomparable with modern ones advanced by Edward Lipinski (Semitic Languages: Outline of a Comparative Grammar p.41-47) for example which asserts that they emigrated to Arabia and Africa via Syria and Palestine.--Rafy talk 20:55, 13 December 2011 (UTC)

Maybe a silly question, but do we need to propose that there is a consensus at all? Maybe we should just list the main theories as in Afroasiatic Urheimat. It is always very difficult to agree about consensuses in some types of subject.--Andrew Lancaster (talk) 16:20, 29 December 2011 (UTC)
The Afroasiatic Urheimat is not the same as the Proto-Semitic Urheimat, and several candidates for the former would not be very relevant to the latter. Traditionally (50 years ago or so) I think Northeastern Africa / Southern Arabia was somewhat widely viewed as the leading candidate for the general region of the Proto-Semitic Urheimat. But I don't know what the 21st century consensus is, or whether a 21st century consensus even really exists... AnonMoos (talk) 17:07, 29 December 2011 (UTC)
Agreed, my impression is it is on the level of individual scholars' theories and no firm consensus. --JWB (talk) 18:49, 29 December 2011 (UTC)
I just came across a good article which also mentions Early Semites reaching the middle east up the Nile ans settling in the fertile crescent. The article is very intriguing when it shows that Akkadian made contact with pre-Semitic peoples from the Arabian peninsula. --Rafy talk 22:18, 17 January 2012 (UTC)
But is this relevant to this article, or maybe better discussed on another article such as Proto Semitic?--Andrew Lancaster (talk) 11:09, 19 January 2012 (UTC) Oops. I did not realize which article I was on!--Andrew Lancaster (talk) 11:10, 19 January 2012 (UTC)
Since no one seems to mind I will rewrite the origin from scratch based on modern sources.--Kathovo talk 15:40, 30 June 2013 (UTC)

Date Eras[edit]

Currently, the era for the dates in the article are not consistent with each other. I would suggest using the BCE system for its neutrality, especially considering the largest number of people who use Semitic languages today are not christians, as well as most scholarly books and articles for the past couple decades use the CE/BCE system. However, it should be noted that the era first introduced into the article was not the CE/BCE system (it was that other one), and so the default should be that other system unless consensus can be made to switch to the CE/BCE system (see WP:Era). Can consensus be made to go with BCE? If not, someone needs to change it to the christian system (it won't be me). — al-Shimoni (talk) 18:20, 22 April 2012 (UTC)


What kind of brilliant idea was moving all the info about Proto-Semitic phonology into the already bloated article on Semitic languages? A description of Proto-Semitic is something one expects to find in the article on Proto-Semitic. -- (talk) 21:40, 8 November 2012 (UTC)

Levelly, I agree. It is easy to check in the history that this was Benwing's doing. I think I'll ask him. 4pq1injbok (talk) 07:35, 9 November 2012 (UTC)
I just saw this. The problem was that there was an enormous amount of redundancy between the two articles, which was rapidly leading to "bit rot" as the two copies gradually diverged. Probably 80% of the stuff in Proto-Semitic language duplicated stuff in Semitic languages, including the vast majority of the phonology. The Semitic languages article actually increased very little in size. I'm not averse to moving the stuff back to the Proto-Semitic article, but I absolutely don't want to see a lot of duplication, and the problem is that a lot of the phonology does seem to belong in the Semitic languages article. I'm welcome to suggestions of how to split things; sorry if I don't respond too often, though, as I'm busy trying to get a PhD thesis done! Benwing (talk) 05:00, 14 November 2012 (UTC)
No sweat, I'm not fast either. Anyway, my suggestion of how to split things would be to put section "Phonology" of Semitic_languages up through the end of subsection "Fricatives" here, and leave the diachronic outcomes there. 4pq1injbok (talk) 22:25, 14 November 2012 (UTC)
This sounds reasonable. Some sections lifted from here could well use merging with certain sections previously in the main Semitic languages article, like the vocab section or the example words table on fricativ correspondences. --Trɔpʏliʊmblah 23:53, 15 November 2012 (UTC)
There, how's that for a first pass? 4pq1injbok (talk) 02:55, 26 November 2012 (UTC)

BC or BCE[edit]

The article currently starts off with BCE in the wp:lead then changes to BC. I would like to change it to consistent usage of BCE. Any comments? Editor2020 (talk) 18:46, 30 June 2013 (UTC)

I usually use BC/AD but since I believe my contribution created inconsistency I don't mind using the common era dating.--Kathovo talk 11:57, 1 July 2013 (UTC)
Actually the article's original era setting is the BC convention which has been partly changed by editors introducing BCE or some overtyping BC with BCE so I have reversed those to make consistent use of BC as required by WP:ERA.--CouncilConnect (talk) 10:34, 5 May 2017 (UTC)


Much of the excerpt relies on a single source (Lipinsky 2001), I have read the referenced pages and there are several caveats:

1. The terminology is incorrect and anachronistic, often in blatant contradiction with the works mainstream scholarship (for instance, the excerpt speaks of Semitic speakers living in the Neolithic Subpluvial in the 5th millenium BC [in accordance with Lipinsky's claims] while most experts see Proto-Semitic breaking up c. 3500 BC [per Kitchen et al 2009 for instance] so the correct label for the aforementionned "Semitic speakers" would be Pre-Proto-Semitic or, more accurately, "Boreafrasians" [Ehret 1995]).

2. The Afroasiatic urheimat's location is still highly contentious, though the African (especially Red Sea) hypothesis indeed seems to enjoy much support in academic circles, in such regards the link between Niger-Congo and Afroasiatic has been firmly established... Not by Lipinsky but by Dan Dediu & Stephen C. Levinson 2012 (Abstract Profiles of Structural Stability Point to Universal Tendencies, Family-Specific Factors, and Ancient Connections between Languages). So once more, there's an obvious amount of bias towards a single source.

3. Elamite as Afroasiatic is an extreme fringe theory, not even worthy of mention as it is rejected by nearly all scholars of Afroasiatic.

4. There are several elements which make an African origin for Semitic highly unlikely, among them the presence within Proto-Semitic vocabulary of petroleum-derived products, oak trees, ice and the horse (Militarev 2009)... Without even mentionning the fact that PS vocabulary itself points to a typical Southwest Asian Bronze Age community, not a foraging group which lived in the Sahara during the Neolithic Subpluvial 7000 years ago.

5. Lipinsky's internal classification of Semitic is completely out of touch with mainstream scholarship (Hetzron, Bender & Häberl come to mind).

Last but not least, Juris Zarins' Circum-Arabian Nomadic Pastoral Technocomplex tends to fit nicely with the data and expansion models proposed to have led to the development of Pre-Proto-Semitic into Proto-Semitic itself (let it be from an archeological, linguistic or genetic POV with Chiaroni et al's 2009 study providing support to Zarins' theory).

I hope all these elements will be taken into account, since the current version is heavily biased towards a single author and makes the whole excerpt look kind of Afrocentric. (talk) 01:31, 3 August 2013 (UTC)

I have rewritten the Homeland section from scratch since it was composed of incoherent trivia mostly from classical Greek and early 20th-century works.
I read both Lipinsk's and Blench's and combined their theories since they more or less corroborated the basic idea. I currently don't have access to those sources. You or anyone else is more than welcomed in introducing other theories with proper citations.--Kathovo talk 20:18, 13 August 2013 (UTC)
I don't really see the problem with Lipinski. Both Lipinski and Blench accept that the homeland of Proto-Semitic as such was in the Levant, in the 4th millennium BC. This is in full agreement with the cited study from 2009.
The African connection comes in only with respect to Pre-Proto-Semitic. Lipinski argues that Berber is the closest relative of Semitic. This is not incompatible with the "Northern Afro-Asiatic" view espoused by Blench, and no more controversial than alternative views such as that Egyptian was the closest branch. The subclassification of Afro-Asiatic is highly controversial in general and there is no consensus that Egyptian is closest to Semitic. In fact, according to what is stated in a section above, Berber is now believed by most scholars to be the closest branch, not Egyptian. So Lipinski is far from isolated on this point and his scenario makes some sense.
The Afroasiatic Urheimat is generally thought to be in Africa. The alternative view, that it lay in the Middle East, while formerly popular, and among Semitic- and Egyptian-oriented scholars perhaps still favoured, is apparently a minority view now, Militarev being its main proponent. I don't know Juris Zarins' point of view, but the way his name is mentioned in the Urheimat article leads me to suspect he was not opposed to the African urheimat. You too seem to accept it. If Proto-Semitic was spoken in the Levant in 3500 BC, Pre-Proto-Semitic a full millennium earlier (or more) can very well have been spoken in North Africa, even by foragers or (more likely perhaps, since the article explicitly speaks of a "vibrant Neolithic culture" in the Sahara) pastoralists. (Calling these people "Semitic speakers" is a bit imprecise but allowable, since precise terminology can be quite cumbersome.) You seem to construe a contradiction where none exists. Just think of the Persians: pastoralists in the 2nd or even early 1st millennium BC, advanced civilisation 1000 years later, which is also reflected in the Proto-Iranian vs. Old Persian reconstructed lexicon. There is the distinct possibility that the Pre-Proto-Semitic speakers infiltrated a more advanced civilisation, although in the 5th/4th millennium BC none of the agricultural/pastoral cultures was particularly advanced. --Florian Blaschke (talk) 22:18, 25 January 2014 (UTC)


As for the section Dating, there are the following problems:

Researchers in Egypt also claim to have discovered Canaanite snake spells that "date from between 3000 and 2400 BC".

This appears to be a contradiction with Canaanite shift, which dates Proto-Canaanite implicitly to the 2nd millennium BC. In the 3rd millennium BC, the dialects in question could be classified only as Northwest Semitic or even Central Semitic, I think.

The specific appearance of the donkey (an African animal) in Proto-Semitic but total absence of any reference to wheeled vehicles rather narrowly dates Proto-Semitic to between 4,800 BC and 4,500 BC.

I have added this part back because the reasoning appears valid; only the dates contradict Donkey and Wheel. The dates seem to be pretty much exactly one thousand years too old, which cannot be explained by a simple confusion of BP with BC. It is also worth pointing out that the according to de:Afrikanischer Esel, the African wild ass was formerly present in Mesopotamia as well. However, the animal was apparently only domesticated ca. 4000 BC in Egypt. If the Proto-Semitic word in question can indeed be assumed to have referred to the domesticated animal (the Asian wild ass has never been successfully domesticated), this would – along with the wheel argument – be excellent support for the consensus dating 4000–3500 BC. --Florian Blaschke (talk) 01:12, 26 January 2014 (UTC)

Ah, see the section above. I forgot that it originally did say BP – I changed it myself! However, the problem remains that BP gives dates that are too late and BC gives dates that are too early. While a century off (an uncertainty of about ±2% in this case) is acceptable, a millennium off (an uncertainty of about ±20%) is pretty extreme in this case. At least the datings for evidence of wheeled vehicles are not that uncertain.
Should the statement be removed? Strictly speaking, it could be classified as OR/synth, but I find it a very trivial case. Of course, it would be best to find a published argument to this effect. --Florian Blaschke (talk) 02:02, 26 January 2014 (UTC)

Out of the Arabian Peninsula hypothesis[edit]

I was reading this article and interested to know more about the Urheimat of Proto-Semitic. The introduction said that Arabian Peninsula Urheimat hypothesis has been abandoned, and that's it. Seems to be quite dismissive of the whole theory. The Arabian Peninsula Urheimat theory was actually quite popular and was the predominant view, so I think its worthy of more elaboration. After all, there's always the possibility of revisiting this theory as our understanding of climate change in the peninsula improves. I think it's quite un-encyclopaedic to leave out the theory outright, so I suggest a special subsection for the Arabian Peninsula Urheimat hypothesis, on par with the sections for Levant and Africa Urheimat Hypothesis. The current status of this hypothesis can be elaborated more in its new, own section. (talk) 06:50, 10 July 2015 (UTC)

Check the genetic[edit]

No african semitic genetic in pre-pottery/neolithic/bronzeage syrian. That is pre-european, pre-dravidian and pre-sumer/elamite/akkad. Semitic word-order is different from sumer, akkad, ebla, indo-iran, proto-greek/minoic and hittitic, but it is near to egypt and south-arabian. and here is bronzeage egypt genetic. the homeland is not in canaan, libanon, syrian. — Preceding unsigned comment added by 2003:46:D46:AAFC:80EF:D9E4:7542:FBC (talk) 22:20, 20 August 2015 (UTC)

Hypothesised vocabulary; & tri-consonantal roots ?[edit]

It seems very odd that while this article gives acres of space to how PS might have sounded, or where it might have originated, there seems to be next to nothing about vocabulary, or even word-structure.

In particular, one of the most striking things for somebody coming from a Western European language to, say, Biblical Hebrew is the almost universal construction of the vocabulary on triconsonantal roots -- to a degree that can almost sometimes makes the language seem artificially constructed. (No of course not really, but that is how it can sometimes seem, such is the degree of regularity of this).

It would be useful to discuss how far back this pattern goes. Is the vocabulary of proto-Semitic also considered to be built over a similar pattern of roots then inflected in different ways to give different meanings and different parts of speech? Are there particular characteristics that these proto-Semitic roots may have that should be presented here? (eg the Semitic root article, in its (rather limited) section History, suggests that some presumably very old roots, denoting Stone Age materials, may quite often be biconsonantal; whereas materials discovered during the Neolithic apparently are exclusively triconsonantal). It would be interesting to know more about this -- for example, how extensive a corpus are we talking about?

More generally, it would be useful to know how wide a proto-Semitic vocabulary has been hypothesised; plus, for example, what proportion of Akkadian or Ugaritic roots may go back to proto-Semitic; or may have apparent cognates in classical Hebrew or Arabic -- and how work in this area has developed.

Wikitionary gives an interesting handful of examples at wikt:Appendix:List_of_Proto-Semitic_stems (though I'm not really sure why that is at Wiktionary -- it looks more like encyclopedic than dictionary content to me). But a wikt page of some sporadic examples is no substitute for a more systematic and thorough overview of what vocabulary of common ancestral roots has been hypothesised.

Besides, that reconstruction and its extent is surely also fundamental, before one can start making pronouncements about sound-shifts etc -- so it really is something that ought to be presented and evaluated here. Jheald (talk) 20:58, 23 June 2017 (UTC)

  1. ^ Versteegh, Kees (2001) The Arabic language p.13