Talk:Afroasiatic languages/Archive 2

2 maps
No problem with either of the maps, but should there really be two which essentially tell the same story?--Andrew Lancaster (talk) 01:52, 4 August 2009 (UTC)
 * Funny, I just changed it to one as you have typed this. The more detailed map is included.  The other one was pointless. Azalea pomp (talk) 02:34, 4 August 2009 (UTC)

structure problem
Currently sub grouping proposals are discussed as a SUB section of the history of discussion on the subject rather than the other way around, and the current linguistic arguments for this are handled in a new separate section. Shouldn't this be reversed a little? This also perhaps gives a way to handle mentioning origins theories from a purely linguistic perspective, which was the apparent consensus approach discussed recently.--Andrew Lancaster (talk) 01:54, 4 August 2009 (UTC)
 * Origins theories from a linguistic perspective will be impossible due to the lack of any consensus by scholars. Azalea pomp (talk) 02:50, 4 August 2009 (UTC)
 * Perhaps, but the article structure seems to need a little tweaking to me.--Andrew Lancaster (talk) 23:04, 8 August 2009 (UTC)

Afroasiatic or Afro-Asiatic?
Both show up in web search. My impression is that Afroasiatic is slightly more common in the non-Wikipedia results.

The article states that the original term was Afroasiatic. Joseph Greenberg's final book Genetic Linguistics uses Afroasiatic. --JWB (talk) 15:47, 12 August 2009 (UTC)


 * Heine & Nurse also use Afroasiatic, as does Hayward in that volume. kwami (talk) 19:20, 12 August 2009 (UTC)
 * I've moved the article accordingly. --JWB (talk) 22:20, 17 August 2009 (UTC)

More data
External Links section of the article:
 * Afroasiatic: Theil, NACAL, Ratcliffe
 * Afro-Asiatic: Ethnologue, Blench (but rest of Blench's site uses Afroasiatic), Multitree (but references within Multitree mostly use Afroasiatic)
 * Afrasian: Militarev

LINGUIST list archive search:
 * Afroasiatic: 181
 * Afro-Asiatic: 140

Google search: Results are mixed and top 10 "Afro-Asiatic languages" results seem to be either Wikipedia mirrors, or actually use Afroasiatic.
 * "Afro-Asiatic languages": 83,200
 * "Afroasiatic languages": 85,500

My guess is that Afro-Asiatic has been popularized recently by Wikipedia and/or Ethnologue. --JWB (talk) 22:31, 17 August 2009 (UTC)

Number of speakers of Afroasiatic, Semitic and Arabic languages
The numbers are contradictory if you compare this article with the article on Semitic languages. Is there somebody here who can help? Otherwise, people could get the impression that WP is unreliable. Aborig (talk) 10:01, 6 September 2009 (UTC)

The rectangular Venn diagram
Is there any particular reason there is no outer ring for Militarev and Ehret? Neither denies the family's validity as a whole (contrast with the "Omotic excluded" opinion).

I'm not sure if it would be better to add one for both (for informational clarity), or to do away with them for the others too, adding insted separate Omotic rings (for visual clarity).

Anyway, at any rate Ehret's Cushitic ring should be extended to encompass Omotic (ie. coinciding with Militarev's).

It might also help visual clarity if coinciding rings were thinner, so as making up a "single" ring (or perhaps with lengthwise alternating colors?) -- Trɔpʏliʊm • blah 11:29, 6 October 2009 (UTC)

"More"
The text says that "two or more" branches were grouped together by medieval scholars. In the only example given, only two are mentioned. This sole example does not justify the use of the word "more". —Preceding unsigned comment added by 86.157.178.16 (talk) 11:30, 27 November 2009 (UTC)
 * Just because the example chosen only links two of the groups, doesn't mean that only two were linked by others. For example, there is the widely held view of Semitic versus Hamitic, which links more than two groups in Hamitic.  (Taivo (talk) 13:07, 27 November 2009 (UTC))
 * The "widely held view" was not held in the Middle Ages. No one mentioned "Hamitic" or
 * "Hamito-Semitic" until about 1840. I don't count 1840 as "medieval". —Preceding unsigned comment added by 78.105.36.65 (talk) 13:32, 27 November 2009 (UTC)
 * Krapf (1810-1881) was the first to use the term linguistically. —Preceding unsigned comment added by 78.105.36.65 (talk) 13:41, 27 November 2009 (UTC)

Aleut?
Taivo, I hit the enter key too early before finishing my comment. What I was trying to say was that Holger Pedersen already included Eskimo in Nostratic, so although Greenberg classed Eskimo-Aleut as Eurasiatic, all he really added was Aleut. VikSol (talk) 03:16, 2 December 2009 (UTC)
 * But writing that Greenberg added Aleut is incorrect. It implies that Greenberg treated Aleut as separate from Eskimo.  He did not.  Greenberg added Eskimo-Aleut.  While for some of the elements you just add to what Pedersen included, you cannot say that Greenberg added Aleut because that has a totally different implication than Greenberg added Eskimo-Aleut.  (Taivo (talk) 05:41, 2 December 2009 (UTC))

Templatic morphology
The text says that a templatic morphology exists. This seems to refer to the importance of consonants in Semitic languages. Verbs with stable vowels exist in Somali and Hausa. Thus, a misleading impression has been given. —Preceding unsigned comment added by 86.177.253.233 (talk) 11:05, 28 January 2010 (UTC)
 * The statement in the text does not exclude the existence of stable vowels. Templatic morphology is another term used for what is called nonconcatenative morphology. This is not only restricted to the Semitic root-and pattern morphology. It only implies that in these languages changes happen to word roots, including all kinds of ablaut or reduplication. I think that is a fair statement about Afroasiatic languages. Landroving Linguist (talk) 14:09, 28 January 2010 (UTC)
 * I can't find mention of infixation in Somali grammar. Also not sure the word "templatic" adds anything to the description and it is not a standard linguistic term AFAIK. --JWB (talk) 20:36, 28 January 2010 (UTC)
 * Not quite correct. The article mentions an infix on the past and present continuous, but admittedly this is a non-standard use of the term infix, and no manifestation of nonconcatenative morphology. In any case, the article mentions noun plural formation by reduplication, and that would qualify as nonconcatenative morphology. Be that as it may, I agree that we should only contain statements which are referenced to a good source. Since no reference is given on the templatic claim, I guess you are free to remove it. Landroving Linguist (talk) 06:57, 29 January 2010 (UTC)

Alleged present tense
The box in the article speaks of a "present tense". The Classical Arabic imperfect given is not a tense and can refer to any time. —Preceding unsigned comment added by 86.157.177.242 (talk) 09:42, 1 February 2010 (UTC)
 * The word "present" has been taken out now. —Preceding unsigned comment added by 94.30.71.244 (talk) 13:50, 28 May 2010 (UTC)

Coptic verb
The Coptic verb is mentioned in the article. Normally, all the rules of comparative linguistics mean that the oldest form of a language is used. I am not sure that the Coptic inflection given is found in Ancient Egyptian. —Preceding unsigned comment added by 86.177.253.233 (talk) 13:02, 1 February 2010 (UTC)

Greenberg
Greenberg's list of cognates for Afroasiatic should be included in the article, to give everyone a chance to study them. —Preceding unsigned comment added by 86.177.254.83 (talk) 12:32, 27 May 2010 (UTC)


 * A full list of all cognates, and just from one source which is an older one? There are two potential problems with such proposals:
 * 1. Would it be overboard? Wikipedia policies ask us not to turn Wikipedia into a dump of raw data. There are debates about what is reasonable, but I am thinking this would be pushing it? See WP:ISNOT
 * 2. There have been many lists of proposed cognates published. Why only list Greeberg's? See WP:DUE
 * --Andrew Lancaster (talk) 12:37, 27 May 2010 (UTC)
 * First, Wikipedia isn't the place for a data dump. Second, Greenberg's list of cognates is probably the least reliable of all the proposed lists because he was not a specialist in any of these languages.  --Taivo (talk) 13:03, 27 May 2010 (UTC)


 * Greenberg had some knowledge of Hausa and Hebrew. —Preceding unsigned comment added by 94.30.71.244 (talk) 16:17, 27 May 2010 (UTC)


 * I agree, that the list of Afroasiatic cognates has a tendency to evolve with the passage of time. Perhaps a separate web-site can be created, with the contributions of Ehret and others. —Preceding unsigned comment added by 86.179.51.100 (talk) 10:39, 28 May 2010 (UTC)

Additional Cognates
Independent of Ehret, I've found other possible cognates: Can this be added, in addition to other words and their respective cognates that I have found? -Ano-User (talk) 10:29, 15 October 2010 (UTC)
 * *y-m 'sea' attested in Berber (for example, Tuareg àman 'water' offers a possible cognate), Egyptian (iumā / imā 'sea'), and Semitic (Hebrew *yam).
 * If you have published this somewhere, it would be welcome. Landroving Linguist (talk) 14:23, 17 October 2010 (UTC)
 * Well, actually I have not published this anywhere, but I did research on my own using a French-Tuareg dictionary, an Egyptian Hieroglyphics Reading Book, and a Hebrew Dictionary. I decided to look for words that would be in a Swadesh word list and I chose a word randomly; the word in this case was 'sea.'  I have, however, found more possible cognates that I have not listed, using the same sources.  -Ano-User (talk) 04:22, 18 October 2010 (UTC)
 * Do any of those books actually point out this correspondence or are you the discoverer? This is important because of WP:NOR.--Andrew Lancaster (talk) 05:42, 18 October 2010 (UTC)
 * The books do not point out this correspondence, I am the discoverer of it, but since putting the information in the article would go against the WP:NOR policy, then it should not be added. Sorry for not having realized that.  -Ano-User (talk) 06:05, 18 October 2010 (UTC)

The Euler Diagram
The Euler Diagram is completely incomprehensible to me and I've studied linguistics and several sciences, can someone make some other kind of diagram? I'm certain almost no one could figure out what it is actually saying. —Preceding unsigned comment added by 72.187.167.161 (talk) 03:04, 28 December 2010 (UTC)
 * You could have said it more politely, but I agree with you entirely. Landroving Linguist (talk) 06:23, 28 December 2010 (UTC)
 * I did not make the diagram but I understand it, so if the above represents a request to the community perhaps you should also explain more about what would make the diagram better?--Andrew Lancaster (talk) 08:02, 28 December 2010 (UTC)
 * Ok, I'll try to give my interpretation of the diagram: If a language or sub-family is circled in a certain colour, then the linguist associated with that colour (e.g. dark blue for Diakonoff) includes this language into AA (I don't understand the significance of the different font sizes for languages). I would then deduce that when a language or family is not circled in that colour, the linguist would exclude this from AA. If this is the case, why then is Omotic only circled in green for Militarev? We see in the table below (which is definitely more helpful) that only Newman and Diakonoff exclude Omotic from Afroasiatic - the diagram makes the inclusion of Omotic appear to be a crackpot fringe opinion, which it definitely is not. Of course, I may misread the diagram, but that is the point of this complaint - if it is so easily misunderstood, maybe it should better be left out. As I say, all the information is perfectly arranged in the table below the diagram. Landroving Linguist (talk) 05:57, 29 December 2010 (UTC)
 * Well, I suppose the circles are only meant to define any level of grouping, but I see what you mean by the implication that they could be interpreted differently, making it seem like, for example, Ehret does not believe Egyptian is Afro Asiatic, which he does.--Andrew Lancaster (talk) 08:14, 29 December 2010 (UTC)
 * So I did mis-interpret the diagram. The caption actually gives an indication how to read it, but on a challenged mind like mine it just appeared to be a maze of colourful lines. Once you understand that uncircled languages are not meant to be excluded, but just not grouped with other languages, things become clear. However, the diagram has then no way to tell you that Diakonoff and Newman in fact do exclude Omotic from Afroasiatic. And still I don't get the significance of the bold print for Cushitic. Landroving Linguist (talk) 12:41, 29 December 2010 (UTC)
 * I am guessing that Cushitic is being presented as a large/diverse group in itself, compared to the others?--Andrew Lancaster (talk) 14:23, 29 December 2010 (UTC)

Bias towards Arabic
This article seems to minimize the extensive influence of Berber in Africa. Berber is spoken throughout North Africa, Mali, Niger, Burkina Faso and parts of Egypt, Libya and Mauritania. It has a long history, much longer than Arabic on the continent and millions of native speakers. Please do not attempt to minimize this influence by claiming it is only spoken in North Africa or the "Sahel" it is much more extensive than that and has an extensive reach into West Africa. Please use the talk page, especially as you are an anonymous editor. Sincerely, — Preceding unsigned comment added by Otherperson2012 (talk • contribs) 15:42, 24 June 2011 (UTC)
 * I don't see how there is a bias towards Arabic when it is only mentioned in one sentence in the introduction. The Berber languages have also not in any way been minimized. It is indicated that "Other widely-spoken Afroasiatic languages include the Berber languages, which are spoken natively across Morocco, Algeria, and Tunisia, as well as parts of Niger, Libya and Mali." The fact that Afro-Asiatic languages are spoken primarily in the Middle East, North Africa, the Horn of Africa, and parts of the Sahel is also referenced whereas the assertion that the phylum is dominant in West Africa in general (where the Niger-Congo languages actually predominate) instead of just the Sahel is not. 76.65.172.191 (talk) 16:02, 24 June 2011 (UTC)

What about Guanche language?
Guanche language Böri (talk) 08:49, 13 April 2012 (UTC)

"Macrofamily"
The term "macrofamily" is not widely used for this language family. It is not a technical term used in historical linguistics. --Taivo (talk) 14:48, 10 March 2011 (UTC)


 * I think it's rather obvious Afro-asiatic/Afrasian should be labelled as a macrofamily rather than simply a family. In fact, the illustration's title read as a "One of the world's major language families", so please don't keep undoing my editing. — Preceding unsigned comment added by Talskubilos (talk • contribs) 14:56, 10 March 2011 (UTC)


 * No, since "macrofamily" is not a standard term in linguistics, your opinion of "rather obvious" is immaterial. The great majority of historical linguists call this a language family, not a "macrofamily".  In Campbell and Mixco's A Glossary of Historical Linguistics, for example, the term is said to only apply to controversial and speculative groupings.  Afroasiatic is not in that category.  Your editing is not based on references or common usage in linguistics.  --Taivo (talk) 15:03, 10 March 2011 (UTC)

I'm afraid the term "macrofamily" is mostly used among language comparativists (e.g. Nostraticists), but it's still valid when applied to a large grouping like AA. This is the only reason why it has "controversial and speculative" connotations (for that matter, I think the definition in Wikipedia's article is rather misleading, so I'm going to revise it before coming back to this article).

For the specific use of this term, see for example the expression "Afro-Asiatic Macrofamily" in The Tower of Babel when referring to Militarev's articles (talk) 15:23, 10 March 2011 (UTC)
 * One usage of "macrofamily" in reference to Afroasiatic does not constitute sufficient reason to change it here. In far more sources it is called a "language family".  "Macrofamily" applies to Nostratic because Nostratic is controversial in a way that Afroasiatic is not.  The claim that Afroasiatic is not a valid genetic grouping is marginal, not mainstream, whereas Nostratic's position is the opposite.  --Taivo (talk) 15:29, 10 March 2011 (UTC)
 * I have clarified the macrolanguage intro with a reference to Campbell and Mixco to make its usage clear--it is only used for speculative groupings, not uncontroversial ones such as Afroasiatic. --Taivo (talk) 15:40, 10 March 2011 (UTC)

I'm afraid you're saying "macrolanguage" when actually referring to "macrofamily", which AREN'T synonyms. I've changed Campbell-Mixco by Trask, who literally says "a macro-family is an exceptionally large language family", which is the case of Afrasian/Afro-Asiatic. The confusion arises because this term (better spelled with hyphen) has negative connonations because it's mostly applied to controversial groupings like Nostratic, not because it's specifically devised to designate them. Talskubilos (talk) 09:06, 11 March 2011 (UTC)
 * Yes, both "macrolanguage" and "macrofamily" are rather ad-hoc and not very usable, but I did, indeed, mean "macrofamily". It's still not appropriate to Afro-asiatic because only Trask uses it.  "Language family" is a far more common term here.  --Taivo (talk) 09:27, 11 March 2011 (UTC)
 * First, you must get a consensus for changing "language family" to "macrofamily" before you make the change. Read WP:BRD before you go editing again.  Your change was opposed, so before making it again you must build a consensus here on the Talk page BEFORE making the change again.  You have not built a consensus.
 * Second, Trask's usage of "macrofamily" in referring to Afro-asiatic is a minority usage, so I switched the order of the Campbell & Mixco reference (which you should NOT have deleted) and the Trask reference at macrofamily to reflect the most common usage first and the less common usage second. --Taivo (talk) 09:35, 11 March 2011 (UTC)

I've re-edited macrofamily to unsure the proper definition hierarchy is maintained. Firstly, a macro-family is broadly defined as a very large language family. What happens most proposed macro-families are speculative or controversial groupings of existing language families, hence its restricted use. I see no problem in defining Afro-asiatic as "one of world's largest language families or macro-family" provided the bare definition (that is, without further connotations) of this term is applied. Talskubilos (talk) 09:52, 11 March 2011 (UTC)
 * The following sources use "language family" to describe Afro-Asiatic:
 * William Bright, International Encyclopedia of Linguistics (1992, Oxford) [as well as other works by Bright]
 * Merritt Ruhlen, On the Origin of Languages (1994, Stanford) [as well as other works by Ruhlen]
 * Vitaly Shevoroshkin & Alexis Manaster Ramer, "Some Recent Work on the Remote Relations of Languages," Spring from Some Common Source, Investigations into the Prehistory of Languages (1991, Stanford), pp. 178-199. [They use "macrofamily" only for proposals of deep relationship such as Sino-Caucasian and Nostratic, and "language family" for established groups such as Afro-Asiatic and Indo-European]
 * These references demonstrate the most common usage of "language family" versus "macrofamily" and show that Afro-asiatic is most consistently called a "language family" and not a "macrofamily". --Taivo (talk) 09:54, 11 March 2011 (UTC)

Of course AA is a "language family". In absence of consensus, I'll leave out the usage of "macrofamily" for AA but change it for Nostratic in the footnotes. Talskubilos (talk) 10:00, 11 March 2011 (UTC)
 * That is the correct usage of "macrofamily". Now you need to stop your edit warring at macrofamily.  You are the "innovator" there since the text as I found it already emphasized the "proposed" nature of the term.  You are still subject to WP:BRD there and must reach a consensus before putting any emphasis on a minority position in linguistics.  --Taivo (talk) 10:11, 11 March 2011 (UTC)

You could also try a Google search with the terms "Afro-Asiatic macro-family" or "Afrasian macro-family" and you'll be surprised. Omotic languages quotes a Diakonoff's article which mentions the expression "Afrasian macrofamily". So you've got two references of Afro-asiatic as a macro.family. Talskubilos (talk) 11:34, 11 March 2011 (UTC)


 * I would be careful of calling Afro-Asiatic a linguistic "macro-family" since the term is often applied to speculative, and/or controversial, groupings of languages. Personally, I think a language family of over 300 languages is not "exceptionally large" when compared to the Austronesian languages or Niger-Congo, for example.  In any case, I'd say Afro-Asiatic should be left as a "language family" on the main article. -Ano-User (talk) 15:15, 11 March 2011 (UTC)

I see the problem lies on the controversial use of the term. From printed sources, it looks like the taxomonical terms "family/"stock" and "macro-family"/"phylum" are used to describe two different levels of language grouping. But AA happens to be labelled as a family on Ruhlen's book A guide to the World' Languages. Talskubilos (talk) 18:36, 11 March 2011 (UTC)

In his new book '''Languages of the World. An Introduction''' (Cambridge, 2012), the linguist Asya Pereltsvaig refers to Afroasiatic as a macrofamily, so it appears to be a metter of scholar disagreement (Russian vs. American). Talskubilos (talk) 09:00, 11 June 2012 (UTC)
 * Well, that's not quite accurate. While the Pereltsvaig book does mention the term "macro family" in relation to Afro-Asiatic, he also refers to it elsewhere as a "family" and includes the following statement: "If Afroasiatic is a bona fide language family; that is, a grouping of languages descending from a common ancestral language, we expect to find patterns of similarity both in vocabulary (in the form of cognates) and in grammar. And indeed both types of similarities are found in Afroasiatic languages." The book also doesn't say anything about whether or not Russian linguists in general refer to AA as a macro-family. Only the author himself apparently sometimes does; and per the o/r policy, sources must directly support the material being presented. If you disagree, please provide the exact page number in the book where you believe that this information is indicated (it wasn't included in the page edit). Middayexpress (talk) 14:01, 12 June 2012 (UTC)

As you can read above on this section, other Russian linguists such as Militarev refer to Afroasiatic/Afrasian as a macro-family in their works. I also don't understand why did you replace Pereltsvaig's number of languages (374) to ethnologue.com (375), as printed sources must have preference. Talskubilos (talk) 19:36, 12 June 2012 (UTC)
 * Ethnologue is also a printed source, as I understand. Also, since the Pereltsvaig book does not indicate that Russian linguists in general refer to AA as a macro-family, you'll need a separate source that explicitly states this per the o/r policy. Middayexpress (talk) 20:02, 12 June 2012 (UTC)

Pereltsvaig explicitly states AA as a macro-family, and so she includes it in the chapter deveted to them in her book. Talskubilos (talk) 21:05, 13 June 2012 (UTC) Russian sources usually refer to AA as a macro-family: see for example http://www.kunstkamera.ru/en/museum_exhibitions/encyclopedia/subsaharan_africa/population/languages/ or the reference "Afro-Asiatic Macrofamily" in http://starling.rinet.ru/texts_new.php?lan=en. This isn't "original reseach" but encyclopaedic reseach. Talskubilos (talk) 08:51, 14 June 2012 (UTC)
 * I think perhaps you misunderstand the o/r policy. Wikipedia draws a distinction between unsourced material (material not yet attributed to a reliable source but may be attributable) and original research (material that cannot be attributed to a reliable source because it is original thought). It is not original research to state that Pereltsvaig in parts of his book describes Afro-Asiatic as a macrofamily, whereas in other parts he labels it a family (as quoted above). It is also not original research to say that the Kuntstkamera Museum's website that you just linked to refers to Afro-Asiatic as a macro-family (Nilo-Saharan and Khoisan as well, incidentally). It is, however, original research to then generalize based on this that Russian linguists at large refer to Afro-Asiatic as a macro-family. Pereltsvaig does not assert this anywhere in his book, nor does the Kuntstkamera Museum on its website or the other website you've also linked to. Per WP:ATTRIB, "material added to articles must be directly and explicitly supported by the cited sources". This means that you need a reference specifically supporting the assertion that Afro-Asiatic is "also referred as a macrofamily in Russian sources" because the term "Russian sources" alone is much broader than just Pereltsvaig, Kuntstkamera and the other website. Middayexpress (talk) 14:32, 14 June 2012 (UTC)
 * The Kuntstkamera Museum also apparently elsewhere refers to Afro-Asiatic, alongside Indo-European and Altaic, as a language family; albeit a large one (c.f. ). Middayexpress (talk) 14:51, 14 June 2012 (UTC)

Definitively, Pereltsvaig lists AA as a macrofamily in the chapter 11 "Macro families", section 1 of his book. The inconsistence you mentioned in Kunstkamera is probably due to co-authoring. Talskubilos (talk) 20:58, 14 June 2012 (UTC)
 * Talskubilos, you need to chill out. One linguist calling AA a macrofamily does not make it an appropriate entry for the first sentence.  The vast majority of linguists call AA a family and that's that.  It is appropriate to mention lower in the article (where it is already mentioned I believe, properly, as Middayexpress states, without calling it a "Russian" POV), but not in the first sentence.  It's simply a fringe, extreme minority position at this time.  One or two linguists using a term that is not normally used with this meaning does not justify mentioning in the lead.  --Taivo (talk) 21:28, 14 June 2012 (UTC)
 * I couldn't quickly find where "macrofamily" is mentioned in the article, so if it needs to be added in some section, that's fine with me, but not in the first sentence of the article which should always reflect the majority view and important minority views, not the views of tiny minorities. --Taivo (talk) 21:37, 14 June 2012 (UTC)
 * Just got my copy of Pereltsvaig's book. First, it's not a scientific treatise, it's an introductory book for students, not even a "standard reference work", so we shouldn't put as much weight on his comments as we would put on, say, one of the two reconstructions of Proto-Afroasiatic/Afrasian.  Second, his use of "macrofamily" isn't really standard--he considers a macrofamily to be a family of families, whether proven (as Afroasiatic and Uralic) or unproven (as Nostratic and Altaic), but he doesn't have any real definition for "family" that would make a family different from a proven "macrofamily".  Third, he's inconsistent on whether he calls Afroasiatic a family or a macrofamily, using both.  Fourth, nowhere does he make a distinction between Russian Afroasiaticists and non-Russian ones, but we already suspected that.  In conclusion, calling Afroasiatic a macrofamily is, as I have already argued, inappropriate for the lead.  --Taivo (talk) 22:27, 19 June 2012 (UTC)

THE Oldest Language Group, not "one of the oldest."
I write this because I just came from the Indo-European language page which stated that the family of languages was the second oldest AFTER the Afro-Asiatic family. Please change for consistency. — Preceding unsigned comment added by 71.233.213.76 (talk) 03:58, 9 August 2011 (UTC)


 * Not sure how much sense it makes to speak of the age of a family. Proto-AA is one of the oldest reconstructed langs. You'd need a good ref for it being the oldest: dating reconstructions is unreliable. — kwami (talk) 23:41, 19 June 2012 (UTC)
 * And, in essence, ALL language groups are equally old since no language family springs anew without coming from something older. What we're talking about is the point at which extant (and evidenced non-extant) daughter languages start to diverge from a proto-language, not the absolute age of a family.  --Taivo (talk) 23:47, 19 June 2012 (UTC)


 * I suspect that they mean "earliest attested". If so, this should be clarified.  -- TimNelson (talk) 12:06, 28 June 2012 (UTC)


 * No, it means a solidly-established language family whose original proto-language must have been spoken at a rather early date compared to the proto-languages of other solidly-established language families... AnonMoos (talk) 15:26, 22 August 2012 (UTC)
 * That comment really doesn't make linguistic sense. I've edited the paragraph to be more neutral in tone without the grandstanding.  --Taivo (talk) 15:40, 22 August 2012 (UTC)
 * It does have a "valid linguistic meaning". Proto-Indo-European was almost certainly close to being a single language ca. 4000 B.C., while Proto-Afroasiatic almost certainly didn't come close to being a single language until well before 4000 B.C. Of course we know a lot more in detail about Proto-Indo-European than we do about Proto-Afroasiatic (nothing remotely close to Schleicher's fable would be feasible)... AnonMoos (talk) 19:31, 22 August 2012 (UTC)
 * You act like these dates are firm and fixed. They are all only guesses, so any absolute comments about this or that language group being the oldest is simply blowing hot air.  The only reason we know that the Afroasiatic family is unified is because the record of these languages is so old.  It gets us 4000 years or so closer to the proto-language.  There may very well be language families in the Americas that are older, but we can't prove it because our record of those languages starts 4000 years later.  We have to be very careful about the way that such comments are phrased and avoid at all costs absolutes.  --Taivo (talk) 20:15, 22 August 2012 (UTC)
 * Actually, 4000 B.C. was obtained by adding a certain safety margin to commonly postulated (but of course somewhat inexact) dates for Proto-Indo-European. In any case, the date for proto-Afroasiatic is certainly significantly more remote than that for Proto-Indo-European. I commiserate with you about life not being fair for language groupings without lengthy written traditions in the Americas, but it's not about fairness, it's about what can reasonably be deduced with some degree of confidence in the current state of knowledge. AnonMoos (talk) 23:33, 22 August 2012 (UTC)
 * You didn't get the point. The point is that we cannot say with any absolute certainty that "Afroasiatic is the oldest".  We are tied to the evidence and the guesses at dates.  Since there can never be any real certainty of dates (short of using a time machine), we cannot make absolute statements.  There is no actual evidence for the dating of Proto-Afroasiatic.  All the evidence is circumstantial, there is nothing concrete.  So in the absence of concrete evidence, there can be no absolute statements.  --Taivo (talk) 03:27, 23 August 2012 (UTC)
 * In any case, any such statement would need a first-rate reliable source to back it up. To write this here on the basis of your own (however careful and conservative) estimation is original research, and we don't do that in Wikipedia. Landroving Linguist (talk) 18:51, 24 August 2012 (UTC)

(outdenting) In response to Taivo's claim "The only reason we know that the Afroasiatic family is unified is because the record of these languages is so old", this is certainly false. Only two of the constituent families are notably old (Egyptian and Semitic), and as for those, (1) Egyptian was already hugely modified even in its earliest attestation 5000 years ago, and AFAIK doesn't play such a major part in A-A reconstruction; and (2) even if we had only the current Semitic languages to work from we could reconstruct most of Proto-Semitic (mostly missing case endings, some original mood endings, and the proper interpretation of the different verb forms). Even a cursory examination of the prefix conjugations of various modern languages — e.g. Yemeni Arabic (Semitic), Kabyle (Berber), Somali (Cushitic), Hausa (Chadic) — shows a clear resemblance that cannot be due to chance. Similarly, it is easy to find basic vocabulary items shared across most or all constituent families using only modern languages. Certainly, it helps to have a lengthy history of writing, but I don't think it would matter all that much except for Egyptian. In this case, the A-A languages have just been unusually conservative in their development. I think the non-concatenative morphology plays a major part, leading to strong analogical pressure that militates against conditioned sound change in consonants. You can clearly see the effect of this in the history of the Semitic languages, for example, where most consonant sound changes are unconditional -- very different from e.g. Indo-European. The result is that e.g. most Proto-Semitic consonants are reflected largely unchanged in the modern languages -- e.g. t, d, p, b, k, g, s, z, r, l, m, n, w, y, etc., which usually either show up totally unchanged or with unconditional and usually easy-to-understand changes, e.g. p -> f, w -> v, g -> dʒ. The only significant conditional changes I can think of off the top if my head are the begedpeket lenition changes; contrast the endless stream of palatalizations, lenitions, nasalizations, etc. in the Indo-European languages such that French "leaf" is less than 2,000 years removed from its perfect Spanish cognate  "leaf" but the connection is now totally unrecognizable, and the connections are difficult to see even between the common ancestral form  (Proto-Romance) and either of the above reflexes. Benwing (talk) 09:01, 25 August 2012 (UTC)

"Transcontinentally"
I am unable to see the content of Semitic Languages: Outline of a Comparative Grammar. What does it say there? It is attached to the claim that ''The term "Afroasiatic" (often now spelled as "Afro-Asiatic") was later coined by Maurice Delafosse (1914). However, it did not come into general use until Joseph Greenberg (1963) formally proposed its adoption. In doing so, Greenberg sought to emphasize the fact that Afroasiatic was the only language family that was represented transcontinentally, in both Africa and Asia''.

Surely there are other language families that are represented on more than one continent, for example Indo-European, Altaic, Uralic, Austronesian and Eskimo–Aleut. --Theurgist (talk) 15:18, 3 September 2012 (UTC)


 * The sentence specifies in both Africa and Asia? Maybe your proposal is to remove the word transcontintentally? Indeed it does not seem to add anything.--Andrew Lancaster (talk) 09:31, 4 September 2012 (UTC)


 * I cannot propose anything before I actually know what the source says. That said, I wouldn't want to intentionally misquote a source just to say what I like. By the way, Afrikaans is an Indo-European language. --Theurgist (talk) 07:59, 7 September 2012 (UTC)
 * Also, Austronesian languages are represented in both Africa (by Malagasy) and Asia (by Malay, among others), strictly speaking. --Theurgist (talk) 21:30, 7 September 2012 (UTC)


 * AA spans Africa & Asia in a block, unlike the others, and assuming a geographical rather than geological or biological definition of Africa. — kwami (talk) 06:34, 8 September 2012 (UTC)


 * But would someone please tell me what the book says actually? For some reason the content of that particular page is not available for me. --Theurgist (talk) 09:56, 8 September 2012 (UTC)


 * "the entire language family was named 'Hamito-Semitic' in 1876 by Fr. Müller in his Grundriss der Sprachwissenschaft ... J.H. Greenberg, instead, considering that this is the only language family represented in both Africa and Asia, proposed to call it Afro-Asiatic in his work The Languages of Africa"
 * Delafosse is not mentioned. — kwami (talk) 22:08, 8 September 2012 (UTC)

Is this a notable alternative view we should include?
Only a linguist can answer this Is this a notable argument and if not is there a obvious flaw with it that refutes the comparisons.?--Inayity (talk) 11:53, 20 September 2012 (UTC)
 * First, this book is not a serious work of linguistics, it is a political, apparently anti-Semitic, rant to "prove" that Egyptian bears more relationship to Africa than to the Near East. Second, the flow of his argument is simply silly.  He claims he wants to destroy the notion of Afroasiatic, but he really doesn't.  He simply wants to cut Semitic out of Afroasiatic so that Egyptian can be solidly linked to Africa.  It's stupid, really.  He shows nice evidence for the connection of Egyptian to the African parts of Afroasiatic and thinks that by doing so he can leave the "Jews" out of the equation.  It's not a serious linguistic argument at all and is utterly fringe.  There is no need to mention this anti-Semitic rant at all.  --Taivo (talk) 12:17, 20 September 2012 (UTC)
 * But what about all the common lexicon in non-Afro-Asiatic languages, does that have merit in how language groups are defined (or is it more syntax commonalities that define languages). Its seems to be more aimed at Arabic than Hebrew, i.e. by showing that Ancient Egypt has similar lexicon to diverse non-Afro-Asiatic languages, more so than to even Arabic or Amharic. I hope i am not breaking wiki no forum rules. Personally i dont see it as antisemitic, but def anti-Semitic (language). I did also notice (as you picked up) a good argument for one thing while ignoring other arguments.--Inayity (talk) 17:23, 20 September 2012 (UTC)
 * Ah, I see what the guy is doing now. I didn't look closely enough after seeing the anti-Semiticism in the first couple of paragraphs.  There are always random similarities between languages, no matter how distant geographically or how unrelated they are.  In this case, the guy has a very tiny amount of knowledge and thinks that he can make a solid claim based on similarities without a firm grounding in actual historical and comparative linguistics.  Most scholars consider the homeland of Afroasiatic to be in Africa.  Only a tiny fringe look at the Middle East as the homeland.  The ancestor of the Nilo-Saharan languages which the author uses would have been a neighbor to Proto-Afroasiatic so borrowing is not surprising.  If the homeland of Proto-Afroasiatic was further to the west, then it would have also bordered ancestral Niger-Congo languages.  In any event, the correspondences between Egyptian and Afroasiatic are in the hundreds while these correspondences that the author notes are in the tens only.  The primary issue in determining relationship is not even lookalikes, but regular sound correspondences.  Did you know that English "do" and French "faire" are a perfect match and related to each other even though they look nothing alike?  That's how linguists work.  We don't look for lookalikes, but for systematic correspondences.  This author does none of that.  This is a fringe view.  Egyptian is very firmly established within Afroasiatic.  --Taivo (talk) 19:42, 20 September 2012 (UTC)


 * There are also lots of pan-African words, such as the word for 'knee', which are generally ignored when doing comparisons, because all they tell you is that the language is in Africa. — kwami (talk) 00:05, 21 September 2012 (UTC)

Egyptian/Coptic is Extinct
The definition of an extinct language is "no native speakers, no living descendants". Egyptian/Coptic fits that definition, therefore the Egyptian branch of Afroasiatic is extinct. --Taivo (talk) 23:16, 12 December 2012 (UTC)


 * The information is unsourced and irrelevant, I don't think which ones are extinct and which ones are still existent needs to be mentioned in that particular list and without any further explanation. If you feel so passionately about keeping that then I suggest you make a new section describing it more thoroughly because the situation with Coptic/Egyptian is more complex than just "extinct". Turnopoems (talk) 15:08, 13 December 2012 (UTC)
 * Turnopoems, three different editors disagree with you. Read WP:CONSENSUS.  If a language has no native speakers and no living descendants, it is extinct.  In this case the entire Egyptian branch of Afroasiatic is extinct and three other editors disagree with your opinion.  --Taivo (talk) 18:15, 13 December 2012 (UTC)
 * The issue is no longer about whether Coptic is extinct or not, which I pointed out earlier. As I said, I don't think just mentioning that it is extinct and leaving it at that will suffice because it does not correctly describe the status of the Coptic/Egyptian language. Adding an info box with a more thorough description of the situation regarding the Coptic language should do the trick. Turnopoems (talk) 20:50, 13 December 2012 (UTC)
 * The info box just gives the basics. It's normal practice to mark which languages and branches are extinct in lists like this. That's entirely appropriate. The details belong in the section or article on Coptic, and indeed, in that info box it's all spelled out. — kwami (talk) 21:30, 13 December 2012 (UTC)

Refinements to the article
Will be gathering sources for proposed refinements...give me a day or three to gather them and report back here. The sources will have implications for this article as well as other Afroasiatic related articles...see the suggested scholarly sources on the Beja people talk page. The information on these respective pages needs to be updated and refined. Found these sources in doing research. I think they can be very helpful in the article(s). A.Tamar Chabadi (talk) 23:56, 2 March 2013 (UTC)

Actual numbers for Afroasiatic speakers
Berber Branch

14 million Atlas languages 5 million Kabyle language 1.5 million Riffian language 1.4 million Shawiya language 1.2 million Tuareg languages 180,000 Nafusi language 30,000 Siwi language 12,000 Ghadamès language 10,000 Ghomara language 5,600 Sokna language 3,000 Awjila language 2,100 Zenaga language

Total = 23,342,700

Chadic Branch

55 million Hausa language 400,000 Ngas language 300,000 Kamwe language 300,000 Mwaghavul language 250,000 Bade language 230,000 Massa language 230,000 Musey language 200,000 Goemai language 200,000 Gera language 150,000 Azumeina language 150,000 Karekare language 130,000 Tangale language 120,000 Ron language 110,000 Kofyar language 100,000 Bole language 25,000 Tumak language 81,000 Nancere language 80,000 Ngizim language 78,000 Warji language 60,000 Dangaléat language 60,000 Ngamo language 50,000 Kera language 50,000 Ngeté-Herdé language 50,000 Saya language 50,000 Boghom language 35,000 Mubi language 35,000 Peve language 34,000 Gabri language 30,000 Tobanga language 30,000 Miya language 27,000 Gwandara language 26,000 Lele language (Chad) 26,000 Fyer language 25,700 Masmaje language 25,000 Galambu language 25,000 Pero language 22,000 Montol language 22,000 Polci language 20,700 Zari language 20,000 Kimré language 20,000 Mesme language 20,000 Migaama language 20,000 Kanakuru language 18,000 Kabalai language 17,000 Kwang language 17,000 Jorto language 17,000 Pyapun language 16,000 Kulere language 15,000 Guruntum language 14,000 Bidiyo language 14,000 Giiwo language 14,000 Yiwom language 12,000 Mokilko language 12,000 Mburku language 11,000 Kushi language 11,000 Ɗuwai language 10,400 Birgit language 10,000 Kajakse language 10,000 Maaka language 10,000 Kwaami language 10,000 Tal language 9,000 Geruma language 8,800 Dass language 8,500 Toram language 8,000 Pa'a language 7,400 Somrai language 7,200 Diri language 7,000 Mogum language 6,500 Ndam language 6,000 Deno language 6,000 Miship language 6,000 Geji language 5,000 Mire language 5,000 Piya language 5,000 Cakfem-Mushere language 3,800 Siri language 3,100 Kir-Balar language 3,000 Koenoem language 3,000 Kutto language 3,000 Tambas language 3,000 Sha language 2,500 Kholok language 2,500 Gadang language 2,200 Zirenkel language 2,000 Sarua language 2,000 Ciwogai language 2,000 Kariya language 2,000 Zumbun language 1,500 Jonkor language 1,300 Jelkung language 1,100 Ubi language 1,000 Kujargé language 1,000 Tala language 1,000 Mundat language 1,000 Jimi language (Nigeria) 250 Miltu language 100 Boor language 40 Buso language 3 Mabire language

Total = 59,263,943

Cushitic Branch

17 million Somali language 4.5 million Eastern Oromo language 3.9 million Southern Oromo language 3 million Sidamo language 1.9 million Maay language 1.4 million Afar language 1.2 million Beja language 980,000 Gedeo language 890,000 Kambaata language 490,000 Awngi language 460,000 Iraqw language 280,000 Alaba-K’abeena language 250,000 Hadiyya language 240,000 Konso language 220,000 Saho language 210,000 Xamtanga language 91,000 Bilen language 70,000 Burji language 69,000 Gawwada language 66,000 Orma language 65,000 Dirasha language 60,000 Daasanach language 60,000 Rendille language 59,000 Libido language 58,000 Garre language 23,000 Tunni language 23,000 Dabarre language 22,000 Jiiddu language 18,000 Tsamai language 18,000 Bussa language 13,000 Waata language 8,000 Aweer language 7,200 Arbore language 5,500 Baiso language 1,700 Qimant language 1,500 Dobase language 400 Dahalo language 12 Ongota language 8 El Molo language

Total = 37,659,320

Egyptian Branch

? Coptic Language

Total = ?

Omotic Branch

2 million Gamo-Gofa-Dawro language 1.6 million Wolaytta language 830,000 Kafa language 350,000 Bench language 240,000 Aari language 160,000 Koore language 95,000 Maale language 93,000 Basketo language 92,000 Yem language 80,000 Shakacho language 74,000 Hamer language 56,000 Gayil language 39,000 Sheko language 38,000 Shinasha language 37,000 Oyda language 34,000 Dizin language 30,900 Dorze language 20,000 Melo language 19,000 Zayse-Zergulla language 13,000 Chara language 7,200 Nayi language 5,000 Bambassi language 3,000 Hozo language 3,000 Seze language 3,000 Ganza language 2,800 Kachama-Ganjule language 570 Dime language 500 Anfillo language

Total = 5,925,970

Semitic Branch

357-397 million Arabic language 25 million Amharic language 25 million Oromo language 7 million Hebrew language 7 million Tigrinya language 1 million Tigre language 940,000 Silt'e language 550,000 Aramaic language 440,000 Sebat Bet Gurage language 280,000 Inor language 260,000 Soddo language 200,000 Mesqan language 120,000 Mehri language 120,000 Harari language 90,000 Muher language 64,000 Soqotri language 44,000 Argobba language 25,000 Shehri language 4,900 Zay language 3,000 Dahalik language 600 Harsusi language 200 Bathari language 100 Hobyót language

Total = 425-465 million

Absolute Total = 551-591 million afroasiatic speakers

166.147.72.146 (talk) 01:07, 26 May 2014 (UTC)

Afroasiatic
The Arab league's population is somewhere between 380-420 million

http://en.m.wikipedia.org/wiki/Arab_League

http://en.m.wikipedia.org/wiki/Arabs

To account for the Berber speakers in the Arab world you could subtract 23,000,000 million.

Which would leave you ranging with 357-397 Arabic speakers, which when tallied with the other Afroasiatic speakers, would still equal somewhere between 551-591 million afroasiatic speakers instead of 362 million speakers. These numbers have been double checked with population numbers in every country to verify the statistics are as correct as possible. — Preceding unsigned comment added by 166.147.72.157 (talk) 18:10, 26 May 2014 (UTC)


 * Dear 166.147.72.157 (or 166.147.72.146, whatever your IP may be at any given time)! Thanks for assembling and researching this impressive amount of information. But before you go and edit any more articles based on that information, please read the Wikipedia policy on original research. In a nutshell, Wikipedia is an encyclopedia of published knowledge, and all information you add or change needs to go back to some high-quality verifiable source. Particularly, you are not allowed to include information that goes back to your own research, such as calculations on numbers that you draw up from a variety of sources, as you have done on this talk page (where it is perfectly fine to do it) and in some articles (where you are not allowed to use these numbers). Also note that Wikipedia pages do not count as reliable sources, and may not be quoted on Wikipedia to back up a claim. Although it may be true that adding up the numbers from all the above Wikipedia pages gets you to the total number you calculate, the nature of Wikipedia does not guarantee that all the numbers on all these pages are good numbers, and therefore such calculations are not permitted. These are the rules of Wikipedia, and not adhering to them will result in your edits being reverted, as has happened before. Another advice: other editors will look at your edits more kindly if you sign them with a proper Wikipedia user name, and not do them anonymously. Landroving Linguist (talk) 08:34, 27 May 2014 (UTC)

Category "Language families"
I think, it is not an error to delete this category, as it present in category "Afroasiatic languages"? Cathry (talk) 14:09, 7 September 2015 (UTC)

Actual numbers for Afroasiatic speakers
oromo is not a semitic language.please correct. — Preceding unsigned comment added by 79.183.23.111 (talk) 10:42, 4 October 2015 (UTC)

West Africa
An editor now for the second time made an edit to the effect that Afroasiatic languages are spoken in West Africa, as opposed to just the Sahel. The rationale is that Arabic is spoken as a lingua franca all over the West African countries (source: edit summary of first edit). I think this notion is wrong on two points: The geographic distribution should only include areas where a language (family) is spoken as a mother tongue. And I think it is an exaggeration to claim that Arabic serves as a lingua franca in any sub-Saharan country. Certainly it is the language of Islam, but no-one ever uses it outside of a religious context - that is not anyone's definition of a lingua franca. Can we get a consensus that we keep the geographic distribution to the Sahel and not all of West Africa? Landroving Linguist (talk) 05:36, 20 September 2016 (UTC)
 * Arabic is an official language in the four Arab League states below the Sahara, but indeed none of these areas are in West Africa. The language family infobox documentation also indicates that this parameter is for the "geographic region in which its languages are mainly spoken." That would be the Sahel, where the Chadic languages are spoken. The main language family spoken in West Africa is instead Niger-Congo, so good call. Soupforone (talk) 16:09, 20 September 2016 (UTC)

que patetico as linguas berbero semiticas tiveram origem entre norte do saara e leste proximo não no congo; proto semitico arabia central, proto berbere possivelmente golfo de sidra e proto berbero semitico possivelmente entre foz do nilo sinai, natufianos etc — Preceding unsigned comment added by 187.64.32.15 (talk) 12:43, 25 January 2017 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified 2 external links on Afroasiatic languages. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20090104234232/http://community.livejournal.com/terra_linguarum/95880.html to http://community.livejournal.com/terra_linguarum/95880.html
 * Added archive https://web.archive.org/web/20100818025156/http://community.livejournal.com/terra_linguarum/95627.html to http://community.livejournal.com/terra_linguarum/95627.html

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 18:36, 5 April 2017 (UTC)

Migration and population replacement
So did the Afro-Asiatic speakers replace or dominate the earlier populations of the Middle East/North Africa? I know that for Europe it is thought that the Indo-European speakers only dominated the local population and most modern Europeans actually descend from the earlier population but there is no information about this matter in this article. — Preceding unsigned comment added by 93.227.29.71 (talk) 07:46, 22 March 2016 (UTC)

Europeans(especially Western Europeans, West and East Slaves) actually descend from the Indo-Europeans, as for the Afro-Asiatic people, they are the ancestors of all modern Afro-Asiatic speakers except the Semites. The Proto-Semites dominated the local population and later merged with them forming the Modern Semites. --90.20.179.160 (talk) 10:31, 2 September 2017 (UTC)

New view
Loprieno, in 1995, has suggested that Afroasiatic might be a merely typological group, and not a genetic one. This opinion should be in the article. — Preceding unsigned comment added by 178.8.40.239 (talk) 09:04, 24 November 2017 (UTC)
 * See https://en.wikipedia.org/wiki/Egyptian_language#Classification for more detail. — Preceding unsigned comment added by 178.8.40.239 (talk) 09:06, 24 November 2017 (UTC)

That is actually not Loprieno's theory (he acknowledges the validity of Afroasiatic ). It is G. W. Tsereteli's allogenetic paradigm ), which has little currency among linguists. Soupforone (talk) 15:49, 27 November 2017 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified 2 external links on Afroasiatic languages. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://archive.is/20121209090902/http://mc1litvip.jstor.org/pss/3628690 to http://mc1litvip.jstor.org/pss/3628690
 * Added archive https://archive.is/20130113032053/http://multitree.org/codes/afas to http://multitree.org/codes/afas

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 13:32, 1 December 2017 (UTC)

Fourth largest language family?
Funny, the exact same thing is written about Austronesian languages with the number 386 million (as opposed to 350 million here). Should it perhaps be changed to fifth? TFighterPilot (talk) 14:42, 13 March 2018 (UTC)


 * No. According to the source quoted here (Ethnologue) the claim is correct for Afroasiatic. It is not clear to me what serves as the basis for the claim on the Austronesian page, there are no in-line citations. Interestingly, the Ethnologue total for Afroasiatic is also much higher (495.000), so that may explain the confusion. Landroving Linguist (talk) 06:57, 14 March 2018 (UTC)

1997
Bender, in 1997, does not mention Egyptian at all, although it is included by most others. — Preceding unsigned comment added by 86.175.127.218 (talk) 11:11, 13 September 2018 (UTC)

Paywall
A recent edit stated that a cited source is not sufficient because it is behind a paywall. WP:PAYWALL states that a paywall is not a reason to reject a good source. Landroving Linguist (talk) 10:34, 26 October 2018 (UTC)
 * Landroving Linguist, The source is not rejected, which is why it is still there. A better source might be less restrictive, which is why the assertion is tagged better source needed.  Adding another source doesn't require removing this one; they can cooexist.  Another way to facilitate verifiability with a source behind a paywall, is to use the   param of the cite web template to quote an excerpt from the page which backs the portiion of content in the article depending on that source, for the benefit of those who are not able to access the source directly. If you have access and can do that, it would be appreciated. Mathglot (talk) 23:18, 26 October 2018 (UTC)
 * Fair enough. Quoting doesn't do much good here, though, as the cited passage in Ethnologue lists the varieties and the countries where they are spoken, and makes no claim about countries where it is not spoken. Somebody making the claim that Arabic is spoken in Somalia as a native language would have to provide a source for that. Landroving Linguist (talk) 12:00, 27 October 2018 (UTC)

Hausa speakers
Landroving Linguist What is not supported? What did you want to revert?Ppdallo (talk) 16:13, 14 November 2018 (UTC)

Hi Ppdallo, sorry for my late response, I was offline for a few days. In Wikipedia it important that the information given on each page matches exactly what the source given in support actually states. You changed the page to the effect that there are more than 65 million Hausa first-language speakers, but you provided no new source that actually claims that. The article still links to the Ethnologue, right after your edit, and Ethnologue states the following: "48,300,000 in Nigeria, all users. L1 users: 33,300,000 (2015). L2 users: 15,000,000. Total users in all countries: 63,154,100 (as L1: 43,654,100; as L2: 19,500,000)." This means that as long as you cannot provide a reputable source that actually states that there are more than 65 million Hausa L1 speakers, the previous edit needs to stand. These are the Wikipedia rules, and you will experience reverts of your edits if you do not abide by these rules. Landroving Linguist (talk) 17:10, 17 November 2018 (UTC)

Hi Landroving Linguist Are you serious? Where did you get that ethnology from? According to CIA word factbook (which I have cited already) the Hausa constitute 27.4% of Nigeria’s 203 million people and if you do the math it will give you 55.6 million people in Nigeria alone. Add that to the 10 million in Niger and what do you get?Ppdallo (talk) 18:11, 17 November 2018 (UTC)

Hi again, Landroving Linguist I have updated Hausa population to over 70 million on Hausa people page as per 55.6 million in Nigeria according to CIA factbook and the rest of other countries where Hausa people live. Pls confirm because I am going to update it here as well.Ppdallo (talk) 20:37, 17 November 2018 (UTC)


 * Yes, I am serious. Not ethnology, but Ethnologue, the source quoted right after the statement of the Hausa language population figure. Ethnologue is the source that is widely accepted in Wikipedia for language information, if not a better secondary source is available. The CIA fact book is *not* a secondary source. Also you are not allowed to make calculations based on different sources (like speaker numbers in million based on stated percentages) and state the results as fact on wikipedia pages, as this is considered Original research, something that no editor must ever do and what gets reverted without fail. Please acquaint yourself with the rules of editing wikipedia here before you make many more edits, as this will save you and others a lot of frustration. Landroving Linguist (talk) 21:51, 17 November 2018 (UTC)

== Added two references as to the research on the classification of Meroitic as an unclassified Afroasiatic language. Added wikilink to Bruce Trigger who proposed that it was. Added two refuting references. Removed 'by whom?' tag. ==

Added two references as to the research on the classification of Meroitic as an unclassified Afroasiatic language. Added wikilink to Bruce Trigger who proposed that it was. Added two refuting references. Removed 'by whom?' tag.


 * Trigger, Bruce, G. (1964) ‘Meroitic and Eastern Sudanic: A Linguistic Relationship?’ Kush 12. 188-194.


 * Trigger, Bruce, G. (1977) ‘The Classification of Meroitic: Geographical Considerations’ in Schriften zur Geschichte und Kultur des Alten Orients 13. 421-435.


 * Hintze, Fritz, ‘Some Problems of Meroitic Philology’, Abdel Gadir Mahmoud Abdalla (ed.), Studies of the Ancient Languages of the Sudan, Sudanese Studies 3, Khartoum University Press, Khartoum, 1974, 73-78.


 * Hintze, Fritz ‘Beiträge zur meroitische Grammatik.’ Meroitica 3, Berlin, 1979, 1-214.

For more detailed information please download this PDF file, written in 2006, or email the author, Professor Kirsty Rowan, University of London, at 110064@soas.ac.uk:

| Meroitic – an Afroasiatic language?

Please review and/or revert as needed. B'H. 172.250.246.150 (talk) 17:40, 3 December 2018 (UTC)
 * Looks good to me. Get yourself a username and provide more of that to Wikipedia! Cheers, Landroving Linguist (talk) 12:42, 9 December 2018 (UTC)

Coptic prefixing verb conjugation
I believe Coptic should be removed from the table illustrating prefixing verb conjugations. Coptic's prefixes are proclitic pronouns, not inherited verbal inflections, and don't follow the pattern described. In Middle Egyptian: An Introduction to the Language and Culture of Hieroglyphs, James P. Allen only describes suffixing verb conjugations. — Preceding unsigned comment added by Criminy1 (talk • contribs) 03:48, 9 April 2019 (UTC)


 * Agree, and done. – Austronesier (talk) 15:50, 10 April 2019 (UTC)

Shared Vocabulary Table
Many of the terms listed are marked as reconstructed (*) even when an example in a language is given. 2600:1700:7A20:5900:54AE:D452:5FF6:9E3E (talk) 00:05, 17 October 2019 (UTC)
 * Only apparently so. For example, when it says Agaw, it refers to a proto-language that is reconstructed from the documented knowledge of a number of currently spoken languages. So as far as I can see, the reconstruction stars are correctly assigned to these examples. Landroving Linguist (talk) 08:20, 17 October 2019 (UTC)

Missing info on what Afroasiatic languages have in common
To persons who have some familiarity with an Indo-European language such as English or Spanish, and an Afroasiatic language such as Arabic or Hebrew, there are some striking differences. I know more about Hebrew than Arabic, so for some of these points I don't mention Arabic due to lack of knowledge. The article should explain which of these and other patterns that are strongly different from Indo-European patterns are common to all Afroasiatic languages and which ones apply to only some branches. (The articles on the main branches of Afroasiatic languages should do the same.) For example, if all Afroasiatic languages have mostly 3-letter verb roots, the article should say so. Also, if any of these properties are found only in Afroasiatic languages and not in any other language groups, the article should say so. —Anomalocaris (talk) 02:34, 31 March 2020 (UTC)
 * Hebrew and Arabic are written in abjads (or impure abjads), in contrast to most or all Indo-European languages that are written in "normal" alphabets.
 * Hebrew and Arabic have verbs with typically three root letters, in contrast to Indo-European languages, which have roots of arbitrary length.
 * Hebrew has binyanim, and Arabic has something analogous, which have mostly predictable meaning shifts such as intensification, passivity, causality, and reflexiveness; there is nothing analogous to binyanim in Indo-European languages.
 * Hebrew and Arabic have certain patterns by which verb roots often become nouns. Indo-European languages also have patterns for making verbs into nouns, but they work completely differently.
 * Hebrew has only a very small number of verb tenses (past, present, future, and imperative), in contrast to the much larger number of verb tenses of most Indo-European languages.
 * Hebrew and Arabic lack the present tense of the verb "to be".
 * Hebrew lacks a verb for "to have" and uses a workaround to express possession.
 * Hebrew lacks a word for "same" and coaxes the "et" indicator of a definite direct object to denote sameness.
 * Short answer: none of these points are common to all branches of Afroasiatic languages. They may be appropriate to the article on Semitic languages. Landroving Linguist (talk) 07:37, 31 March 2020 (UTC)
 * Agree with Landroving Linguist. I will comment hwoever that there is nothing "normal" about Indo-European alphabets. The Indo-European alphabets are derived from the Phoenician alphabet (or abjad) which was a Semitic language. In many Semitic systems vowels are usually not needed as they are easily inferred.--Bob not snob (talk) 06:47, 2 April 2020 (UTC)

Map of Urheimat
User:Wadaad has repeatedly removed the map illustrating one of the hypotheses for the Afroasiatic Urheimat, giving the following rationales in the edit summaries: "Removed misleading map, the Omo Valley was populated by Mota-like HGs of the E-M329 subclade not found elsewhere in Afro-Asiatic communities. Afro-Asiatic in the Horn has been suggested to come from a more northern population (Egypt, North Sudan, Red Sea see 10.1126/science.aaw6275", "Unlikely Urheimat, extremely erroneous map" and "It is extremely unlikely given the many latest genetic studies + moreover, there are many AA urheimat theories, it is not a settled matter, therefore this map is not encyclopedic." But it needs to be noted that a) this map is based on a reliable secondary source which is quoted in the map's metadata, and b) that Wadaad so far has given no better reason for this removal than that he thinks that the map is erroneous, misleading and unencyclopedic. Such statements, going against a reliable secondary source, amount to Original research, particularly as the evidence brought against this map is Wadaad's personal opinion that bio-genetic evidence trumps linguistic evidence when it comes to the history of language families. I hereby declare that I have no strong opinion on whether Blench's Urheimat hypothesis is more likely to be true than other researchers' hypotheses, but on a topic such as African language history a hypothesis by Blench, one of the leading figures in the field, has as much weight as any other hypothesis, and there is no reason to remove material that is based on that hypothesis, as it does not violate any of Wikipedia's policies - it is certainly not unencyclopedic. The article of course would be better served by balancing the map with other maps illustrating the other hypotheses, but as long as these do not exist, the Urheimat map based on Blench will have to stand alone. With this understanding I will now re-insert the map into the article, and I call on Wadaad to not remove it without further discussing it here. To just state that he thinks that Blench must be wrong is not sufficient for removing this clearly sourced material. Landroving Linguist (talk) 21:19, 21 August 2019 (UTC)
 * I totally agree with you that the edit summaries by are characterized by giving emphasis to a single POV, and bordering to OR when it comes to his opinion that "bio-genetic evidence trumps linguistic evidence" as you aptly summarize it. And the map is definitely not "unencyclopedic". Thanks for restoring the "stable version" and bringing this to the talk page.
 * Yet, I also have doubts about whether the map should be kept. If had written in his edit summary "a single map gives too much emphasis on one out of two hypotheses (WP:NPOV)", we would have little to argue in defense of keeping the map, in spite of its crystal-clear legend "Map showing one of the proposed Afroasiatic Urheimat (Eastern Sahara theory.)". Maps are iconic and stick to the readers memory, so just having one map is imbalanced and indeed somehow misleading. So even though is is still a discussion and not a vote, I would suggest to delete the map unless it is complemented by another map illustrating the competing hypothesis. – Austronesier (talk) 05:33, 22 August 2019 (UTC)
 * Thanks for weighing in! Yes, I agree that the map should be balanced by maps illustrating other hypotheses, but as long as these maps are not created (and I'm not in the habit of creating maps) I believe that the inclusion of this map is defendable as a visual illustration of the concept of Afroasiatic Urheimat - by displaying one of the several hypotheses floating around in the academic discussion of the subject. As you say, the caption of the map makes it very clear that there are other ideas. Then again, if there is a consensus that the map should stay out, I will not resist it. Landroving Linguist (talk) 06:31, 22 August 2019 (UTC)
 * & genetic evidence is important as it reveals more fine-grained information than linguistics can ever achieve. For example, recently the Indo-European Urheimat (linked the genetics section of that page) has been identified through the help of ancient DNA and modern population genetics (bolstering the Kurgan hypothesis (Ukraine) and dismissing the Anatolian one). So, I would not dismiss this claim so easily. About why I think this map is very misleading and extremely unlikely: Haplogroup E-M329 is prominent in Southwest Ethiopia where this map points to as the so-called urheimat of the Afrasian language family, yet this tens-of-thousands of year old paternal lineage (much older than Afro-Asiatic language family) is largely restricted to Southwest Ethiopia and not found anywhere else (excluding a handful of Peninsular Arabians who may have acquired it through the Arab slave trade). This clearly dismisses the claim that Southwest Ethiopia can be the urheimat, possibly even Ethiopia at large. Secondly, ancient DNA from Southwest Ethiopia reveals that populations existed there only 4,500 years ago that completely lacked Afrasian related autosomal components, such as the ancient sample from the Mota cave. This strongly points to Southwest Ethiopia not being the Afrasian urheimat. Additionally, more recently an ancient genetics study that sampled likely ancient South Cushitic pastoralists from Kenya and North Tanzania showed that they carry Northeast African/Red Sea genetic markers and autosomal components. The researchers from that study point to a northern population movement. Their paternal haplogroup E-V1515 originated along the Red Sea coast and spread southwards more recently long after the creation of the Afro-Asiatic language family. Lastly, a population genetics study focused on the Horn of Africa shows an epipaleolithic back-migration from the Middle East to the Horn region affecting most of the Afro-Asiatic speaking populations there. The Ethio-Somali autosomal component in that study was found to be related to the Maghrebi autosomal component and the researchers suggested a population split somewhere in Egypt. This all strongly points to a more northerly origin for the Afro-Asiatic language family (likely the Red Sea region - anywhere between Egypt and Eritrea). Wadaad (talk) 17:38, 23 August 2019 (UTC)


 * This is not a forum. We could discuss at length about which method is the most adequate to learn about prehistorical population movements. Genes, languages, and material culture each spread in different ways; pots don't talk, genes don't talk, languages are not passed on by genetic inheritance, traded goods can move around for thousands of kilometers with little actual population movement involved. Input from each discipline needs to be considered to solve the puzzle.
 * But again, this is not a forum. You need not convince us about your POV, neither should I convince you to share mine. Out task here is not to prove which hypothesis or method is the best, most adequate, or whatever. Our task is to present the consensus view about a subject matter from the current state of research. And if there is no consensus, as in this case, we have to present a balanced overview of the currently existing proposals. No cherry-picking, no personal preferences. Even if your preferred POV happens to be the most plausible and most appropriate, wait for peer-consensus to settle the matter. And of course: no synthesis. E.g. articles A and B which do not explicitly address the Afroasiatic question do not belong here, even if they happen to discuss the same genetic signatures as articles C and D, which explicitly identify these genetic signatures with (Proto-)Afroasiatic speakers.
 * So only the question is: does this map adequately represent the variety of current Urheimat proposals? As I have said before, visually presented information has a strong impact on our readers, so a visual presentation has to be as neutral, inclusive, and balanced as possible. That said, I would welcome input from other editors, too. Maybe someone could even volunteer and draw the other maps; like, I have to admit that I am not good at creating maps either :) – Austronesier (talk) 19:57, 23 August 2019 (UTC)
 * about the consensus issue, the problem is there are many competing AA urheimat theories and there is no clear consensus on the matter. Some support North Africa, others the Red Sea, others Ethiopia, and some even the Levant or the Middle East. I provided you genetic evidence why Southwest Ethiopia in particular can be pretty much ruled out. This map provides unduly emphasis on the Southwest Ethiopian hypothesis, which I find to be both misleading and not encyclopedic. When I get the time I might look into updating it using other sources like Christopher Ehret whose AA urheimat is more realistic. Wadaad (talk) 04:33, 24 August 2019 (UTC)
 * It appears as if you are not getting Austronesier's main point. None of the four articles you quote come to the conclusions you present here; only article 4 talks about the Afroasiatic Urheimat and allows for it to be located in the Horn of Africa, among other places. So your judgment that "Southwest Ethiopia in particular can be ruled out" sounds like a strong statement, but it appears as if there is no academic source actually sharing your view in these strong terms. In other words, the evidence that you shared here, impressive as it sounds, amounts to original research, and that cannot guide us to change the content of wikipedia pages. There are countless pages on Wikipedia that I know to contain incorrect information, but since I don't have a reliable secondary source to back up my knowledge, I refrain from changing these pages, as all edits must only be guided by such sources. Again, I appreciate that in the end your conclusions may prove to be true (I really don't have a strong opinion about the Afroasiatic Urheimat). But if the HOA-hypothesis could really be so easily ruled out, then Blench would probably not have proposed it, or would have retracted it by now. In the same way, just as Ehret has many followers, there are other just as credible linguists who don't agree with his statements. As Austronesier says, the academic debate is still going on, and it is not for us Wikipedia editors to act as if this debate is already settled, or to even settle it in advance. Landroving Linguist (talk) 21:20, 24 August 2019 (UTC)
 * The issue is not with HOA hypotheses in general (the southwestern Red Sea coast is in any case part of the HOA!), but with Blench's hypothesis in particular, which focuses on southwest Ethiopia as the AA Urheimat. Blench's position appears to be an outlier, if not wholly disproven (I might add that Blench is a respected linguist, but his area of expertise is the Niger-Congo language family, not Afro-Asiatic)...and yet the article chooses to show a map for a proposed AA Urheimat only based on Blench's work, to the exclusion of others. The map appears to be unrepresentative of the most widely held positions, and in that sense it is misleading.
 * Just to recap, the article states: "The main theories of Urheimat are the Levant/Fertile Crescent,[44][45][46] the Eastern Sahara,[47][48][49][50][51] North Africa and the Horn of Africa."
 * Jacob D (talk) 14:07, 16 September 2020 (UTC)Jacob D

Citation style
please be specific. What is unclear about the citation style? – Finnusertop (talk ⋅ contribs) 22:01, 1 April 2021 (UTC)
 * Happy to help! So, started with External links: external links should be just that: links. They should just be websites that may be interesting to someone researching the topic, and should be formatted that way as well. Many of these, however, are formatted as sources, and likely even provided some of the information in the article. Links that do not comply with the policy at WP:EL should be discarded. "Links" that have actually been consulted in the creation of the article should be included as inline citations. If information from the links were used in the article, but it's not clear how, they should be migrated to the "general references" section. Now, General references: General references have their place, but in many cases in an unfortunate result of editors not properly properly creating inline citations. This General references section is lengthy, and it would be ideal if someone could go through them and see if they actually do substantiate information in the article or not. If they don't, they should be removed or added to a "further reading" section. If so, someone should attempt to convert them to inline citations. If they can't, then they can remain in the section, but should be consistently formatted (they currently aren't). The policy WP:GENREF can provide guidance on this as well. And finally, the Citations section (which corresponds to the Works cited section). Honestly, this is such an undertaking, I would prefer to just start working on it myself. But, if you or anyone else would like to join in, WP:INCITE is useful.--Esprit15d • talk • contribs 13:06, 28 April 2021 (UTC)

shared vocabulary
In the second table of shared vocabulary, appears the proto-word *siwan- (no. 25) which means know and it's written as it doesn't exists in the semitic family. But in Hebrew there is a word "Shinun" (שינון), root: SH.N.N (tr. repetition, review; memorization), maybe it does should fit in? Lilijuros (talk) 20:53, 24 December 2021 (UTC)

Hamito-Semitic
James P. Allen uses “Hamito-Semitic” in all of his books on ancient Egyptian that I’ve seen, all of which have been published since 2000 I believe - so why does this article’s section on terminology suggest that it is no longer current except in “some European universities” and the Indian census? (Why would the Indian census use that term anyway?)—Ermenrich (talk) 13:33, 7 December 2021 (UTC)
 * I correct my previous statement: I can't find "Hamito-Semitic" outside of one work by Allen: These peculiarities identify Egyptian as a distinct branch within the HamitoSemitic language family, with no close relatives of its own – perhaps, therefore, closer to the common ancestor of Hamito-Semitic than to either of the other two branches. (Allen, James P.. The Ancient Egyptian Language (p. 2). Cambridge University Press. Kindle Edition, 2013.) In his most recent book, he still refers to Proto-Hamitic, explaining Hamitic for non-Semitic Afro-Asiatic languages. (Allen, James P.. Ancient Egyptian Phonology (p. 77). Cambridge University Press. Kindle Edition, 2020.) Slightly less recently there's also the Hamito-Semitic Etymological Dictionary by Orel and Stolbova, 1994 . All of this suggests to me that the unsourced statements in the article about its current use are incorrect.--Ermenrich (talk) 14:01, 7 December 2021 (UTC)
 * Fair enough. I deleted the offending statement. In Wikipedia we don't have to hunt down and identify or even classify the sources that still use terminology seen as obsolete by most others in the field. LandLing 14:04, 7 December 2021 (UTC)
 * It's "Semito-Hamitic" in India, and indeed still used in the 2011 census. Looks like residual bureaucratic usage, since the term was also used in the 1961 census. –Austronesier (talk) 14:25, 7 December 2021 (UTC)
 * On this issue, my five cents: Hamitic refers to an older racial classification of humanity and an associated linguistic groupings, ultimately erroneous or in dispute. In addition, due to its association with colonial practices the term can be viewed as somewhat offensive, although that is certainly beside the point. From this discussion I can see that there are a few instances where it is still used, but it is widely viewed as outmoded, not least because there is not agreed-upon reason why non-Semitic languages would belong to one particular group. This is also explicitly stated in the sources now cited, e.g., Katzner and Hetzron.  This can then be stated explicitly, rather than that it is “sometimes also used” which indicates that the terms are exchangeable and in equal standing among the scientific community. The majority view in the research community is that Hamitic is outmoded. ~ Modainel (talk) 07:40, 30 March 2022 (UTC)


 * Be that as it may, the language family is still sometimes called Hamito-Semitic.—Ermenrich (talk) 12:40, 30 March 2022 (UTC)
 * I do not dispute that it is in use. The finer point here is that it is controversial and widely viewed as outmoded, for the above-mentioned reasons. Therefore one arguably might wish to avoid to present the term as neutrally exchangeable or a neutral equivalent to Afro-Asiatic. Words are sometimes not just words, but imply different views, even worldviews. I'd suggest to acknowledge that it is still in use, along the lines of something like: "Afro-Asiatic replaces earlier use of Hamito-Semitic, Semito-Hamitic, or Erythrean, although these are still sometimes in use." ~ Modainel (talk) 14:29, 30 March 2022 (UTC)

Afroasiatic languages: A point in part
As pointed earlier based on research and knowledge of a couple of ancient languages that belong to the Afroasiatic or Hamitic-Semitic  in general from Western Asia to the Easter region of Africa including the Horn of Africa seem to have been influenced and benefitted from language developments in the mentioned regions over the period of great civilizations. Over the centuries branching off to develop their own distinct forms (perhaps because of isolation from the others) which in turn seems to continue to branch off as well while retaining some of the old, thus the similarities or direct presence of ancient characters in their respective languages, This is the observation of a non-(formally trained) linguist, non the less is knowledgeable enough of two ancient languages hence, conducted research as an Independent Researcher for years. As in some Hamitic-Semitic languages that have phonetic system, the representative character, symbol or sigh has to be read. For example, O49 in Gardiner’s Sign List, “niwt” and H8 “egg…”, often appear at the end of a word, phrase or sentence. During my research and reading of texts from Ancient Egypt, reading in in hieroglyphics when each sign was read according to the already established alpha-sound, in transliterated words, often, then, (my research was consistently leading to:) niwt (as two words) that means “it is far” and “niw” or “neuw” that translates to far as one word, including in Ancient Egyptian Language) in one of the Semitic-Hamitic language in one and, “ is”  from the Verb to be in a related but distinct language. Similarly, “KMT niw “ translates to, “ let me estimate /measure far”  when preceded with “look or look at me when I measure far or estimate far” but, when instead of the soft “k/K”, the writing was with the stronger or emphatic “Q/q” and so was rhe “T/t” then the phrase “QmT niw” would mean, “it is a settled one” or,” It is put aside”. In this example, the new/neuw is the “is” from the Verb to be used for a masculine singular (n/s). In the related language the H8 from Gardiner’s sign list when it appears at the end of the word, it translates to, “it is” as well for a feminine singular (f/s). For example, “netsa means free while netsa nt” would mean “she is free (as two words) and, “netsa t/netsa net means freedom. Any word could be written in any script but the true difference could be in the meaning hence the differences and the need to analyze each word within the context in which it presents itself. The ancients used their signs and symbols when called for for their sounds thus can appear anywhere within a word, as we do in the spelling of any word using the 26 alphabets in an orderly combination. In the concentration of my research, as a rule, their sentences used to end in a form from a Verb that was tied to: “mainly, the Numbers; Pronouns, and the Gender within a sentence.”(excerpts from my research ). In my I research, and reading of some texts from ancient Egypt’s rich  history, I had come across words on noticed the ancient Egyptians were using H8 and O49.for their words and within their texts in general, and the similarities are striking..(Alphalang). Alphalang (talk) 12:42, 21 May 2022 (UTC)


 * I'm not exactly sure what your point is, but the key-phrase in your contribution is "in my research." If you have the wish to change anything on this page based on the results of your research, you can't, as this would constitute original research. Any change to the content of this (or any other) page needs to be based on reliable published sources, and you need to present these sources together with the change. LandLing 07:10, 22 May 2022 (UTC)

The shared vocabulary section
I'm uncertain of the current presentation in the "Shared vocabulary" section. Ehret and Orel/Stolbova have both been heavily criticized, for one thing. A more honest approach would be to give the lemma as reconstructed in both dictionaries, at least.

I'm also unclear where our supposed list of "fragile scholarly consensus" etymologies is coming from.--Ermenrich (talk) 22:53, 14 February 2023 (UTC)
 * I would add: I think adding reconstructions like those found in Güldemann of the pronouns (essentially as consonants) would also be better than our current practice.--Ermenrich (talk) 22:57, 14 February 2023 (UTC)
 * I've made a first attempt at how I would fix this here. There are already two large problems:
 * Orel/Stolbova do not include pronouns, for whatever reason. I've supplements by using Sasse 1981 as cited in Güldemann, but these more properly belong to morphology (and don't include the lemmas for where). The lemmas for "what" are included in e.g. The Egyptian Etymological Dictionary, but this does not include reconstructions, only cognates.
 * Ehret is rather notorious for his position that, if a root has 2 consonants in common with another root, he can prove the any additional consonants/vowels to be affixes. We thus have a case with his root for "to lick" where he uses nouns meaning "tongue" to deduce it, whereas Orel/Stolbova deduce two roots, one for "to lick" and one for "tongue" and merely suggest they could be related.
 * Ehret has also been criticized for taking dubious words for cognates. I highly doubt that other scholars regard words such as Egyptian jdmj (red linen) as related to the AA root for blood. It's not included in Orel/Stolbova, for instance. (turns out this is in Takacs, although he keeps the relationship more loose than Ehret). I think a good example can be found with the cognates of "brother" given: because Ehret reconstructs a lemma "its" and assumes that the -n is a suffix, (and also has some rather idiosyncratic sound laws), he connects Hebrew "aẖ" and Berber "uma" to this root. Orel/Stolbova are also noted for making some rather wild comparisons, of course. This leads to the question of how we should present cognates: do we need to list who said they were cognate?
 * There are still I think two hundred lemmas they share that could be compared on the table (and compare also our "fragile consensus").--Ermenrich (talk) 14:28, 15 February 2023 (UTC)
 * I don't think a full display of the shared vocabulary is called for. It would be more encyclopaedic to provide a list of good examples (the size of which shouldn't be any larger than what we have now) to illustrate the relationships between the branches, and then leave it to the readers to go to the sources if they want to see more.
 * If you conclude that some of the other items among the 200 lemmas would be less controversial than what is currently shown (because all authors agree on them, including Sasse and Güldemann), then I think you could and should include these instead of the more offending ones found in the current list.
 * And yes, we should definitely have a few pronouns, even if Orel/Stolbova don't include them. LandLing 08:47, 16 February 2023 (UTC)
 * Points taken. There seems to be general agreement on the Afroasiatic correspondences between the (non-ejective) labial consonants (*b, *p, *f, *m), and, to a lesser extent, also *r, *l, *n, *glottal stop, *ayin, *t, *d, *k, and *g. The fricatives and ejectives are most of what's tripping things up, as even a cursory comparison of the proposed correspondences between Orel/Stolbova, Ehret, and Takacs shows. This means that most of the good matches don't include these consonants (or if they do it's because the form is so well documented, like third person s-). And then vowels are a huge issue: Takacs makes no effort to reconstruct them, while Orel/Stolbova and Ehret come to different conclusions about them. I will try to find some more widely attested and agreed upon words like "p-r" (to fly) and remove those where the reconstructions lead to cognates that vary too much.
 * Probably it would also be more interesting to our readers to use words in attested languages rather than reconstructed proto-forms for the branches where possible. I'll try to do that too.--Ermenrich (talk) 15:34, 16 February 2023 (UTC)
 * Yes, I agree about the preference of attested languages over reconstructed languages. Saves you a lot of trouble. LandLing 16:20, 16 February 2023 (UTC)
 * I decided to split off the pronoun section and do a more straightforward comparison of all forms with a given meaning, whether they can be proven to be related or not, using Gragg 2019 as a guide.
 * I'm trying to get my hands on some more sources to display uncontroversial cognates between the languages.--Ermenrich (talk) 21:20, 17 February 2023 (UTC)
 * Looks good.--Andrew Lancaster (talk) 16:31, 19 February 2023 (UTC)

Agree, looks great! Still needs some cosmetics (e.g. be bold, but use bold?) here and there, but that a minor thing. –Austronesier (talk) 20:19, 19 February 2023 (UTC)
 * The bold there is left over from the cognates table - I'll get rid of it.--Ermenrich (talk) 20:52, 19 February 2023 (UTC)
 * Any chance someone can add more of the meanings of the example words?--Andrew Lancaster (talk) 15:53, 20 February 2023 (UTC)
 * Do you mean in the tables I've added? Unfortunately, the sources don't always say, but it should be possible to find out most of them if we do some digging... I can add a few already.-Ermenrich (talk) 15:57, 20 February 2023 (UTC)

I've gone ahead and replaced the table with a new one with ten cognates that are "widely accepted" according to Gragg and Hayward - I added the 7 that Gragg included (with a winnowing down of the proposed cognates for both etymological dictionaries to the most widely accepted) and added three more from Hayward. Hayward includes a further 8 we could consider adding, including lexemes for "to not be there," "sole," "to bloom", "to enlarge", "man", and "house".--Ermenrich (talk) 18:36, 22 February 2023 (UTC)

Dating and Urheimat
I've moved on (or am also working on) rewriting the section on the dating of Afroasiatic and the Urheimat. See my sandbox for the current (still rough) draft. Ermenrich (talk) 14:29, 24 February 2023 (UTC)
 * Keep in mind that there is a sub-article for the urheimat.--Andrew Lancaster (talk) 19:57, 24 February 2023 (UTC)
 * Indeed - one of my goals is to make the section(s) here shorter and more coherent. It looks to me like that other article needs some work in some areas, but I'm not ready to throw myself into that at the moment.--Ermenrich (talk) 20:06, 24 February 2023 (UTC)
 * The other article is in a not so good, but also not so bad shape. Only the "Genetics"-section is a bit of a trainwreck full of WP:SYNTH (as usual). We had a discussion in Talk:Afroasiatic_homeland and largely agreed that something must be done, but with little practical results (as usual too LOL). Apart from the opening rant, the gist of my proposal for a way to go is here. –Austronesier (talk) 21:05, 24 February 2023 (UTC)
 * Good stuff, - I managed to find a quote by Frajzyngier (2012) that will satisfy RS/AC that "most linguists" support the African origin, with Militarev being the main exception (one source [maybe Frajzyngier, can't remember at the moment] refers to him as a "one of the rare supporters"). I think two relatively pithy paragraphs on each side in the question question should flesh it all out here. On dating there's probably even less to say, except to add that not everyone agrees that AA has to be "the oldest language family in the world".--Ermenrich (talk) 21:14, 24 February 2023 (UTC)
 * What do people feel about the current expanded draft - it's three paragraphs rather than two, but I think it's still more concise than the current version. Am I missing anything important? (I don't mention genetics at all.) Do I go into too much detail anywhere?
 * I don't have much to add on dating and honestly, I'm not sure we need more (I know I've seen someone question whether we really need 4000 years of development between the first attestations of Egyptian and Semitic somewhere, but I can't find it and that probably indicates that it's not a widespread criticism). Found it, it was Güldemann - probably worth mentioning.--Ermenrich (talk) 15:59, 25 February 2023 (UTC)
 * The draft looks great. We can think about adding a sentence or two about hypotheses that seek correlations between the spread of languages and demic dispersal as evidenced by genetic markers, but we should TNT-cleanup the main article first. I could imagine mentioning here Ehret et al. (2004) as an interdisciplinary work; plus Lancaster (2009) and Hodgson et al. (2014), as both have been cited in linguistic works. –Austronesier (talk) 20:08, 26 February 2023 (UTC)
 * I've tried to incorporate a paragraph on genetics - if someone could look it over I would appreciate that. Or did you mean at the other article that needs to be TNT'd? I might prefer that. I'll move over the section without it now.--Ermenrich (talk) 00:34, 27 February 2023 (UTC)

Subgrouping
I'm trying to fix up the subgrouping section now and I'm a bit unsure how to precede. As there is no agreement, there aren't really even "schools" on subgrouping, just various proposals. A few may be more common than others, but as it is, any scholar can proposal a subgrouping, which means that the section can simply be extended infinitely for every time any scholar states anything about two groups being related. The article currently cites one that was proposed in a footnote to a printed lecture (Newman 1980).

There must be some way of limiting this. The main proposals seem to be Diakonoff, Ehret, and Orel/Stolbova (vs. Cohen and Greenberg who argue for five/six coequal branches). No proposal has any real support that I can see except Cohen/Greenberg, which is essentially an agnostic position (and which, strangely, is the only one that Wikimedia commons has no image to illustrate).

Anyone have any ideas?--Ermenrich (talk) 15:23, 27 February 2023 (UTC)
 * I'll note: Carsten Peust counts 27 proposals total; Güldemann discusses three of them Fleming, Bender, and Ehret. Perhaps it would behoove us to limit ourselves to those three, perhaps adding Diakanoff as discussed or mentioned in passing in other sources? Also: this paper (which we can't cite) discussed Ehret, Diakonoff, and Blench (2001).--Ermenrich (talk) 16:46, 27 February 2023 (UTC)
 * I've outlined five studies that seem to get discussed in other sources - I'm waiting to get my hands on some of the sources still to cite them more fully. I understand that the book coming out March 14 of this year will also included discussion of this problem.--Ermenrich (talk) 15:18, 28 February 2023 (UTC)

Thinking about it some more, I wonder if it really makes sense to discuss any of the individual proposals. Maybe just summarizing a few trends and then the problems with attempts to-date would suffice.--Ermenrich (talk) 14:40, 4 March 2023 (UTC)

New book coming out
The other day I stumbled across this book: https://www.amazon.com/Ancient-Egyptian-Afroasiatic-Rethinking-Languages/dp/1646022122/ref=sr_1_1?crid=VUAUS6WL4JQ8&keywords=ancient+languages+of+the+near+east&qid=1671287206&sprefix=ancient+languages+of+the+near+east%2Caps%2C142&sr=8-1

Once it's out (2023), it seems like the chapters will be a lot of use for this article.--Ermenrich (talk) 14:28, 17 December 2022 (UTC)
 * Here's the publisher link:. The list of authors is impressive. –Austronesier (talk) 14:59, 17 December 2022 (UTC)
 * Still haven't been able to get my hands on the book, but the table of contents shows that it will be very useful for this article.
 * Comparative Afroasiatic linguistics and the place of ancient Egyptian within the phylum / M. Victoria Almansa-Villatoro and Silvia Štubňová Nigrelli
 * Data limitations and supplementary methods in placing Egyptian / Chelsea Sanker
 * Afroasiatic lexical comparison : an Egyptologist's point of view / Jean Winand
 * Egyptian morphology in Afroasiatic perspective / Andréas Stauder
 * Proto-Semitic and Egyptian / John Huehnergard
 * Some common features of Akkadian and Egyptian revisited / Elsa Oréal
 * Lexical, phonological, and morphological evidence / Aren Wilson-Wright
 * Reconstructing Proto-Semitic nominal and verbal systems in the context of Afroasiatic languages / Vit Bubenik
 * Ancient Egyptian's place in the Afroasiatic language family / Christopher Ehret, Deven N. Vyas, Shiferaw Assefa, J. Lafayette Gaston, Tiffany Gleason, and Andrew Kitchen
 * Testing the hypothesis -- theoretical and methodological issues : the relationship of Egyptian and Afroasiatic / Zygmunt Frajzyngier and Michael Avina
 * Restructured or archaic? : the hunt for shared morphological innovation involving Egyptian / Lameen Souag
 * In Pascal's and Boole's footsteps : measuring the mathematical probability of genetic kinship between language families / Leo Depuydt.
 * If we're lucky we'll get some statements of academic consensus. One problem with a field like this is that individual authors too often contradict each other and thus make a presentation here difficult.--Ermenrich (talk) 13:57, 22 March 2023 (UTC)

One ultimate source, two homelands?
The current version of "Origins" says: The source for "Northern Africa" is Scheinfeldt, Soi & Tishkoff (2010), for "the southeastern Sahara or adjacent Horn of Africa" it's Ehret, Keita & Newman (2004). So far so good. But when I looked up the former for what they actually say, I found this: "Although the origins of the Afroasiatic language family remain contentious, linguistic data generally support a model in which the Afroasiatic language family arose in Northern Africa >10 kya (36)". Citing this for "Northern Africa" seems legit, right? But what is their source "(36)"? Oddly enough, it is Ehret, Keita & Newman (2004)!
 * Most scholars place the homeland of Afroasiatic near the center of its current distribution, in a region stretching from Northern Africa to "the southeastern Sahara or adjacent Horn of Africa."

Looks like we're artificially differentiating things that actually refer to the same thing. Austronesier (talk) 21:22, 22 March 2023 (UTC)


 * Yes, I agree. I’m going to revert the addition for that reason.—Ermenrich (talk) 21:50, 22 March 2023 (UTC)


 * It seems, if I am correct, Ehret, Keita & Newman (2004) based their location on haplogroup data: "A critical reading of the data in genetic analyses, specifically those of Y chromosome phylogeography and TaqI 49a,f haplotypes, supports the hypothesis of populations moving from the Horn or southeastern Sahara northwards to the Nile Valley, northwest Africa, the Levant, and Aegean (14,15). The geography of the M35/215 (or 215/M35) lineage, which is of Horn/East African origin, is largely concordant with the range of Afroasiatic." Another one, Kitchen et al. 2009, has another proposal based on 'Bayesian phylogenetic analysis', showing the homeland of Afroasiatic in the" Red Sea coastal region of Egypt (Figure 1.), but funnily enough also citing Ehret 1995, and 2004 ("An origin of Afroasiatic along the African coast of the Red Sea, supported by comparative analyses (Ehret 1995; Ehret et al. 2004)". Either way, perhaps just stating:
 * Most scholars place the homeland of Afroasiatic near the center of its current distribution, somewhere in Northeast Africa."
 * would be a better solution? If my first observation about Ehret 2004 is correct, than we should also take Hogdson et al. 2014 into account, as their findings are more recent than that about haplogroups. I do not object the revert, although I continue to argue that "an origin within Africa" is to be preferred over "an African origin". What do you think?Krause96 (talk) 22:05, 22 March 2023 (UTC)
 * I don't see why a less precise formulation like "northeast Africa" would be preferable to the more specific statement we currently quote (which is stated to be the communis opinio). I think you are drawing a difference where there isn't one, just minor differences in formulations. The Horn of Africa and the southeastern Sahara include the Red Sea coast, for instance, and both the southeastern Sahara and the Horn of Africa are in Northeast Africa.--Ermenrich (talk) 23:14, 22 March 2023 (UTC)
 * I know from past discussions about population genetics articles that some academics have used some of those broad terms in quite specific ways. From memory the Rome team who studied E-M35 for example tended to use NE Africa to refer to Egypt, and NOT the Horn of Africa. For some people I presume northern Africa will imply Mediterranean countries, while for others it will stretch to the Horn of Africa. These are just observations though, not proposals. Naming the adjacent seas, deserts or rivers might help avoid problems though.--Andrew Lancaster (talk) 05:42, 23 March 2023 (UTC)
 * I agree with Andrew Lancaster, the terminologies vary. Furthermore, is the 2004 paper representing the 'communis opinio', as stated by Ehret, Keita & Newman (2004)? Especially when younger papers make suggestions of the Red Sea coast in Egypt, the Sinai, and or Northern Africa. I think they all have a reason why they used these specific regions, rather than "the southeastern Sahara or adjacent Horn of Africa". All these regions would however fall in the 'Northeast African region', stretching from Egypt to the southeastern Sahara or adjacent Horn of Africa. Thus I think that the term 'Northeast African' is more reliable and practical, as it includes all these different regions and does not base the whole model on the formulation of one cite. The Red Sea coast of Egypt does not seem to be the 'southeastern Sahara', which more corresponds to Sudan and partially overlaps with proposals for the Nilo-Saharan* homeland. (*if indeed a valid family, but at least some of them would have be spoken there, proto-Eastern Sudanic etc). I would at least mention Egypt, specifically 'southeastern Egypt' as Ehret actually did too. Scheinfeldt et al. 2010 for example also mentioned the Kebaran culture and] Mushabian culture in the Sinai, as possibly associated with proto-Afroasiatic. And I repeat that the 2004 paper of Ehret (and co.) does make their statement based on than proposed haplogroup distribution, making the argument that this region is the most likely place, although not ruling out other regions, as their main point is to give a response/clarification why they believe the homeland is placed within Africa. Yet, the other papers which cite them do refer to slightly different regions, such as specifically Northern Africa, Egypt, Sinai. Do these papers become irrelevant? They do not cite the 'southeastern Sahara or adjacent Horn of Africa' but rather the model that Afroasiatic originated within Africa. Taking all this together, the 'communis opinio' in 2023 is hardly this narrowed region, but rather includes several regions in Northeast Africa (stretching from Egypt to the Southeastern Sahara). Perhaps the new book will give some clarification.Krause96 (talk) 06:43, 23 March 2023 (UTC)
 * Again, I see no reason to replace the specific formulation of Ehret et al. when they are the ones everyone else is citing. They state: [Diamond and Bellwood] fail to engage the five decades of Afroasiatic scholarship that rebutted this idea [of an Asian origin] in the first place. This extensive, well-grounded linguistic research places the Afroasiatic homeland in the southeastern Sahara or adjacent Horn of Africa. We are speaking here of linguistic, not genetic research. The fact that geneticists cite them and then change the formulation somewhat is therefore not really necessary for us to mention. The only change that I would propose to the article is perhaps changing "most scholars" to "most linguists" before that specific claim.
 * It seems to me, moreover, that the vagueness of "Northeastern Africa" has the added disadvantage of potentially implying that Proto-AA speakers might not have had a "Black" appearance (for, as AL notes, many will take it as implying the Mediterranean coast, which many consider to be "white"). This weakens the impact of the fact that the majority of scholars believe the language family originated in Africa and are agnostic about its "racial" appearance. This is not an area we should be wading into.--Ermenrich (talk) 16:00, 23 March 2023 (UTC)


 * I am not sure if I get your arguments, especially regarding the "vagueness of Northeastern Africa" and fringe concepts about "race" ("black" or "white" or their associations with certain regions). As you stated afterwards, linguists do not make statements about "racial appearance", neither do (serious) geneticists. Furthermore, geographic regions are not defining "racial appearance" either. The whole region, including the Horn of Africa, obviously had a vibrant history, as noted by Hogdson et al. 2014 and succeeding papers on African history. This argument does not seem related to the points I brought up and discussed here. My concern is that the narrow region of "southeastern Sahara or adjacent Horn of Africa" is not the sole proposal and neither the 'communis opinio'. It may have been in 2004, but probably not in 2023, as following papers did not use this regions specifically, but more broadly Northeast Africa, or specifically the 'Egyptian Red Sea coastal regions', 'Sinai', or 'Northern Africa'. As I understand, you argue that the succeeding papers citing Ehret 2004, and their regional proposals, are irrelevant. Not sure if this view is correct, but I get the point. We will have to wait for further publications which will perhaps clarify it. I will not object the current version if all others agree with it, especially as there currently is no new paper specifically favoring a certain region (although Kitchen et al. 2009 graphically argues for the Red Sea coastal region of Egypt). It's just a pity that we must rely on a 2004 paper IMHO. Anyway, thanks for your points.Krause96 (talk) 17:40, 23 March 2023 (UTC)