Talk:Pama–Nyungan languages

number of speakers
I'd like to see some information here about the number of native speakers of this 'language family' (the one described by Hale and disputed by Dixon) and also the languages with the highest number of native speakers   —Preceding unsigned comment added by 76.104.195.115 (talk) 06:20, 7 May 2011 (UTC)

Rewrite required
This article needs to be rewritten and expanded, but meanwhile I removed mentions of an Australian language family. Few Australian linguists subscribe to that idea anymore, and indeed R. M. W. Dixon no longer believes even Pama-Nyungan is valid. Although most Australianists hold to Pama-Nyungan, and many are looking for external links to other Australian families, traditional "Australian" should be relegated to macro-family hypotheses such as Nostratic.

Historically, Australian languages were first suggested to form a single family when only Pama-Nyungan were well enough described to provide much data. --kwami 02:24, 21 Mar 2005 (UTC)

It'd be great if someone rewrote this to make it a little less Dixon-focussed. There's a lot more research around these days that is better supported. Claire (talk) 19:46, 29 June 2009 (UTC)


 * This was stricken from the article itself, but about being Dixon-oriented, I believe I'm not trespassing here. So here comes the deleted passage - and I believe most would agree with me about Hale's chouice of words.


 * "The reader should however aquaint him or herself with Dixon's argument in his own words (Dixon 2002). To a layman at least, he makes a convincing argument, and Hale's over-the-top phrasing is definitely not the dispassionate rejoinder one should expect from a scientist. It sounds more like a man defending his religion. Furthermore, fifty basic cognates (note that the article is silent about how they sprang to being from his pen) are not very impressive, considering the latitude comparatists often allow themselves - and it should be mentioned that this comment comes from a firm comparatist. The method just doesn't always work given enough depth of time to obscure relationships. The time we are talking about here is longer than the time required by Nostraticists, and hardly anyone believes that theory. Dixon deserves a read, if only what his opponents are arguing against. For instance, he comes up with a perfectly plausible explanation about the drift of language features, given the 40000 years or so, that those features had available to drift. Hale's age estimate is not supported in this article anyway." All the best Io (talk) 19:25, 23 July 2010 (UTC)


 * PS: Medeis, you apparently watch this article like a hawk. Don't remove this, please - it's a perfectly good input for discussion. All the best Io (talk) 19:30, 23 July 2010 (UTC)

Please don't act as if my politely reverting your addition of your own personal comments, made in the improper form of a reference to the article, and suggesting that you find a quote of Dixon to place in the article instead, has anything to do with trying to prevent you from discussing anything. I am glad you have found the discussion page, which is indeed the proper place for your concerns. Why you think I would stoop to vandalizing the talk page by deleting your remarks is beyond me.

That being said, I have read Dixon. He does not offer any refutation of these specific forms. He offers a priori theories as to why comparison may fail in general. Hale's emotion was notable as the opinion of a professional and was based on his exasperation with this sort of contrarian skepticism. Dixon, frankly, sounds like the lawyers for O.J. Simpson or Amanda Knox, asking you not to consider the obvious, but to focus instead upon theories for which they offer no evidence but their own speculation. Hale's article was peer reviewed and no one doubts his scholarship. I suggest you actually read Hale's word list. I am sure your library can get the book referenced in the article. The cognates show repeated regular sound correspondences. There is no dubious "leeway" and everything is sourced and verifiable - not a creation "[sprung] to being from his pen."

Of course, so far as the article goes, my personal opinion is as worthless as yours. What is relevant is notable, reliable and verfiable sources. Hale fits that bill to a tee.μηδείς (talk) 21:26, 23 July 2010 (UTC)


 * Don't worry. I have come across people both on this Wikipedia and others who delete on sight. You were polite, and after your deletion I thought the discussion page was the place which it apparently is. The most compelling reason I find Dixon believable is the time span involved and the demonstrated speed by which cultural (he makes a special note of circumcision) innovation can spread. The same applies to language, allowing for a "wipe-out" of credible correspondences in a much shorter timespan than Australia has been inhabited. I haven't read Hale, but his quoted words were anything but trustworthy - more like spitting venom, and I doubt any library in my small country has copies of his works. Of course, he and Dixon may have had a previous history of enmity of which I am not aware. Spurious "correspondences" are actually fairly likely among ca. 300 languages and the number 50 isn't impossible to have arisen by chance - and yes, with apparent sound laws along with them. Give it a few thousand years. (I'm not contradicting myself here - I'm saying that words can spread, they can spread along with formation rules and grammatical features spread as well. For all we know we might have a number of Sprachbünde, and we have enough time to have it all realised.) Dixon himself is no anti-comparatist. He just found, at least to his satisfaction that the remaining traces to compare are not enough, and whether he is right or not, he is at least convincing. All the best Io (talk) 22:03, 23 July 2010 (UTC)


 * Do search out and read the source book, which includes serious work by many authors. Hale is rightly insulted by Dixon's implication that they are not doing real linguistics.   Hale does not provide spurious resemblances.  His has published his forms and the burden of proof is on his critics to show how his analysis is false, not to merely assert that analysis can be false.  After all, the assertion that any belief can be false also applies to itself. Hale also identifies forms which obviously are borrowings based on their irregular shape and their resemblance to a motivated form in a neighbouring language.  Hale lists basic vocabulary words, including uniquely sharted pronouns such as the first personal dual inclusive, found in multiple languages, from the Cape York peninsula to the West coast. While Australia has been inhabited for tens of milennia, he posits a shallower time depth for the spread of Pama-Nyungan,  The mere fact that words can diffuse is not proof in any specific case that a word has diffused.  A serious claim that borrowing has occured bears the burden of proof.  An argument that forms might have diffused from disparate original stocks which doesn't identify those original stocks is nothing more than speculation and about as worthy of serious consideration as O. J. Simpsons' pledge to search for the real killers.  In any case, without recourse to Hale's paper, this is baseless speculation.  I refer you to his work.μηδείς (talk) 22:43, 23 July 2010 (UTC)

Articles and groups?
How far do we go with creating new articles for groups, sub-groups, etc. Is there a Wikipedia standard for this? Eg: Pintupi is in the Western Desert sub-sub-group, which is in the Western Desert-Warnman sub-group, which is in the South-west group, which is part of the Pama-Nyungan family. Do we have to have an article for every group? Dougg 05:45, 19 September 2005 (UTC)


 * I don't see why we couldn't do more subgroups and have articles on them. I guess it is just who wants to write them.  Since some subgroups are different according to the linguist, I have kept it rather neutral and just listed the major groupings and the languages.  I still have more languages to add. Imperial78

Specifics of Classification

 * I think it's pretty much agreed these days that Tangkic is non-Pama-Nyungan. Claire 02:56, 13 January 2006 (UTC)
 * Re Pama-Nyungan, Barry Alpher has a wonderful article on the evidence for Pama-Nyungan as a cohesive subgroup. It's in an incredibly expensive book but if any of the editors of this section would like a copy send me an email and I'll send you a pdf - I'm one of the editors of the volume so I'm depriving noone except myself and my coauthor of royalties :) It's in Bowern and Koch (2004) Australian languages: Classification and the Comparative Method. John Benjamins. Claire 02:56, 13 January 2006 (UTC)

Spelling issues
To what extent do you want to use established community orthographies? Yuulngu as you have it has had community literacy for about 50 years now, and it's Yolŋu in the estiablished orthography. I'd be in favour of giving priority to established community spellings as well as giving variants, if that's ok? Claire 02:57, 13 January 2006 (UTC)

more article links
There are a fair number of language stubs we don't have links to, which can be found at Category:Australian_Aboriginal_languages and Category:Indigenous_Australian_language_stubs. Some of those articles should be moved to "X language". Arrernte has a fair amount of language data too. Sorry, but I'm operating on only 64MB of memory at the moment (an archaic computer that's the only one functioning), which is way too slow to go through all these myself. kwami 10:37, 5 April 2006 (UTC)

How Well Established is Pama-Nyungan
Is this language super-group a hunch based on some shared characteristics and a driving desire to find order in randomness... or is there a fully reconstructed proto-language with specific sets of transformations leading to each of the present day language groups?

Could you present a little bit of the phonology, lexicology, and grammar that lets you describe Pama-Nyungan as fact and not (vague?) theory in this article?


 * They are all based on the pro-nouns that are similar to some extent (Blake 1992), this is the proto-form of the overall language family.

The table below is a reconstruction of the original language that is proto-Pama-Nyungan. 1st person is 1, 2nd person is you and 3rd person is they.

nyan (she) I cant place. Enlil Ninlil 05:13, 14 December 2006 (UTC)


 * The forms above do not match other sources. O’Grady and Hale ("The coherence and distinctiveness of the Pama-Nyungan language family within the Australian linguistic phylum", Bowern and Koch, 2004) give, inter alia, 1ps *ngayu, 1pd *ngali.

μηδείς (talk) 22:14, 25 July 2010 (UTC)

Redirects
Should Non-Pama-Nyungan really redirect here? I suggest it should be a separate article outlining why it's even used as a psuedo-natural class, and use it basically as a linking page to all the non-Pama-Nyungan families as well as Pama-Nyungan proper.

Aidhoss 04:01, 6 November 2007 (UTC)

Hale vs Dixon
(I have copied this from above to place it at the bottom of the page since it is the most recent dialog.μηδείς (talk) 23:08, 23 July 2010 (UTC))


 * This was stricken from the article itself, but about being Dixon-oriented, I believe I'm not trespassing here. So here comes the deleted passage - and I believe most would agree with me about Hale's chouice of words.


 * "The reader should however aquaint him or herself with Dixon's argument in his own words (Dixon 2002). To a layman at least, he makes a convincing argument, and Hale's over-the-top phrasing is definitely not the dispassionate rejoinder one should expect from a scientist. It sounds more like a man defending his religion. Furthermore, fifty basic cognates (note that the article is silent about how they sprang to being from his pen) are not very impressive, considering the latitude comparatists often allow themselves - and it should be mentioned that this comment comes from a firm comparatist. The method just doesn't always work given enough depth of time to obscure relationships. The time we are talking about here is longer than the time required by Nostraticists, and hardly anyone believes that theory. Dixon deserves a read, if only what his opponents are arguing against. For instance, he comes up with a perfectly plausible explanation about the drift of language features, given the 40000 years or so, that those features had available to drift. Hale's age estimate is not supported in this article anyway." All the best Io (talk) 19:25, 23 July 2010 (UTC)


 * PS: Medeis, you apparently watch this article like a hawk. Don't remove this, please - it's a perfectly good input for discussion. All the best Io (talk) 19:30, 23 July 2010 (UTC)

Please don't act as if my politely reverting your addition of your own personal comments, made in the improper form of a reference to the article, and suggesting that you find a quote of Dixon to place in the article instead, has anything to do with trying to prevent you from discussing anything. I am glad you have found the discussion page, which is indeed the proper place for your concerns. Why you think I would stoop to vandalizing the talk page by deleting your remarks is beyond me.

That being said, I have read Dixon. He does not offer any refutation of these specific forms. He offers a priori theories as to why comparison may fail in general. Hale's emotion was notable as the opinion of a professional and was based on his exasperation with this sort of contrarian skepticism. Dixon, frankly, sounds like the lawyers for O.J. Simpson or Amanda Knox, asking you not to consider the obvious, but to focus instead upon theories for which they offer no evidence but their own speculation. Hale's article was peer reviewed and no one doubts his scholarship. I suggest you actually read Hale's word list. I am sure your library can get the book referenced in the article. The cognates show repeated regular sound correspondences. There is no dubious "leeway" and everything is sourced and verifiable - not a creation "[sprung] to being from his pen."

Of course, so far as the article goes, my personal opinion is as worthless as yours. What is relevant is notable, reliable and verfiable sources. Hale fits that bill to a tee.μηδείς (talk) 21:26, 23 July 2010 (UTC)


 * Don't worry. I have come across people both on this Wikipedia and others who delete on sight. You were polite, and after your deletion I thought the discussion page was the place which it apparently is. The most compelling reason I find Dixon believable is the time span involved and the demonstrated speed by which cultural (he makes a special note of circumcision) innovation can spread. The same applies to language, allowing for a "wipe-out" of credible correspondences in a much shorter timespan than Australia has been inhabited. I haven't read Hale, but his quoted words were anything but trustworthy - more like spitting venom, and I doubt any library in my small country has copies of his works. Of course, he and Dixon may have had a previous history of enmity of which I am not aware. Spurious "correspondences" are actually fairly likely among ca. 300 languages and the number 50 isn't impossible to have arisen by chance - and yes, with apparent sound laws along with them. Give it a few thousand years. (I'm not contradicting myself here - I'm saying that words can spread, they can spread along with formation rules and grammatical features spread as well. For all we know we might have a number of Sprachbünde, and we have enough time to have it all realised.) Dixon himself is no anti-comparatist. He just found, at least to his satisfaction that the remaining traces to compare are not enough, and whether he is right or not, he is at least convincing. All the best Io (talk) 22:03, 23 July 2010 (UTC)


 * Do search out and read the source book, which includes serious work by many authors. Hale is rightly insulted by Dixon's implication that they are not doing real linguistics.   Hale does not provide spurious resemblances.  His has published his forms and the burden of proof is on his critics to show how his analysis is false, not to merely assert that analysis can be false.  After all, the assertion that any belief can be false also applies to itself. Hale also identifies forms which obviously are borrowings based on their irregular shape and their resemblance to a motivated form in a neighbouring language.  Hale lists basic vocabulary words, including uniquely sharted pronouns such as the first personal dual inclusive, found in multiple languages, from the Cape York peninsula to the West coast. While Australia has been inhabited for tens of milennia, he posits a shallower time depth for the spread of Pama-Nyungan,  The mere fact that words can diffuse is not proof in any specific case that a word has diffused.  A serious claim that borrowing has occured bears the burden of proof.  An argument that forms might have diffused from disparate original stocks which doesn't identify those original stocks is nothing more than speculation and about as worthy of serious consideration as O. J. Simpsons' pledge to search for the real killers.  In any case, without recourse to Hale's paper, this is baseless speculation.  I refer you to his work.μηδείς (talk) 22:43, 23 July 2010 (UTC)

Here is the book, Australian languages: classification and the comparative method By Claire Bowern, Harold James Koch at Google Books. Unfortunately the preview does not include Hale's paper, but you can see the general scholarly tone of the work.μηδείς (talk) 23:10, 23 July 2010 (UTC)


 * Thanks for the link. All the best Io (talk) 14:09, 24 July 2010 (UTC)
 * PS: I just read Dixon's appendix, The "Pama-Nyungan" Idea again. I can't find anything insulting there at least, and if he's right that the main tools of Pama-Nyunganists are mass comparison and glottochronology, then caution is in order. The first is dubious at best (you may be lucky or not, see Greenberg's African classification where he was and his Amerindian where he was not) and the second can't be taken seriously, although I admit that the original idea was good. It just didn't correspond to reality. And regarding mass comparison, reseachers who do resort to it may find a good place to start but that's about it. They undeniably often allow themselves a great deal of latitude, semantic and phonetic, just to use that word again. But maybe the link you gave me will convince me otherwise. All the best. Io (talk) 15:04, 24 July 2010 (UTC)
 * Just out of curiosity. Could Claire who called for less Dixon be the same Claire Bowern who coauthored the book which the link leads to? Add that to the fact that Pama-Nyuangan apparently is Hale's baby, and people are protective of their babies, I think Dixon, who is an accomplished and convinced comparatist in his own right (see The Amazonian Languages, coedited and cowritten by him) is bound to have an uphill struggle here. Anyway, the link you gave me sadly did not give me permission to read about Pama-Nyungan, and my book budget right now is zero in any currency. But altogether, both sides (Hale included) seem to have taken a religious approach to the whole thing. And, to quote Dixon from memory, he wanted to believe in Pama-Nyungan - he just couldn't make it fit and having read his side of the story, I'm still inclined to believe him. All the best. Io (talk) 16:49, 24 July 2010 (UTC)
 * Just two more things.


 * The quotations from Hale in the article "extravagantly and spectacularly erroneous", "wrong-headed" and "so bizarrely faulted, and such an insult to the eminently successful practitioners of Comparative Method Linguistics in Australia, that it positively demands a decisive repost." do not come from the pen of a dispassionate scientist. It's more like a spitting contest or in this case, a father protecting his child. This kind of phrasing should never see the light of day in respectable publications. I don't know if Dixon ever engaged in this kind of rhetoric but if he did, someone might point it out. All I have is the Cambridge Survey of Australian Languages, in itself a sign that he is not totally detested by the entire community - or maybe the series editors simply didn't know better when they released the book - after all, they might know nothing about linguistics and just have got the job by chance.


 * Another thing. How can you posit (as Hale did) a time depth for languages whose written descriptions reach back to the 19. century AD at best? To be more precise, he assigns an age comparative to the Indo-European languages, which is definitely stretching things very far, considering the documentation available for Indo-European languages and the lack of same for the Australian. Glottochronology run amok? All the best Io (talk) 17:24, 25 July 2010 (UTC)


 * Just because I like talking to myself, here's another point to consider. Armenian is attested from the 5. century on. It took until 1877 that Hübschmann was able to establish that Armenian is an Indo-European branch of its own, not Iranian. The inherited Armenian vocabulary (i.e. Indo-European non-borrowings) consists of about 450 words. With results like that for the most thoroughly researched language family in the world, and a language with an attestation like that, how is one to to trust lexicostatistics and glottochronology for a group of languages, not attested until Captain Cook (if he did take notes) and whose documentation even today is sparse? With every case I can think of, I'm leaning more towards accepting Dixon. (And, just as a reminder, "Silence is acceptance"?) All the best Io (talk) 18:04, 25 July 2010 (UTC)


 * Surrender accepted. May we never meet again. All the best Io (talk) 18:47, 25 July 2010 (UTC)

A case has not been refuted until it has been stated at its strongest. -Christopher Hitchens


 * That's just a phrase and goes to prove nothing. 85.220.123.77 (talk) 19:30, 29 April 2011 (UTC)

Hale's published work speaks for itself.


 * I'm sure it does, none too favourably. You never answered the questions about glottochronology and mass comparison. I suspect out of fear or for a lack of response. So just a final question, again: Were those Hale's methods? It's a very simple question, but perhaps you're too much of a medeis. 85.220.123.77 (talk) 19:30, 29 April 2011 (UTC)

You, however, protest too much.


 * Also just a phrase out of the blue. Somebody insisting on a point isn't necessarily wrong. (For instances, see any courtroom.) 85.220.123.77 (talk) 19:30, 29 April 2011 (UTC)

Your second-hand opinions on Hale, whom you have not read, based merely on your reading of Dixon's ad hominem remarks (Dixon does not address a single one of Hale's well-documented cognates) simply do not merit further comment.μηδείς (talk) 19:16, 25 July 2010 (UTC)
 * Just to ask again, since I seem to recall, that I was the one who taught you what an ad hominem is: Where does Dixon engage in that? Also a simple question. But you are right about Hale's comments not being ad hominem. They are just plain invective, which avoids a logical fallacy, but still is considered a mark of the man. 85.220.123.77 (talk) 19:30, 29 April 2011 (UTC)
 * I haven't seen an ad hominem in Dixon. You might tell me where. You have not refuted anything. And last of all, you have not taken note of your own addition (not a forum etc.). So since you haven't answered a single thing, I must still continue to assume that Hale, who has been proven to be venomous, at least in this respect, based his conclusion s on mass comparison and glottochronology. Those two methods are, after all, the only ones available for this undertaking. I would read Hale, but I can't get hold of him. (And yes, I know he's no longer with us, requiescat in pace.) I have read Dixon, and he presents a good case. But I mentioned the religious fervour earlier and you just sound like an angry acolyte. After all - arrogant though I may be (and that certainly applies to you), you haven't answered anything. 50 lucky strikes, given the freedom the reconstructor has in picking and choosing both sounds and semantics, having thousands of years to operate with does not make a convincing case. You might think of the Armenian example above. How, exactly, did Hale put the time depth of Pama-Nyungan in the vicinty of Indo-European? The only things available in this linguistic situation is guesswork based, perhaps on experience, but the only method, which has actually been proposed to assign actual time, i.e. glottochronology, is not taken seriously. In any case, with only A Survey of the Australian Languages to work with, where was Dixon supposed to find the space to pick the list apart? Io (talk) 19:34, 25 July 2010 (UTC)


 * Sorry, I meant within the confines of what I have. Dixon may very well have cursed like a sailor and all that, but that is not available to me. In the meantime I recommend


 * Aikhenwald & Dixon: Areal Diffusion and Genetic Inheritance all the best Io (talk) 19:56, 25 July 2010 (UTC)
 * Sigh! Little medeis strikes again. Grow up and learn to read. —Preceding unsigned comment added by 157.157.73.182 (talk) 19:50, 2 January 2011 (UTC)

hale's criticism
Please stop deleting notable verbatim criticisms from the article. Indeed, having an entire section devoted to Dixon borders on undue weight and Hale's quite notable comments serve to balance a section devoted to his objections alone. Reverence for Dixon is fine, and I suggest you find reliable sources to quote in praise of him in his article. The comments you describe as "ad hominem" are no such thing - ad hominem is the argument that because a person is evil or the like his arguments are invalid - no such claim is made here, the emotion of the comments does not make them ad hominem.

In the case of controversies, attributed verbatim quotes are relevant, which is exactly what we have here. "Forcefully" is an unsourced evaluation that you reflects your opinion of Hale's words - the critic's words speak for themselves.

μηδείς (talk) 21:19, 2 January 2011 (UTC) Æ, éttu það sem úti frýs. I won't be heard from again. —Preceding unsigned comment added by 157.157.73.182 (talk) 15:49, 3 January 2011 (UTC)

Is the hypothesis in such a sorry state that references have to be repeated?
See the reference to Evans and McConvell.

And apart from Hale's venom, is there any chance of evidence ever being presented on this page?

Just saying that one has used the --- gasp --- comparative method isn't particularly convincing.

And these hundreds of cognates, where are they?

The mainstream appears to consist of Evans, McConvell and Bowern, more or less, with "Nobody" doing the cleanup --- or mostly clicking the undo-link. Even Lyle Campbell, who should be pretty dispassionate about the subject calls for much more information and research and assigns Pama-Nyuangan the label of family status not confirmed.

And please don't let poor Medeis do all the rollup. He's just a camp follower.

Please present something convincing and, most of all, answer the question, whether Hale committed the sins of glottochronology and mass comparison. (Dubbing the last one lexicostatistics doesn't make it any more respectable.) Io (talk) 01:39, 22 May 2011 (UTC)


 * PS: I almost forgot. Having Bowern here on board as an interested party does not lend credibility to unbiasedness, Io (talk) 01:44, 22 May 2011 (UTC)


 * Pama-Nyungan appears to be generally accepted by historical linguists, even if most of them (or even almost all of them) may never have investigated the evidence closely by any measure. But the same is true for such large families as Austronesian, Northeast Caucasian, Sino-Tibetan, Afro-Asiatic, Niger-Congo or Nilo-Saharan, let alone the most obvious comparison, Trans-New-Guinea, all of with are supposedly highly ancient. In all these cases, there are extremely few experts, but other scholars, just like the general public, basically take the experts' word for it. That even though my impression of (say) the evidence for Niger-Congo (let alone Nilo-Saharan) has never been particularly good (and I'm not the only one with a very sceptical attitude in this case), and even though the big three African families are supposed to be on the order of at least 10,000 years old, hence perhaps twice as old as Indo-European, Semitic or Pama-Nyungan are estimated to be (and the fact that the membership of Niger-Congo and Nilo-Saharan, perhaps also Afro-Asiatic, has lately been reduced significantly could be seen as bolstering not only the case of the defenders, but also the suspicions of the sceptics, in a way). However, there's simply no way to make a general assessment of plausibility for such proposals (let's assume the time-depth estimates are roughly correct for the sake of the argument). Perhaps many non-IE families have really simply changed considerably more slowly (there are various cases which seem to confirm this impression) and borrowing hasn't confused the picture significantly (intra-family borrowing wouldn't matter for the general coherence of a family as such). Semitic, for instance, surely does not seem to have changed as quickly as IE. So I wouldn't be so quick to dismiss Pama-Nyungan out of such general considerations. Wikipedia can only report the general consensus in the field as reflected in relevant sources, and if the consensus is in favour of Pama-Nyungan and various other big families, we can't help it if the consensus is based on poor evidence and method and blind acceptance of the pronouncements of small circles of experts – it's not our job to judge the consensus. Scholars should not fall prey to the appeal to authority, sure enough, but Wikipedia is all about appeal to authority. We can't engage in our own research to confirm or debunk Pama-Nyungan and similar cases. I don't like this situation either, but it's simply the way academia works, and the way Wikipedia works, respectively. --Florian Blaschke (talk) 20:17, 15 April 2014 (UTC)


 * The only real doubter I'm aware of is Dixon, and other Australianists dismiss his arguments as methodologically flawed. Even Glottolog, which does not accept Nigor–Congo, Eastern Sudanic (let alone Nilo-Saharan), or even Omotic as demonstrated families, and which trims TNG significantly, accepts Pama–Nyungan.  That might be because the editors are better informed about non-Australian families, but there aren't many sources that evaluate all the families of the world.  — kwami (talk) 01:05, 16 April 2014 (UTC)