September 17, 2015


This is something I've been chewing on for more than a year now and yet never got myself to blog about (although I have mentioned in private or in comments here and there). Impelled by the minor but quite apparent NE African influence, genetic and cultural, on the Neolithic peoples of the Levant, whose offshoots eventually landed in Greece triggering the European Neolithic, I decided in the Spring of 2014 to explore, via mass-lexical comparison, if Basque language (and by extension the wider Vasconic family, which I believe now to be that of mainline European Neolithic) might have any relation with Nubian languages. I did not expect to find anything but noise but to my surprise the number of apparent cognates is quite significant. 

My primary analysis was this one but now I have combined it with a comparison with Proto-Indoeuropean (PIE), which is also very probably related to the roots of Vasconic: LINK (open office spreadsheet). 

The synthesis is as follows:

Of course the "cognates" are only apparent cognates at this stage of the research and the evaluation is necessarily subjective. But judge yourselves. 

If we discard the "weak" apparent cognates, the vocabulary correlation between Basque and Nubian and between Basque and PIE is pretty similar. But, in my understanding, both are well above the noise threshold, an example of which could be the PIE-Nubian apparent cognates, which are many many less. 

I must say anyhow that the oblique apparent cognates, that is when one word sounds much not like its strict synonym but a related one (for example words meaning hot and fire), look all very solid and most intriguing. 

Also, when attributing probabilities to origins of Basque words, Nubian appears to be at the origin of almost double the words (26%) that can be attributed to PIE (15%). Of course, for lack of data or because they actually have other origins, the unknown origins apply to the majority of words (56%), double than the Nubian origin ones.

However Nubian here is constituted of three different languages (Dilling, Nobiin and Midob), while PIE is just a single theoretical construct. This last must be done this way because many modern and historical IE languages, notably in Europe, have other Vasconic substrate influences, which must be studied separately from general PIE-Vasconic shared vocabulary. This kind of late Vasconic influence is very much unlikely in the case of Nubian instead. In any case I don't know of any a proto-Nubian Swadesh list readily available. 

Finally I must mention that because the PDF format is horrible for copy-pasting, I chose to re-transcribe the Nubian words according to my best approximation using a normal keyboard (not always the same characters that the original list uses).

Strongest Basque-Nubian apparent simple cognates

  • Basque - Nubian languages (English)
  • azal - àzì, àzzì-di (bark)
  • haragi - árízh (meat)
  • odol - ógór, èggér (blood)
  • buru - úr (head)
  • oin - ó:y (foot)
  • esku - ish-i, ès-sì (hand)
  • hil* - di-ìl (to die)
  • euri - are, ara, áwwí, áré, árí, áró (rain)
  • harri - kugor, kakar (stone) [notice also the pre-IE root *kharr- speculated to be at the origin of Karst, etc.] 
  • lur - gùr (soil, ground)
  • haize - irsh-i, éss-í (wind)

There are some others that are shared with Indo-European and with similar subjective "weight", not listing them here to keep things clear. There are also other apparent cognates that are arguably less clear like bat - be (one) that I'm also skipping here but you can find in the spreadsheet.

*Hil (meaning both to die and to kill in Basque, which can't be confused because they conjugate differently) seems ancestral to English ill and kill (this one via a Germanic precursor).

The intriguing oblique cognates

Notice that these words do not mean the same, yet their meanings seem strikingly related.
  • Nubian (English) - Basque (English)
  • hor, koy, kà:r (tree) - harri (stone) [notice that zuhaitz (tree) can be interpreted etymologically as zur-haitz = wood-rock, so the relation is not that weird]
  • ok-i, og (breast) - ogi (bread)
  • a-l (heart) - ahal (can (verb), potential, power)
  • azh, àz-ír, àzza (to bite) - (h)ortz (tooth), aitz (rock, peak) [some argue that originally "to cut", present in many cutting tool names: aizkor = axe, aitzur = hoe, aizto = knife, etc.*]
  • shu, zhúù (to walk) - joan (to go) [often pronounced jun or shun]
  • é:zhi (water) - heze (wet) [also archaic particle *iz-, meaning "water" by all accounts: itxaso = sea, izurde = dolphin, izotz = ice, and common in Vasconic river toponymy]
  • zhuge (to burn) - su (fire) 
  • zhùg, sù, sú:w (hot) - su (fire)
  • úr-i, úrúm (black) - urdin (blue) [archaic also green, grey]

*This one is an obvious and very prevalent Vasconic substrate infiltrator in Western IE languages: axe, adze, azada (hoe in Spanish), etc.

How can this be possible?

It is of course a mere working hypothesis and ultimately you judge but I find it hard to disdain. However there is no apparent connection, notably no significant genetic connection, between Basques and Nubians. So how can we explain this?

I have it reasonably clear myself, so I made a map to explain it:

Basque is after all just the last survivor of a once much larger family (Vasconic), a family that most likely corresponds to the languages spoken by the early European farmers (mainline Neolithic of Aegean roots). As that expansion was largely done in about a mere thousand years, I estimate that when both branches met near the Rhine, the two peoples could still understand each other, even if with some difficulty. Only the Southern/Western branch(-es) survived long enough to leave historical evidence, so it is hard to guess how the Northern branch evolved anyhow.

The Nubian linguistic connection is anyhow not the only thing that requires the Levant or Palestinian Neolithic step, also Y-DNA E1b-M78 (mostly V13 in Europe, attested in some early farmers and still very important among Greeks and Albanians particularly) and probably the so-called "Basal Eurasian" component that Lazaridis detected among early European farmers and that could well be the signature of African genetics from the Nile.

Linguistically, also the very notorious presence of Semitic (an Afroasiatic branch) in West Asia is surely another legacy of the same African influences in the Mesolithic Levant. Before this research, I thought it was the only one but now I strongly suspect that at some point Nubian (Nilo-Saharan) languages were also present in the region. Maybe one (Nubian evolving towards Vasconic) corresponded to Natufian proper and the other (proto-Semitic) to Harifian, the semi-desert pastoralist facies of the same wider culture. Can't say for sure.

The chain was once long but now only some of the most distant links remain unbroken. It is difficult to imagine that they were ever connected at all...

To do...

A lot remains to be done, of course:
  • These mass lexical comparisons only apply to a few families in the region and the rest should also be tested for. My energies are limited and so are my qualifications as "linguist", so I encourage others, hopefully more energetic and knowledgeable, to expand.
  • Grammatical features cannot be analyzed by this methodology. Again my means are limited. 
  • Anthropological research would be an interesting complement. So far the only shared cultural trait I could spot would be the use of bells attached to ankles for dancing but there could be others. 
  • ...


  1. I find the concept very intriguing for a host of reasons and I hope to see this theory develop.
    I'll first recommend an ebook "Prehistoric Iberia: Genetics, Anthropology, and Linguistics:" by Antionio Arnaiz-Villena, Jorge Martinez-Laso. It can be googled.

    There are several discussions on specifics of a Hispano-Badarian or Hispano-Nubian connections. Overall, this is all geared within an Early Neolithic expansion.

    Here's what I find most interesting.

    1. The distribution of Cardial-Impresso Pottery Culture also occupies North Africa and Nubia (not just Southern Europe), the leaf-shape arrowhead, obliques and iffles also being the most abundant of the Neolithic. Culturally, people very similar to Mr. El Torcs were also settled at one time in Northern Sudan, which became more widespread in the south as the Saharan climate worsened. It is plausible that a sizable genetic component of this still exists in Africans below the Sahel.

    2. More broadly, I'm curious about the speculated Nilo-Saharan family and Basque, for the reason that NS was theorized by Greenberg to have been spoken in the Green Sahara. Structurally, an ergative language like Songhay may be percentage of Neolithic basque-like language and native tropical languages. You have made the case for NS Nubian which is NS. It's also been theorized that Afro-Asiatic Berber has a Basque-like substrate.

    I can't speak to directionalities or Vasconic theory or whatever, however I would not be surprised to see real connections between the two. There are unquestionable material commonalities, and genetic as well.

    1. "The distribution of Cardial-Impresso Pottery Culture also occupies North Africa and Nubia"...

      Not as far as I know. There is a minor scatter to North Morocco and that is all.

      There is also a late influence on the Bilblos facies of the Amuq-Biblos culture, but we can't consider it Cardium Pottery culture, just Cardium Pottery in another culture (which may be related or whatever but not the same culture in any case).

      As for "Nubian Cardium" I know nothing about that. Would you care to provide some (hopefully good quality) documentation?

      I know you have mentioned some Egypt-Iberian possible pottery design links dating to much later periods than the one I am hypothesizing about but when I search for Naqada I, on my own I see mostly designs that are very much unrelated even to what you have mentioned (for example: and definitely nothing that is "Cardium" but, if anything, would be related to Iberian Chalcolithic and Zambujal, several millennia later. And nothing about Nubia/Sudan in any case.

      So would you mind clarifying your apparently very far fetched claims?

      "I'm curious about the speculated Nilo-Saharan family and Basque"...

      Me too but so far I have only worked with Nubian languages. And for one simple reason: they are the northernmost branch of Nilo-Saharan and hence the best candidate within that family for a possible connection.

      "It's also been theorized that Afro-Asiatic Berber has a Basque-like substrate".

      Who theorizes that? There used to be one of those junk amateur lists of random words with randomly attributed meanings and similitudes. My reading was that just another wacko claim, regardless that maybe a few words could still be representative of mutual influence... or mere noise. There are many opportunities for Vasconic to have influenced Berber (Megalithism) or for both to share a related "Solutrean" Paleolithic substrate and the "evidence" I've seen so far is so weak that the only conclusion one could produces is that there is no apparent relationship at all. Similarly my preliminary comparisons with Semitic also produce negative results (noise and nothing else).

    2. The distribution of Cardial-Impresso Pottery Culture also occupies North Africa and Nubia"...

      I can't answer this fully tonight because I am on a mobile device. But...I will qualify this as the circum-Mediterranean and North Pontic Maritime Impresso tradition which "sometimes" uses shell-tempered Cardial impressions in its pottery; that's somewhat different than shell tempered cockle shell as being the main discriminator and/or being unique to the S. Europe coast.

      There's a couple of very similar traditions between N. Pontic and Nubia impresso. I was only able to google this in short time.

      "I know you have mentioned some Egypt-Iberian possible pottery design links dating to much later periods than the one I am hypothesizing about but when I search for Naqada I, on my own I see mostly designs that are very much unrelated even to what you have mentioned"

      The way I interpret this is that North Africa has phases that are very similar to Europe from the Paleolithic. It has a Neolithic transition with demic immigration of maritime impresso farmers from the levant who meet hunter peoples. Somewhat related to the Euro Middle Neolithic, it has an Eastern influx of pastoralists from further East, namely Mespotamia and the Levant, but possibly the N. Pontic.

      So the rockerstamp or rockerbandt or whatever from Sudan/Chad is similar to the same from the North Pontic in a Neolithic context. The painted bi-chrome solar pottery of Naqada IIb and Tell Hassuna, Yamna and Beaker follow a tradition originating from Western Iran.

      While Berber is fully Afro-Asiatic, I read a paper this last year about some structural anomalies in Berber that were viewed as Basque-like. I'll try and find this tomorrow.!!

    3. I can't find the Basque-Berber paper. I don't think it was Mukarovsky 1969, which was lambasted by R.L.Trask in two books. It's all highly speculative and a little off topic from Nilo-Saharan.

      On the nature of the Maritime Impresso phenomenon, here's another paper that interested me concerning the North-Pontic area (again a little off topic)
      "A new approach to the problem of the neolithisation of the north pontic area"

      It's possible(??) that some Maritime Impresso groups had R1 lineages such as V88 & M269 and these were distributed in such places as Western Iberia, the Adriatic, the Ukraine, the Levant and Upper Egypt. Sounds weird, I know.

      I will triangulate this since one of the two Iberian farmers had a lineage close to V88, but more importantly, look at the correspondence between the V88 of Sudan/Chad and that of the North Pontic (where V88 also is present and diverse)

      Both Klysov and Maglio tried to reconcile this in one way or another. Mike Maglio had them leaving Iberia to Sudan and then to Eastern Europe. Klyosov had another route across North Africa.

      All off topic somewhat, but the underlying connections and timing do present this as a possibility.

    4. What is interesting is that there was apparently some early scatter of Impressed pottery in the Black Sea and that this may be related to the Dniepr-Don genesis that follows. At least the second paper is not in Russian!

      I wouldn't dare to associate with any lineage at this stage: CPP people were for all we know similar or identical to Painted/Linear Pottery ones, with G2a being the dominant (but not only) Y-DNA lineage detected to date.

  2. 1. So, to be sure that I'm clear what you did, your Swadesh list seems to have 207 words of which you were able to locate the Nubian word in 110 instances, and then found cognates within that list of 110 words. Is this correct?

    2. The linguistic designation "Nubian" can be a bit ambiguous because there are several communities which speak different languages from different language families that live in the Nuba Mountain region in the Kordofan Province of Sudan and sometimes people use a regional designation to refer to any language spoken in that region.

    One of those languages is a small group of probably related Niger-Congo language also sometimes called Kordofani, several languages that are probably Nilo-Saharan languages and, and more than one Afro-Asiatic language spoken in the area (the autochronous ones would be Cushitic and of course there would be dialects of the Semitic Arabic language spoken).

    While obviously, I know that you don't mean Arabic, I'm not clear which language family affiliation the Nubian language you are referring to is. Is it the small family of Nilo-Saharan languages linked above?

    1. Ignore my second question, I see that you clarified it in the body text in language that my eyes skipped over the first time. I do have a little trouble discerning from your second spreadsheet how many of the Swadesh words you found PIE words to compare to, however.

      Interesting hypothesis anyway. Do any of your cognates suggest Neolithic or post-Neolithic technologies? I don't see any that do at first glance?

      Is there any evidence of non-lexical similarities in grammar or phonetics?

    2. This comment has been removed by the author.

    3. Thank you for reading, even if late... ;-)

      As for your first question, I didn't quantify that part because I had one initial list, which does work pretty much as you say (good catch!) and then when mixing it with the IE-Basque comparison list it became expanded in order to incorporate these IE-Basque comparisons, limited by the PIE words listed in the Swadesh list available at Wikitionary (which are different from Nubian ones in many cases).

      So the preliminary quantification of a year ago (June 2014, I believe) is in the first link and for this entry I expanded with the PIE comparison (taken from another draft I did many years ago) and a re-evaluation of the "strength" of affinities (minor changes only).

      "Do any of your cognates suggest Neolithic or post-Neolithic technologies?"

      Only the fact that "aitz-" ("az(t)-") appears in many Basque (and non-Basque but surely Vasconic substrate) tool names seems to relate to Neolithic. In any case it'd correspond to a Vasconic and not "Vasco-Nubian" stage of the development. Good catch again, Andrew.

      "Is there any evidence of non-lexical similarities in grammar or phonetics?"

      That's something I'm also intrigued about and that I did once take a preliminary look at, using again online materials, but never really got to work on it seriously, so I can't say but that I keep a very vague memory of some features being suggestive and meriting a second look that I never took.

    4. I've been looking at the grammar and phonetics today after reading your post. The vowels seem identical. At first glance, it seems like Basque has more consonants, but often less studied languages like Nubian miss subtle distinctions in consonant sounds. The word order appears to be the same.

      One of the most striking points, because it is a litmus test for Nilo-Saharan classification is that Nilo-Saharan grammatical number has three categories that include a marked singular and an unmarked collective number and a plural. The Nilo-Saharan unmarked number doesn't seem to be quite the same as the Basque unmarked grammatical number, but having three kinds of grammatical number with the intermediate value being the unmarked form is very distinctive.

      Basque inflection of multiple forms of speech is far more complex than even Old Nubian, but that could be a function a time depth. Languages that haven't experienced mass waves of language learners tend to get more complex over time grammatically. While the oldest atttested forms of Old Nubian (one of the oldest attested non-Afro-Asiatic languages of Africa) was recorded at a time when there were at least three languages in widespread use in the same place for different purposes (Coptic liturgically, Greek in commerce, and Old Nubian in daily life) which might very well have resulted in grammatical simplification relative to early forms of Nilo-Saharan.

      Nubian languages are tonal, but apparently, linguists consider Basque to have some residual tonality in the form of residual semantic pitch.

      I haven't had time to really be systematic about it, but there are definitely similarities there on the grammatical as well as the lexical basis. Also, your Sino-Basque and Uralic-Basque comparison have helped establish a baseline for random noise which helps validate these conclusions and might be good to add to your chart.

    5. All this is very interesting. I appreciate your expansion and I'm glad that you share my excitement on this issue.

    6. "The Nilo-Saharan unmarked number doesn't seem to be quite the same as the Basque unmarked grammatical number, but having three kinds of grammatical number with the intermediate value being the unmarked form is very distinctive".

      I would think so. We call it mugagabea ("undefined").

      "Basque inflection of multiple forms of speech is far more complex than even Old Nubian, but that could be a function a time depth".

      I wonder about the true origins. Considering the rapid expansion of European Neolithic and the clear increase of admixture with HGs as the cultures moved west, I'd say that some quite massive amounts of new speakers were indeed incorporated at some point, at several points rather.

      As for tonality, it is subtle enough in Nubian to haven't been recognized in many many decades, confusing it with stress. It is a common African feature, while tonality is almost non-existent in Europe instead, so areal influences may be at play here.

  3. "the very notorious presence of Semitic (an Afroasiatic branch) in West Asia is surely another legacy of the same African influences in the Mesolithic Levant"

    I would agree that Semitic in West Asia is a legacy of African influences in the Levant, but I doubt that Semitic has been in West Asia since the Mesolithic. The earliest attested presence is Akkadian in the Copper Age, which surely isn't fair because Akkadian or a close relative was surely spoken at the time that the earliest written languages (Sumerian and Coptic) came along to attest it.

    LD based estimates of admixture dates in Cushitic populations, surprisingly, suggest West Eurasian admixture at about the same time plus or minus a few centuries as Ethio-Semitic, which suggests that this language family may be much younger than conventionally assumed, perhaps dating only to the arrival of the Neolithic via the Nile Valley to Sudan and Ethiopia, rather than in the Mesolithic or Upper Paleolithic as commonly assumed.

    The comparative similarity of Berber and Semitic languages relative to Coptic, also suggests that Coptic may have been a late arrival to Lower Egypt, perhaps imposed by an Upper Egyptian dynasty at sometime around the time of the unification of Egypt under the Scorpion King, before which it might have been a local Upper Egyptian dialect that was unexceptional except for the fact that its ancestral speakers would produce a powerful king in one of the earliest true states. But, it can't have happened that long ago, because Berber is quite young as language sub-families go..

    I tend to see the current range of Afro-Asiatic diversity has arising from an expansion that starts when the Neolithic revolution reaches Egypt (where populations increase by 100 fold in pretty much the minimum amount of time that human biology permits), spreading south to produce Cushitic, west to produce Berber (or a lost pre-Berber AA language), and east to produce Semitic (ca. 6000 BCE-5000 BCE). The original AA language of lower Egypt (which would have been intermediate between Berber and Semitic) is lost when Coptic predominates in the Copper Age. Chadic comes into being, derivative of Sudanese Cushitic when a bunch of R1b-V88 pastoralists assimilate a lot of Cushitic women and then move to Lake Chad ca. 5300 BCE. Omotic is probably a complex hybrid of Cushitic and Nilo-Saharan arising from proximity to both for a prolonged period of time, explaining its outlier status.

    Nilo-Saharan is very likely older than Chadic and isn't nearly so obviously Neolithic, so your Mesolithic expansion into the Levant works for it. But, I think the scenario of a subsequent AA arrival of proto-Semitic shortly after the Egyptian Neolithic makes better sense.

    1. The are two reasons to consider the expansion of Afroasiatic primarily Mesolithic:

      1. It is clearly a very very old language family, borderline the time depth where distinction becomes too blurry to be acceptable, and that is, by all accounts, some 12-10 Ka years old.

      2. The expansion of proto-Berber to NW Africa only seems to fit the chronology (and ultimate Upper Egyptian/Nubian origins) of Capsian culture, which is Epipaleolithic and roughly fits those very early dates. NW African Neolithic is just Capsian II.


      3. Semitic almost certainly expanded from the Circum-Arabian Pastoralist Complex, derived from Harifian via the PPNB disruption (of which it is initially a peripheral facies).

      4. Arabian Neolithic seems to have its origins in Palestine, including massive colonization, and again it should point to Harifian.

      There are no other cultural flows that I can consider on this issue.

      Of course the Mesolithic influences would not yet be Semitic proper, but a precursor within Northern Afroasiatic (pre-proto-Semitic if you wish). That should not be a reason for any qualm.

      Other branches of AA, like Chadic, could still expand in the Neolithic time-frame. I haven't researched the matter for lack of data, but they seem mostly a separate phenomenon from Northern Afroasiatic, which rather looks Mesolithic to me.

    2. Check this paper:

      It offers a glimpse of the plausible ethnic complexity of early Neolithic West Asia. I find particularly interesting fig. 4 (showing cultural groupings by lithic tech in PPNA and EPPNB, which may be actual ethnic groups).

  4. This comment has been removed by a blog administrator.

    1. ^^ He was banned long ago. I don't care what he says: he's just not allowed here.

  5. Interesting Theory but your Comparison of PIE and Basque is simply a fruitless attempt..

    1. This comment has been removed by the author.

    2. Please, i am trying to be peaceful here, you have proposed lots of nonsense in the past (actual dogmatic propaganda) so i have no interest in digging them, lets just focus on whats positive!.

    3. This comment has been removed by the author.

    4. You're trying to be peaceful? And then accuse me of nonsense and propaganda (of what?).

      What is clear is that you have an agenda of claiming an "Asian" origin for proto-Indoeuropean and that's why you threw your one-liner, regardless that one can also imagine Asian (West Asian) scenarios for the PIE-proto-Vasconic interaction. You are not dealing with the merits of the matches, just establishing a preemptive cordon sanitaire against any evaluation. And that is what I cannot accept, much less when PIE is still one of the best matching "languages" versus Basque. After Nubian maybe but still pretty high.

      There are words like hartz (bear) and hauts (ash, dust), but not just those, that demand a Vasconic-PIE interaction, particularly because the PIE forms are much better cognates than the Western IE ones. Whether this interaction took place in Europe (via maybe absorption of Paleoeuropean linguistic influence into Vasconic) or in Asia (before R1a and R1b migrated to Europe maybe), I am agnostic, but the relationship is striking.

    5. //What is clear is that you have an agenda of claiming an "Asian" origin for proto-Indoeuropean and that's why you threw your one-liner,//
      I'm not claiming anything, my suggestions and understandings are based on evidence and reasoning yours are what based on Presumptions,and Creationism,
      //There are words like hartz (bear) and hauts (ash, dust),//
      Can you elaborate more on those words please?.

    6. "Creationism"?!


      Hauts is listed in the above link, as for hartz, which is not included in the Swadesh listing (probably because there are no bears in Africa), see:

      It is the leit motif of Roslyn Frank's (collaborator of the late Larry Trask) theory on "Paleoeuropean" as linguistic superfamily or shared substrate that would link IE and Basque. She's onto something quite clearly.

    7. Yes, Creationism.
      I only find hartz attractive.

    8. Why Creationism? I am anti-Bible since I remember.

    9. As I said, I will not dig into something which is negative now, as it will give no results.

    10. You issued the accusation, you insisted on it, the least you can do is explain why? Meanwhile I will not answer any of your questions.

    11. You believe in peace? if you then please can we proceed on the proposed etymologies?.
      Why you are Creationist is not a new thing and it does not matter much now...

    12. I believe in truth, freedom, respect, dignity, social justice... "Peace" without them is just a decoy.

      So truth first.

    13. This comment has been removed by the author.

    14. Its good to know you believe in them, I also believe in a thing which is called change of concept by meaningful understanding, that you lack, the case is worse when you try to defend a certain theory,i think it will be worthless to recite here again why that theory is incorrect, so, i earnestly ask you to discuss the etymologies.

    15. When you explain your accusations. I don't feel you are treating me with respect, so I feel comprehensibly reluctant to play your game in your terms.

    16. I do respect you, Unfortunately a certain theory that you respect so deep makes a fool of you, I apologize if i seem to disrespect you but its indirect.

    17. You are showing disrespect to me in two ways: one by throwing personal attacks but the other is the more important one: you are not explaining whatever logic is behind those insults, leaving me defenseless.

      The least I can demand is an explanation of your attacks.

  6. After taking a look, yes, your Nubian and Basque cognates look reasonable, however, some may require a bit explanation as comparisons-
    1.lur - gùr (soil, ground) (They surely look from different roots)
    2. haize - irsh-i, éss-í (wind)
    3. ok-i, og (breast) - ogi (bread), what the connection of Bread and Breast??.
    4. a-l (heart) - ahal (can (verb), potential, power), My objection is same as previous.

    1. The method can only produce a list of apparent matches: it is an statistical product. In a case like this one with such a high frequency of matches there must be something to it but the second phase belongs to certified linguists with a salary, who will dedicate their lives, 8 hours a day, 5 days a week (except vacations) to untangling the mysteries of obscure intercontinental linguistic relationships. Or most likely won't because they have something else to do like documenting the vocabulary of some remote valley dialect or producing the Nth work on how marvelous the national epic is - for its linguistic qualities, of course.

      So, meanwhile...

      "what the connection of Bread and Breast??".

      Don't they seem oddly connected even in English? BREA-d, BREA-st! Staple food: one at the baby stage (or also pastoralist adult stage), another at the adult agricultural stage.

      Heart and "ahala" (power, creative power specifically, might) are very much related intuitively but also in Egyptian mythology. Before the Hebrew genesis there was another version in which Ptah did almost the same but he created with the heart (the Sun) and only later he named with the tongue/word (Moon) - unlike the priestly Hebrew version which is just about words but is clearly an Egyptian (edited) borrowing.

      There is no creation without ahala in Basque and there is no creation without heart (Sun) in the ancient beliefs of the Nile basin. However, in the Abrahamanic religion, creation becomes just the act of chatting and writting, of scribal documenting! What a degradation! The propagandists took over mythology and mythology became meaningless, soul-less, powerless.

    2. I checked the bread/breast apparent connection in English and do seem distantly related words, both from a PIE root (*bherw-, *bhrew- or *bhrews-) meaning apparently to boil, to sheathe or to swell, related to "brew".

      Interestingly "bero" means "hot" in Basque. Another oblique cognate?

    3. I cannot find any relation between Bread and Breasts:

    4. You only answered to two,anyway, the cognates seem impossible, but Bread and Breast have a root level connection yes, bero can be related but i don't have a Basque Etymological dictionary to be sure,but of course the connection is in IE not Basque-Nubian

    5. You won't find an etymology for "bero" IMO: it's a basic word. As there are no other living relatives of Basque to compare with and internal etymology is a dead end in this case, that's about it. Unless someone finds a cognate in Iberian or something, what I'm not aware of.

      This word seems indeed connected to IE and not Nubian. The Nubian word for "hot" (and "to burn") seems to produce "su" (fire) in Basque, as mentioned in the original entry. I think those double origins are interesting and that's why I posted Basque-Nubian along with Basque-PIE data.

    6. By Connection i meant Bread-Breast in IE languages in root level, the case of Nubian-Basque is still to be established in root level :) i.e. if the root also has given words such as related to Swell or Brew in Basque-Nubian, then it will be good.

    7. You know or should know that there is no such wide linguistic corpus for Basque or Nubian. The kind of depth in analysis available for IE (not always right but at least available) just does not exist for these lesser families: either it cannot (for lack of enough survivor languages to provide the necessary contrast) or it has not been researched (for lack of interest of the academia).

    8. That's a disappointment of course! but your idea i think has future :), anyway, I'm focusing now on PIE and Basque cognates, i think its important to investigate them, as i said i found the case of Bear tenable, can you list some more that you think are good here?, If you do then we can talk more and may found more interesting stuff!.

    9. It's all in the linked spreadsheet. Are you having any problems downloading it?

    10. Yes i have seen them, i will discuss them one by one here, anyway, to start i plead you to give your Top 5 now :).

    11. First you answer to my question: why "creationism"?

  7. *psten- breast Ancient Greek στῆθος ‎(stêthos); Old Armenian ստին ‎(stin); Sanskrit स्तन ‎(stána); Avestan [script needed] ‎(fštāna); Lithuanian spenys; Latvian spenis; Old Norse speni; Old High German spunni; German Spanferkel; Irish sine; Old Prussian spenis; Persian پستان ‎(pestān); English spean

    *peg- breast Sanskrit पक्ष ‎(pakṣá), Latin pectus, Latvian paksis, Russian пах ‎(pax), Tocharian A päśśäṃ, Tocharian B päścane, Old Irish ucht

    peysḱ- Found only in West Indo-European languages, namely the Italic, Celtic and Germanic branches. Perhaps derived from *peh₂- ‎(“to feed, to guard, to nourish”) and thus cognate to Proto-Slavic *piťa ‎(“food”), Sanskrit पितु ‎(pitu, “food”), Lithuanian piẽtūs ‎(“lunch”), Old Irish ith ‎(“corn”), Latin pānis ‎(“bread”), English food and German Futter ‎(“fodder”).[1]

    *bʰera- to split, beat, hew, struggle1
    to cut, scratch, split, rub, beat, hew, struggle1
    English: bread
    Proto-Germanic: *braudan

    to heat, boil, brew1
    to boil, seethe1 2 3 4 5
    English: bread
    English: brending
    English: broil
    English: broil
    French: brûler
    German: brühen
    Old English: broþ
    Proto-Germanic: *braudan

    to split, beat, hew, struggle1
    to cut, scratch, split, rub, beat, hew, struggle1
    English: bread
    Proto-Germanic: *braudan

    1. I was talking of "breast" and "bread" in English, not in Greek or other IE. They can perfectly originate in different PIE (or other) roots.


    2. Interestingly the actual PIE→English etymological relation is about swelling/boiling and not about "staple food" as I thought initially for the Nubian-Basque apparent phylogeny. I am positively surprised and bow to this IE logic in order to draw the Vasco-Nubian parallel.

      Initially, i.e. just hours ago, all I recognized was that breast and bread are curiously similar in English. Now I think I know a bit better, so thanks for the conversation.

    3. The classification of Nubian languages is uncertain. Maybe they are part of cordial ware expansion, but to the south, during the Green Sahara period.

  8. I can only a vague resamblance between peg and peh2

  9. So I'm starting to talk on the proposed IE-Basque cognates by Maju, I have to thank him for the effort!, i personally know its not an easy task to perform, I intend to go step by step, so now i pick 5 words from the Top of that list which are-
    1. Hi (Zu) 'You' Compared with IE Tuh (You).
    2. Gu ''we'' PIE *we
    3. Hiru ''Three'' PIE root *trei- "three"
    4. Luze ''Long'' with PIE *dlonghos ''Long''
    5.Labur, Motz ''Short'' with PIE *Mreghu
    On No. 1 I think its really weak the sounds don't match and even if there is a connection it can't be said to be of cognate type of relation.
    No. 2 is also problematic the same way.
    No. 3 same as before, however, Basque Bi which is listed as ''Two'' is quite close to the PIE *duwa which has Given Late Latin binarius "consisting of two , bini "twofold" is similar and can be a later IE borrowing.
    No. 4 is also not tenable due to the difference.
    On No 5 , Comparing Labur, Motz with Mreghu i think again very problematic.
    Ok I will venture more later....

  10. Supposing basque and PIE have a common proto language, doing the backwards transformation:

    H -> T
    G -> w

    1.Ti(zu) PIE *Tuh
    2.Wu PIE *we
    3.Tiru PIE: *trei
    4.Luze PIE: *dlonwtos
    5.Labur,Motz, PIE: *Mrewtu

    1. Notice, Daniel, that hi is like thou, tu (the original 2nd person singular) and zu is you, vous (i.e. a plural that has become singular by medieval courtesy usage). What I compared with thou, tu, was zu, and what I compared with Nubian was hi.

    2. You can't compare Zu and Tuh as cognates but distantly related roots.

    3. This comment has been removed by the author.

  11. 6. Har ''Worm'' with PIE *wrm worm.
    7. erro, sustrai ''Root'' with PIE root *wrad- "twig, root"
    8. Azal ''Skin'' with PIE *Pel,*Twakas ''Skin''
    9.Hezur ''Bone'' with PIE *Hasthi ''Bone''
    10. Aho ''Mouth'' PIE hohs ''Mouth.
    No. 6 it seems the IE root has come from the root *war To Turn as in Versus, or Skt. Vartana, or Wheel so my question is how you think they can be connected in root level? is there are other similar instances in Basque which show Turning movements etc?
    No. 7 yes it looks distantly related.

    1. No. 8. Seems quite Impossible! .
      No. 9. is interesting yes But it probably reflects an an Archaic relation the Laryngeal can be observed like in Hittite hastai, in Kurdish hestī so it means the relation again is close to Caucasus,however, the words are as notable are a bit different and one can even discard! but i'm keeping it as a probable one, though only as related not as cognates.
      No. 10 is very basic it seems and not special and also different, the Basque term misses the S sound.

    2. Number 6, 7 the mapping fails. But the pronounciation is close.
      I don't know about baske, but if the first sillable is stressed, it could come from a degenerated H, so Azal, could come from Tazal and Twakas
      9. If Hezur is segmented like, Hez Ur, zur being a stressed consonant, and applying the case above, we have Tez Tur. So Tez Tur vs. Tasthi
      10.Assuming stress on A, we have, Taho vs. Tohos

    3. Daniel, you are forcing evidence, the evidence that we have clearly tells that the cognates are weak! and mostly implausible, though there are some connectable of course and i intend dig deep on them also.

    4. Daniel: stress in Basque is normally distributed in two syllabes. So ázál (not ázal nor azál) and more usually (basic nominative): azálá. I'm using here stress markers ("tildes") as in Spanish, they are never written in Basque.

      Similarly: hé-zúr and he-zú-rrá.

  12. 11. Hortz ''Tooth'' with PIE g̑m̥bh/*hdonts ''Tooth''
    12. Mihi ''Toungue'' with PIE dn̥g̑hū, dn̥g̑hu̯ā ''tongue''
    13 . Jakin ''Know'' with PIE g̑en-, g̑enə-, g̑nē-, g̑nō/ Weid ''Know''
    14. Gogotu ''want'' with PIE *Tong ''Want''?
    15.Eseri ''Sit'' PIE *Sad ''To Sit''
    16. Izar ''Star'' with PIE *hstar ''Star''
    17.Bide way with PIE *Panth ''Way, Path''.
    18. Egu(-n-) ''day'' PIE *agh "a day"
    19. Urte ''Year'' PIE *Yar ''Year'' /*Wat ''year''
    20. Gorri ''Red'' PIE *(H)rudh ''Red''
    I find only 16. and 18, 19 worth discussing.
    On. 18, i think the PIE root *dhagwh- 'to burn, shine' is the case as in Dutch dag, Gothic dags 'day' Old Irish daig 'fire' Lith. degti 'to burn', Skt. dahati 'to burn', dagdha 'burnt' rather than *agh which is Watkins proposal and seems vague.
    16. is interesting also as well known there is Akkadian istar "venus." the Laryngeal in IE is found in Hittite: haster
    On 19 i think i have to check more it can be really interesting!.

    1. Gogo-tu (v.) may be "to want" but gogo (n.) is much more than that: the soul or mind, and that one is the real root. Buru (head) and buru-tu (to think rationally, to consider) are used for the more rational aspects of the mind, while gogo is more emotional probably. Buru is also used to make the reflexive form (to myself = nire buruari, lit. to my head).

    2. Urte (year), Urtzi (personified sky, "God" in Early Medieval Basque), etc. seem related to the root "ur" (water). Hence one can see urte as "the water cycle" and Urtzi as "the water carrier" but the suffixes are not clear in their meaning. Even the blue (arch. blue-green-gray) color name "urdin" seems related to "ur" as water, much as "zuri" (white) appears related to "zur" (wood), "beltz" (black) to "bele" (crow or raven) and even "gorri" probably not just to "gorroto" (hatred) and "gorrotu" (to hate) but also maybe to "gor" (deaf) and, according to Roz Frank, to Eng. "gore" (bloody flesh, meat...)

      Egun may well have an internal etymology egi-une (zone of bounds or zone of truth) and be also etymologically related to Eki (the Sun or the East) in indirect relation with the interesting verbs egin (to do) and ekin (to take action, to act - not in the theater sense however). Ezina ekinez egina: the impossible is done by taking action.

    3. Note about Urtzi (Ortzi, Ortz, Ost): IMO it should be the same as the Greek Uranus (Ouranos) and therefore an ancient pre-IE god of the sky, particularly the night sky, and astronomy. However the Basque form suggests a primary meaning as part of the water-cycle. It has no mythology but its influence in metereological terms and weekdays is unavoidable: ortzadar (sky-horn: rainbow), ortzarri (sky-stone: lightning), etc. Also Osteguna (day of Urtzi = Thursday, Jupiter's Day in Latin variants), Ostirala (fern field of Urtzi = Friday, which was the ancients' main holiday apparently, when the gods and mortals met to reenact the perpetual creation in their fertility cult, later known as the witches' sabbath).

    4. I am very impressed by your Philosophical explanations!, I salute your depth!.
      But of course most of the words can't be cognates to IE technically because of their structure.

  13. 21. Bero ''Hot'' with PIE *Tap/*ghwar ''Heat''.
    22. Zahar ''Old'' with PIE *Sanoh ''Old''.
    23 .Lehor ''Dry'' with PIE *Tars ''To be dry''.
    24. Izen ''Name'' with PIE en(o)mn̥-, (o)nomn̥, nōmn̥ ''Name''
    25. Ke ''and'' with PIE Kwa/Kai ''and''.
    26. -n ''in'' with PIE hen,hen-ter ''in''
    27. Ile ''hair'' with PIE pulah ''Hair''
    I have excluded 4, 2 of them were discussed and other 2 didn't make sense at all :).
    21.Bero is interesting, there is also Bar in Sumerian ''To Burn,To fire'', i think it can be related.
    25 and 26 are very basal perhaps archaic but as i said before they are not very definitive and found in languages world wide.
    Now, I will dig them deep, this was the first impression, i will let you know!.

    1. Notice that izen (name) and izan (to be) seem tightly related internally in Basque.

    2. Izar (star) could also be related in principle.

    3. Well Name and To be are quite close to the notion of existence. hence, i'm not surprised to see if they are internally related.

    4. No, i don't think its a good idea to Drag Izar with them, because it will dilute the relation with the IE root :).

    5. The question is that the best relation between iz- derivatives and PIE is probably the way of to be rather than name. Izan <> *h₁es- ? Basque simple present conjugation is as follow: na-iz, ha-iz, da, g-ara, z-ara (zarete), d-ira. Compare with English "is" and "are", for example.

      So IMO the Basque root would be iz-, from which izan (to be), izen (name), izar (star) and the fossil root *iz- for presumedly "water" maybe, as in *izaso → itsaso (sea), izotz (ice), izurde (dolphin). However what if *iz- was not "water" but rather "being" and many of those words require a "religious" or "mythological" explanation? What if izar meant something like "high being" and itsaso "ancestor of beings" (-(a)so is used for high ranking relatives: amaso: grandmother, arbaso: ancestor - or maybe was a marking of respect as you use -ji in India, in which case it'd be again "venerable being" or similar). What if izurde is as some have argued "gizurde" (=man-boar)? Or even giza-ur-ti, something like water-man, where water is "ur" and "urde", boar, is a false cognate? If izotz is just "cold being" and the *iz- water toponyms actually refer to nymphs and not directly to water?

      Or whatever. Because I do not have a strong opinion but I do think that izan (being) is older than izen (name) and not vice versa.

    6. About Indic Ji, Its of uncertain origin, some say it derives from Arya ''Honorable'' etc.
      Well i'm not sure but on Iz, i was thinking if we connect PIE root *es- "to be." as in English third person singular Is, seems weak but probably worth thinking.

    7. Just to be sure ''Dira'' is ''are'' in Basque right?.
      The English are comes from the root *er-, *or(w)- ''To Move,To Rise'', I find that difficult to compare!.

    8. "Dira" is "(they) are", our verbs are quite richer than any IE verbs, let alone English ones, so it is different from "(you) are": zara or zarete, or "(we) are": gara.

      For the synthetic intransitive verb izan ("to be" but only with the primary meaning of essence or persistent quality, not that of staying or going through a phase, that'd be egon), the structure is made of three morphemes:

      1st sing. N-A-IZ (ni → n-, izan → -iz, which is considered the radical form of the verb)
      2nd sing. H-A-IZ (modernly zara but that used to be plural)
      3rd sing. D-A
      1st pl. G-A-RA (for some long forgotten reason -iz becomes -ra in plural forms)
      2nd pl. Z-A-RA (ancient, now doubly pluralized as zarete)
      3rd pl. D-I-RA

      Notice that this is just the official list, dialects vary quite a bit: gara for example, as the song explains, can be gare, gera, gire and also gara, of course, depending on the region. This surely explains why it's "dira" and not "dara".

      We are lucky this is just a simple present time intransitive verb, because when we get with transitive forms with direct object and both transitive and intransitive forms with indirect object, it gets a bit more complicated, as both must be properly implicit in the verb. For example "I have" does not exist, there is "I have it/he/she" (dut), "I have them" (dituzte), "I have you" (zaitut), "you have me" (nauzu), etc. The verb's radical is obviously -u- but all the rest is a bit crazy. Luckily most verbs are not synthetic but use auxiliaries but these auxiliaries must be properly conjugated in every occasion and don't dare to say "ni dut", it is "nik dut", because "ni", without the -k, would be the object in a transitive sentence (but the subject in an intransitive one). Nor ("who" is, comes, walks... or is object of someone's else's actions) is totally different from nork ("who" does, says, has or eats).

  14. Thanks Maju, great post. Very well researched and quite compelling. My view is that your theory and Hans Mukarovsky's both have merit. However Mukarovsky (1981) makes the case that Berber language only spread to the west of North Africa from 4kya, and that Mande-like languages may have been spoken further north prior to this [he doesn't mention the 'green Sahara', but that would be my take on it]. There do seem to be a considerable number of equally striking Basque-Mande cognates. You cite E-M78 [E-V13] as possible genetic evidence; however E-M81 could also support Mukarovsky's theory. Maybe the two theories are both correct. I took a look at the 'Vasconic substratum theory' page on Wikipedia. The map shops Iberia, Basque Country, Western France, and British Isles. Looks identical to the distribution of West African HLA haplotype A*29:02-B*44:03-C*16:01-DRB1*07:01-DQB1*02:02 in my opinion [although I know you don't have much faith in the HLA for popgen studies because selection can act on it - but could Y DNA not have a similar issue?] I would also like to somebody to take a closer look at R-M269 in Africa - it's definitely there, not all the African R1b is R-V88. How old is it? Didn't some ancient DNA from Neolithic Canarian Guanches have R-M269?

    1. I'm not familiar with Mukarovsky's theory although I think I heard of it in vague terms long ago. A big problem here is that no or very few West African genetics can be considered old in NW Africa and that instead there is a strong genetic component stemming from NE Africa which seems to correlate well with the Capsian + Berber model. Also Fulani, who are the relevant only West African population with NW African genetics speak a clearly Atlantic Niger-Congo language. Another thing would be if Mande has some relation with NE African languages but I don't know of any.

      "I would also like to somebody to take a closer look at R-M269 in Africa - it's definitely there, not all the African R1b is R-V88. How old is it? Didn't some ancient DNA from Neolithic Canarian Guanches have R-M269?"

      Yes, all that is a most intriguing issue. There is a lot of R1b in Sudan (less in Egypt and much less in NW Africa) that demands a good study or two. IMO (always in wait of further new info) R1b probably originated in West Asia with an Upper Paleolithic time-frame and spread to NE Africa, roughly associated with J1 (which seems also old and diverse in that region, unlike in NW Africa), and probably also T and at least some of the Eurasian mtDNA clades (such as X1, M1, etc.) Later some of that R1b (V88 specifically) spread to parts of the Chadic area, while J1 instead seems to have spread preferentially to the NW, along with E1b-M78.

      Anyhow, the Guanche mummies did have some notable R1b (10%) but also rather high frequencies of I (7%), which almost necessarily arrived from Europe, as well as mysterious K(xP) (10%) and P(xR1) (a singleton, 3%). In my humble opinion, they suggest interaction with Europe, maybe via pre-Capsian NW Africa (mtDNA U6 structure would seem supportive of that) but they also harbored large amounts (~40%) of what I consider Capsian lineages, J1 and E1b-M78, owing surely to the late dates of Canary Island colonization.

      Hence they seem a mixture of fossil "European" lineages (now rare in NW Africa Y-DNA-wise but better preserved in mtDNA), the mysterious local E1b-M81, which also seems old, and the "Capsian" lineages that IMO arrived in the Epipaleolithic and can be strongly related to Afroasiatic or other NE African roots.

  15. Thanks for your reply Maju.

    "A big problem here is that no or very few West African genetics can be considered old in NW Africa.."

    But do we know how what percentage might date back to Saharan dessication [6kya] as opposed to more recent times, eg. Islamic slave trade?

    "Another thing would be if Mande has some relation with NE African languages but I don't know of any."
    Nubian is a Nilo-Saharan language in NE Africa. Mande languages on the other hand do apparently have similarities with Songhai, a different Nilo-Saharan language. Also, it might be worth investigating the structure of the Saharan languages [Tebu] to see where they fit in.

    "Later some of that R1b (V88 specifically) spread to parts of the Chadic area.."

    Some R-M269 has turned up in Chad and also in Nigerian Hausa. Though of course nothing like as much as R-V88.

    "Anyhow, the Guanche mummies did have some notable R1b (10%) but also rather high frequencies of I (7%), which almost necessarily arrived from Europe.."

    Sudan does also have Y DNA haplogroup I though [1.3%].

    Finally, here is a link to an extract from Mukarovsky's thesis if you wanted to have a read [I'm not necessarily endorsing it]:-

    1. I wouldn't dare to quantify but my impression is that the general African (non-Eurasian) components (incl. notably mtDNA lineages) in NW Africa has three origins or layers:

      1. A deep one that may date to the OoA times (Aterian) or maybe secondary later flows. This would correspond with one or two of the "Green Sahara" periods: the Abbassia Pluvial and the Mousterian Pluvial.

      2. A second and most important layer that seems to be sourced to NE Africa (Nile basin around Sudan) and should be related to the Capsian flow of the Epipaleolithic. Again it relates with a "Green Sahara" period: the Neolithic Semi-Pluvial, which actually began much earlier than Neolithic proper and ended soon after the Neolithic began.

      3. More recent trans-Saharan interactions that may be simplified as the slave trade.

      "Sudan does also have Y DNA haplogroup I though [1.3%]".

      It's interesting. Considering one of our best tracks, which is mtDNA U6, the impression I have is that NW Africans influenced the Nile region (bringing U6a with them) and that later some of that back-migrated to the West with Capsian. I guess that the Y-DNA I you mention might be tracked to those pan-North-African flows, speculatively at least.

      I'll take a look at the Mukarovsky thesis. I'm intrigued.

    2. I took a look and have mixed feelings. There are maybe three types of "cognates":

      1. Some that just don't make sense or even use arbitrary Romance-borrowings
      2. Many that could equally be cognate of Basque or English, what requires further analysis of whether, assuming merit, the double relation is via PIE-Basque interaction or via Vasconic substrate in English or Germanic specifically.
      3. A handful that reminded me of the apparent cognates, some oblique, in my Vasco-Nubian list above, so meriting research also about possible Mande-Nubian relations.

      In the best case, I'd probably lower the percentage of valid "cognates" below 10%, what begins being borderline "noise" but I guess that it still merits a second look, particularly on light of the (hopefully more solid) relations I'm proposing here.

  16. You might want to take a look at, which is the "official" lexicostatistical phylogenetic tree of world languages. There, Basque shares a branch with Shabo, a Nilo-Saharan language spoken in Ethiopia. The nearest relatives are a couple of languages from New Guinea and Australia, plus the Salish languages spoken on the coast of British Columbia. Unless a couple of Basque sailors stranded on all of those shores in the 16th/ 17th centuries, and assuming the lexical relations are not just statistical noise, we are talking about a very deep, most likelyy paleolithic root.
    Some of your "oblique cognates" are pretty widespread. I have tracked *ku "to run, go, leave" across a couple of language families worldwide. The root occurs, e.g., in canis "dog" (->the runner), as well as in the Kudu, akkad. gadu (goat), and various African and Dravidian terms for gazelles. é:zhi (water) alludes lat. aqua, Adang (Alor, Indoneisa) si, Aghu (PNG) oxo, Chechen xi, Ashing (ST, Burma) asi, Bafut (NC) ki, Bangka (NC) zo, Cantonese sei, Quiche xa7, Chibcha sie, Chicasaw oka, Khorasani su, plus various others (though bo-/wa- for water seems to be even more widespread). .

    1. Can't find it, as the search feature does not seem to work beyond the intro. I've been looking up and down for a while and only got a feeling of dizziness. I'll take your word from it but I wouldn't trust too much such a massive tree, which in many nodes doesn't seem to make sense.

      "Unless a couple of Basque sailors stranded on all of those shores in the 16th/ 17th centuries"...

      Not likely to be influential even if it happened. The tree seems unable to grasp the PIE-Basque relations, which seem quite apparent, so I don't think it's too credible on first sight.

    2. The world language tree chart strongly supports the case of Basque as a true isolate. The connections to the languages it is linked to are very few (eyeballing it, certainly under 4 matches). Really, the entire chart deserves white out in the noise range of about 0-4 with everything in that range attributed to random noise. The fact that the closest match is Nilo-Saharan is interesting, but that link in the tree is very, very basal to the point where it is really meaningless.

    3. I couldn't find Basque but I could find the Berber cluster and they are in a similar situation: with only thing close being some Native American cluster. And everybody agrees that Berber is Afroasiatic, right?

    4. Everyone absolutely believes that Berber is Afroasiatic.

      Of course, the paper surely does systemically underestimate cognates because it is being done strictly by a computer without a deep understanding of the way that the semantic meaning of a root can subtly evolve into something that isn't as obviously connected. The method works well enough to show close linguistic relationships via lexical similarity, but it isn't well suited to the more subtle cases that it purports to take on.

    5. That's what I figured: that the analysis was computerized, what may be a first step, especially when dealing with such large amounts of info, but I don't think computers can still do most tasks better than humans can.

  17. This is another (amateur) study on Basque's similarity to Niger-Congo languages.


    By GJK Campbell-Dunn M.A. (NZ), M.A. (Camb.) Ph.D.

  18. I see little merit to the article other than "bat" (1) and "bi" (2). It is apparent that the author does not speak Basque. "Lehengo"? "first" is "lehen(-a)".

    So good for his suffix -ko hypothesis, particularly because -ko in Basque means "from" (locative of origin) and is probably related to the common IE suffix *-ikos (English -ic, Latin -icus, etc.), meaning roughly the same. Vide:

    Hence "lehengo" means "from before" ("lehen" also means "before") and has nothing to do with what Campbell-Dunn says. Most other alleged etymologies are also unlikely: k>h is OK but n>h># is totally unacceptable, etc.

  19. Lol, this guy has another one too. I'm not a linguist whatsoever so I can't really give a good appraisal. However to me it looks like all languages have some similarities if one looks hard enough. A bit like looking at 'interference' on your tv screen and thinking that you can see a picture there.



    GJK Campbell-Dunn

    MA (NZ), MA (Camb), PhD (Cant).


    1. Yeah, one problem is that phonemes and phoneme combos (syllables, words) are only so many and therefore there will always be some random matches between any two languages. This may seem odd but it is perfectly normal. That's why I talk of noise level and above noise level. Below 10% of apparent cognates (rough figure, I believe there are studies pointing to more precise figures, maybe 7%?) it's nothing but noise and any real cognate can't be discerned from a mere coincidence.

      Andrew mentioned above that I should have mentioned previously negative analyses for comparison, such as the Sino-Basque skeptic try:

      He certainly has a point, because they are illustrative of what is not affinity but just coincidences, noise.

      Another different case he mentioned was my rebuttal of the proposed Basque-Uralic:

      In this case, I didn't make a mass lexical comparison but just went one by one debunking the reasonings of the proponent.

  20. GJK Campbell-Dunn does agree with you on one point.
    At the end of his third thesis ["THE ETRUSCAN DECIPHERMENT: ETRUSCAN MORPHOLOGY COMPARED WITH NIGER-CONGO" - ] he writes: "African substrate existed in the Aegean and Mediterranean from a very early date."

    1. Not everything "African" is the same: Africa uncomparable to anything else in its diversity, if anything to Asia.

      In any case I prefer a harsh criticism from someone who does things well than "agreement" from someone who does not add up.

  21. Maju wrote,


    •harri - kugor, kakar (stone) [notice also the pre-IE root *kharr- speculated to be at the origin of Karst, etc.]
    •lur - gùr (soil, ground)
    •haize - irsh-i, éss-í (wind)

    Nubian (English) - Basque (English)
    hor, koy, kà:r (tree) - harri (stone) [notice that zuhaitz (tree) can be interpreted etymologically as zur-haitz = wood-rock, so the relation is not that weird]


    I call random (perceived) resemblance.

    Kʰar is an Armenian word for "stone," and хад (had) is a Mongolian word for "large rock, crag, cliff." These seem more similar to the Basque word IMHO. Japanese has ki ~ ko- "tree; wood," which would be a better comparand for Nubian hor, koy, kà:r (tree). Of course, if Nubian kugor, kakar (stone) is assumed to have experienced reduplication (e.g. *kar "stone" > *kar-kar "(many) stones" > kakar), then it might just as well be compared with Basque harri (stone).

    Nubian gùr (soil, ground) seems more similar to Korean horg > heuk ~ heulg- (soil, ground) than to Basque lur. Also cf. Turkish yer (place, spot, position, location; space, room; floor; ground, earth; piece of ground, piece of property; terrain, region, area; the Earth), Mongolian ger "home, house, traditional dwelling of Mongols," Greek gē ~ geō- "earth," Gaia "Mother Earth," Germanic Germanic *er-tho "earth," etc.

    The words for "wind" seem superficially similar to Japanese kaze "wind, air(flow)," Mongolian хий (hiy) "air, gas, bubble; idle, empty, vain, futile; hysteria," Chinese 氣 qì "air, aura; spirit, metaphysical force; gas, bubble, carbonation." The Mongolian word looks somewhat like an early loanword from Chinese, but cf. Mongolian салхи (salhi) "wind," which I suppose might be a compound of *sal + *kʰi.

    Anyway, I think all the word pairs you have presented are just as likely to reflect random coincidences or retained ancestral commonalities as they are to reflect a unique relationship between Basque and Nubian.

    1. As I said before: judge yourselves, I don't pretend to argument beyond doubt but rather to open a window for further exploration. What I do think is that the PROPORTION of "coincidences" is way too high to be just "noise".

      Re. Armenian, recently I was approached by someone asking for feedback on a hypothesis about Basque-Armenian semblance. After I provided it, my correspondent seemed persuaded that there was nothing of relevance. So far all Armenian words that seem to have correspondence with Basque seem to be via PIE, for example (Old Armenian) ačiwn < *h₂eHs- (PIE) <> hauts (Basque). In this case, Wikitionary also claims a PIE root, with even Asian cognates such as Sanskrit karkara, (hard, firm), Greek kárkaros, (hard, rough) and Welsh carn (cairn or stone mound). Latin "cancer" (crab) is also allegedly related but compare with Basque karramarro (crab, crayfish), which also has, even more clear the *k'aŕ- root present. "Mamarro" is critter, so karramarro would be something like rock critter, very straightforwardly if we accept that the *k'aŕ- root is PIE.

      The main problem here would be its presence in Sanskrit (Armenian should have an easy European origin explanation) but something I'm becoming more and more aware about Indo-Iranian genesis is that it is not so directly an offshoot of Yamna but that it has some important Corded Ware derived influence, what may have brought Vasconic terms to it. Hence we must look at Tocharian and Anatolian for the purest PIE controls when comparing with Basque because otherwise it's hard to differentiate between PIE-Basque relations and Early European IE - Basque ones, which may also have reached Indo-Aryan and, of course, Phrygio-Armenian. Sadly, we don't always have references for those two early branches of IE.

      Searching for a most correct etymology of Basque "karrika" (street, possibly meaning "paved": [k]harri-ka, or stony place in Occitan: "garriga"), but maybe from Occitan "carrer" (street, apparently from the same root as cart), I found this paper (in Spanish), which highlights some of the difficulties on the proper etymology of words with the apparent root *khaŕ- or similar (there are other roots that are too similar and cause confusion, such as PIE *skerbh- → curb). The author however finds that the root *khaŕ- is most common in Western Europe (Germanic, Celtic and Italic modern areas, as well as Basque) and seems to support a pre-IE origin.

  22. One thing that caught my attention the other day was that the Songhai word for river is "isa". Which made me think of rivers with similar names in Europe. Besides those rivers named 'isa' by Songhai language speakers in their territory, there is also Oued Isser river in Algeria; Isère river in SE France; Isar river and Isen river in Germany; Yser river in Belgium/France; Oise river in France [originally Isara];Isara river and Eisack river in Italy; Éisra and Istrà rivers in Lithuania; and Jizera in Czech Republic. And in England the River Ure is believed to have originally been 'Isura'; and River Aire was 'Isara'; and Thames was 'Tam-esis' and 'Tam-esa'. Probably there are others.

    1. 'Ure' could be just 'ure': water o the water in Basque (standard form: 'ura', but 'ure' is common too).

      The issue of old European hydronymy's *iss- or *iz- is or should be well known. In Basque iz- does not mean anything anymore (except the root of the verb izan: to be) but it is "fossilized" in many words in which it seems to mean "water" or something similar: itsaso (sea) ← *izaso ← *iz- + -aso (ancestor: arbaso, respected relative such as grandfather: aitaso, grandmother: amaso, etc.), izurde (dolphin) ← *iz- + urde (boar), izotz (ice) ← *iz- + (h)otz (cold), etc. Notice that Germanic "ice", "iss" seems in this context a Vasconic borrowing from the same root. Another possibly related Basque word, which appears in the mass lexical comparison above, is heze = wet, which I tentatively relate to Nubian é:zhi (water).

      As for the presence of an apparent cognate in Songhai... it could be a coincidence or Nubian/Nilotic influence or whatever. A single word can always be a mere coincicence and that's why I emphasize that my hypothesis is not based on some random coincidences, which you can always find between any two languages, but on a significantly high proportion of them. It is the unusually high frequency of "coincidences" what makes the above mini-research a strong indication, if not yet "evidence".

  23. But how exactly would linguistic links between nubian and basque be present without afro-asiatic links?

    1. That is a very good question. I reckon that I haven't yet made a proper mass lexical comparison Basque-Afroasiatic, so the matter remains open, but a quick look at numbers doesn't suggest any strong link, nor it has been ever proposed.

      If this stands, a possibility could well be that Nubian expanded mixed but also somewhat separated from Afroasiatic, much like the waters of the rivers Negro and Solimoes in Brazil run by the same course but only very slowly mix, being very apparent because of their different colors. Another possibility could be that Afroasiatic expanded after Nubian, at least in the sedentary "proto-farming" (and later also farmer) areas, being specialized in more arid, open lands initially. Complicated but there was all that "Green Sahara" back in the day, allowing for a complexity that maybe later was partly erased.

      Requires research in any case, clearly so.

    2. Hmm.. I'm fairly certain afro-asiatic was present in those areas prior to farming. Maybe this nubian connection is restricted to nubian and has nothing to do with nilotic, in which case I'd suggest a link with R-V88 somehow, if we were to accecpt that it enterred Africa from Europe, a scenario of which I have good evidence. And because R-V88 is most likely older than vasconic, then, provided my scenario is correct, we would be speaking of a substrate, from whatever language(s) these R-V88 people might have spoken, shared between basque and nubian. I think that by analysing Sardinia(pre indo-european of course) on linguistic grounds we'll be able to have a better grasp of the matter. It's interesting (actually, its much more than that) that Sardinia shares Y-DNA I-M26 with Basques while at the same time being the european hotspot for R-V88... And also of A-M13 (previous A3b2), a sudanese/nilotic marker.

    3. It must be said that Nilo-Saharan is not a solid family. Nubian belongs to Eastern Sudanic, which may or not be part of the tentative Nilo-Saharan greater family. I focused on Nubian languages for accessibility (Starostin Jr. has been working on them, recently I realized he also published a wider study on "proto-Nubian", which I have not yet compared with) and also because, being the northernmost language of the family, they had greater chances to throw up some relation (if Nubian failed it was surely worthless looking into the wider family).

      I don't draw any connection with R-V88 but rather with E1b-M78 and particularly its Neolithic European main derivative E1b-V13. But it can't be of course matter of a single lineage. V88 is associated to Sardinians and may have expanded also within the early European Neolithic but in Africa it is most tightly related to Chadic, an Afroasiatic language grouping, not Nubian-related, so I don't see any obvious relation on first sight. Some criss-crossing of lineages probably happened but I am not particularly inclined to associate one lineage = one language family, unless there is clear support for it. In this case, if Vasconic is in fact the mainline Neolithic language family, then it's most direct association is with E1b-V13, G2a and also I-M26, of course. Other lineages may or not or just irregularly been related to the process.

    4. But R-V88 is still high in Sudan, representing about 1/8 of the Y-DNA there. It's especially common in Nubians, Copts and Arabs. But the arab expansion probably reduced the frequency of other nubian haplogroups in favor of J1, which seems to be mostly of arabian derivation in Nubia and Sudan, so R-V88 could have been as high as 20% before J1 made its move. Also note the high frequency of both B-M60 and R-V88 in most non arabic-speaking groups of the egyptian/north sudanese region like Siwa Berbers (about 25% each), Copts, about (15% each) and Nubians (with about 10% each). So R-V88 definetely has a connection with the nile, even though we would tend to be blurred by its high success in the chadic area, which might of course be largely due to a founder effect. We cwould of course need the check the diversity level and I bet these would be higher in the northern Nile. Furthermore, V88 is also common in Bantoid groups from Cameroon, with the Fang ethnic group of Cameroon/Gabon/Equatorial Guinea showing as much as 20% of it. As a result, it also took part in the bantu expansion as shown by its zambian représentatives. All this to say that the guy is not necessarily chadic and happens to be associated with this particular afrasian branch because it settled around lake Chad and thus managed to replace a large part of the original paternal makeup of the region.

      In all, I wouldn't be surprised by the role R-V88 might have played in this linguistic connection. Maybe if we could compare the principal afrasian branches and see which of these would have some connection with vasconic, I'd bet chadic would show more of it, just like the next guy pointed out. I also really don't think V88 would have brought pastoralism because I assume it enterred from Europe. On the other hand I can almost prove that T-M184 brought cattle culture from bab-el mandab and more anciently from India, or rather from Gedrosia.

      Returning back to the shared linguistic connection, it's either that or the expansion of E-V13, which would imply that at least part of the E-M78, or even E-M35 as a whole, originally spoke some nilotic languages (the ancientness of E-M78/M35 would maybe a reason for nilotic being a poorly defined languistic family). And then we have E-V22, sister calde of V13, reaching very high frequencies in nilo-saharan speaking groups of Darfur.

      On a side note, we also have I-M170 in the northern nile region but not in North Africa. It would be interesting to see which subclades would be present.

    5. Lineages don't speak languages, people do. Lineages stay, languages change. I have the same Y-DNA lineage as my great-grandfather but my birth language and his are totally different, and without moving from the same small province, just because of acculturation. And my brother's great-grandsons may perfectly speak another birth language such as English or Chinese or Russian or Arabic.

      Otherwise R1b-V88 seems like originating from West Asia (or even in Italy), just as all other major R1b sublineages, and Sudan may well host diverse R1b sublineages: it is AFAIK, a matter awaiting research. But in any case I don't have any reason to imagine R1b of any kind migrating from Africa to Europe (prove me wrong but with data) nor I have any certainty that such lineage alliances are from the Holocene: naively I'd say that R1b in Sudan should have arrived just like J1 or T to the wider NE African region in the Upper Paleolithic.

      "... it's either that or the expansion of E-V13, which would imply that at least part of the E-M78, or even E-M35 as a whole, originally spoke some nilotic languages"...

      They may have spoken different dialects of the Eastern Sudanic and Afroasiatic family. There's no particular reason to believe that they were monolingual across large areas and the diversity of lineages involved allows for diversity in languages too. I'm more interested in cultures than lineages, honestly, and what we see in the West Asian Neolithic, especially at the beginning, is a diversity of cultures that allows for diversity of languages.

      Just like today in the Nile Basin both families overlap irregularly, they might well have done also in the past in wider areas. I'd imagine that Afroasiatic is more related to semi-arid specialization (eventually pastoralist), while Eastern Sudanic seems more related to river basins (and eventually mixed farming). With time and Chalcolithic hierarchization, Afroasiatic has imposed itself in places like Egypt or the Levant, but not so much in the craddle of both families upstream the Nile.

    6. Yes but due to the fact that population densities were much lower in the past, it allows some haplogroups to have the upper hand as time passes. That's even more so in the case of isolated population because they're not receiving any influx that could throw in some new lineages there. Of course, when I say this or that particular haplogroup spoke a certain language, you know it's not to be taken literally but it only means that haplogroups and genetics as a whole tend to be correlated with cultural phenomenons.

      But anyway, I never said R-V88 enterred from Africa, I said it enterred Africa from Europe or from Italy/Sardinia to be more exact. As for Sudan, R1b there is mostly, or wholly, of two kinds, R-V88 and R-M269 with the first one being much more common than the other.

    7. We are talking of whole global regions: Europe, NE Africa, etc., not some remote valley. We can see how many many different lineages persist in the World from the times of the OoA or even before in the case of Africa. Population densities were "low" in Europe as in China, in NE Africa as in West Asia or India... all is a bit relative and I would not make a fetish of that idea of "low population densities": some reasonable diversity should be expected, especially in the warmer or more inhabited regions. In fact a problem is that most high quality ancient samples come from northernly areas, where colder climate helps preservation but also hampered settlement in the past, so they should not be considered representative of the wider regions but rather "peripheral", particular cases within these.

      "But anyway, I never said R-V88 enterred from Africa, I said it enterred Africa from Europe or from Italy/Sardinia to be more exact".

      Heh, that's something I also fail to see: my impression is that while Italy seems to retain some diversity, the West Asian origin is much more parsimonious: other R1b and other lineages like J, etc. behave similarly, West Asian interactions with both Africa and Europe, as well as Central and South Asia, are well documented, instead direct Italy-Africa relations before Phoenicians are not. The evidence in general points to Afro-European interactions being mostly mediated by West Asia, with the exception of those taking place across Gibraltar State, which are easy to identify as a distinct.

      I find a bit annoying when hypothesis about people's migrations are forged on mere genetic reasoning, particularly when this reasoning is not too solid. We need to have ample vision: look not just at single lineages but at patterns affecting groups of them and other markers, and also, crucially, at what we know about prehistory based on archaeological data, which in most cases is very important and can't be just ignored. Only if we consider everything together we can hope to understand, else the tree won't let us see the forest way too often.

  24. 'Isa' and 'Isara'. Could be related to Chadic languages [eg. Kotoko]. - Sara, sare, sari, shari, chari, tsadhe, tsad [corrupted into Chad] - all mean water apparently. Eg. Chari/Shari River, Lake Chad/Tsad.

    Incidentally the River Niger is known as 'Isa' by the Songhai, but also 'Kwara' by the Yoruba, and 'Eghirreu' by the Temashight [Tuareg] and the Hausa. The latter two sound somewhat reminiscent of 'agua'/'aqua' in my opinion.

  25. "iɣzā" = 'river' in Riffian Berber language apparently btw.

  26. I'm now unsure whether 'sara'/'sare'/'sari'/'shari'/'chari' and 'tsadhe'/'tsad'/'chad' are not actually from Kanuri, a Nilo-Saharan language, rather than particular Chadic languages.

    I also noticed that Kikongo [a Bantu language] has 'nzere' and 'nzadi' for 'river', which seem to resemble both 'sara'/'sare'/'sari'/'chari' and 'tsadhe' as used in Chad.

  27. I agree with Squad that a lot of this is likely tied up in the story of R1b-V88. I find it more likely that this would be a back-to-Africa phenomenon than the reverse (isn't Afroasiatic enough?)

    What are your thoughts on the R1b/R1a turnover that seems to have happened in the Steppe? It seems to be a likely place for these Vasconic-PIE contacts to happen, particularly if Proto-Indoeuropean arrived with R1a, and if the previous R1b inhabitants were speaking something Vasconic. Vasconic influence does correspond very well with the distribution of R1b it seems, and we know it was present in Iberia relatively early.

    Also, any thoughts about the analogous situation in the Americas with respect to the spread of agriculture? I can't say I've personally spent a lot of time looking at it, but I think it would be more amenable to study given the more recent timeline. It seems like agriculture was able to "jump" language families as it spread in the Americas, with these later adopters such as the Iroquois continuing the spread of agriculture without any linguistic ties to the Mesoamerican agriculturalist core. Thoughts?

    1. "What are your thoughts on the R1b/R1a turnover that seems to have happened in the Steppe?"

      It is important to underline that the R1b sublineage that has been found in the Volga (and is still nowadays common there) is just a cousin of the main branch that we see in Western Europe, the link being probably in West Asia maybe in the Balcans. It implies nothing re. the steppe.

      On the other hand something that is a common element not just between these two R1b branches but also others under R1b and even R1a itself, is that they all stem from West Asia. So I do consider that parts of West Asia may be at the origin the shared PIE-Vasconic vocabulary. If so, we'd be before a Neolithic connection distinct from that of the Nubian one (they only overlap in a few words and there is no significant PIE-Nubian link I can discern).

      An alternative possibility is that we are before a genuine Paleolithic European language of some sort (the Eastern Epigravettian tongue, for instance) that influenced both Vasconic and PIE (maybe even being a direct precursor of PIE in its core structure and not just vocabulary). There may be other options but these two are the ones I find easier to consider as serious possibilities.

      "Vasconic influence does correspond very well with the distribution of R1b it seems"...

      That would be the classical stand of someone like Venneman but I think that we can also see a lot of Vasconic substrate in places like Italy or the Balcans (Ibar river, "gore" for what in Basque is said "gora"), as well as the importance of the Sardinian connection (genetics, culture and linguistic substrate but not meaningful Western type of R1b nor Magdalenian substrate), led me to propose that Basque and Vasconic are rooted in the Neolithic expansion and are not related to this or that lineage particularly.

      "Also, any thoughts about the analogous situation in the Americas with respect to the spread of agriculture?"

      I only know bits about that but in North America, at least the lands surrounding the Gulf of Mexico, it seems very apparent that the Neolithic spread via cultural diffusion and not by migrations. It's plausible that it was often the case in the New World, not sure why but guess that lack of livestock may have been a partial reason (meat was still procured mostly through hunting or fishing, or in Mexico from extreme resources like flies or even human flesh). At least in South America they had llamas and cavies but in the North all they had was turkey, and only in some areas.

    2. Re: R1b found along the Volga - yah, I realize that it's a sister lineage to the main ones in Europe. I'm just saying that could mean that PIE was influenced by a different branch of the same Vasconic family.

      Re: Vennemen - worth noting that Sardinia and the Balcans are hotspots for I2, another Paleolithic haplogroup in Europe. And northern Italy is mostly R1b too I thought?

      I'm not saying the influence is absent from the non-R1b areas though, just that it seems strongest there.

      I strongly suspect that Vasonic is somehow tied to this hunter-gatherer resurgence in the later parts of the Neolithic. I don't think it necessarily follows that these folks spoke a hunter gatherer language though.

  28. The Vasconic root *iber- meaning river or place where a river meets the sea: clearly related to 'Aber-' and 'Inver-' in Celtic languages, 'ibaiaren' in Basque, etc. However 'baram' = 'river' and 'bahar' = 'sea' in Kanuri; 'ibhar' and 'ibhr' = 'sea' in Tamazight (and 'ibuhayra' = lake). 'Bahr' = 'sea' in Arabic. 'Biyaha' = 'water' in Somali. And river names all over Africa which sound similar: Cal-abar; Tar-aba; Ibra; Ipera (in Tanzania - similar to the Ippari River in Sicily?); Mbere-shi; Ma-mbere; M-bari; Abara; We-mbere; Ru-ipa; Iti-mbere; Mbere; Nana Barya; Atbarah; Ibra; Ok-para; etc. In my opinion the origin is African, and I'm more inclined to opt for an entry into Europe via Iberia and/or Sardinia than via the Balcans.

    1. IBAR or IBER has nothing to do with the sea. It means "river bank", i.e. the parts of the riverine ecosystem mostly used by humans. In Spanish would be "ribera" or "vega" but I don't think English has a single word for that nor that the true equivalent terms "bank" or "riverbank" are used with the same commonality as it is in these languages.

      IBAIAREN is not a word that can be considered as such: it is a declension of the root word IBAI (river) + -A (the) + -REN (of), hence: "of the river" or "the river's". You must consider IBAI and not its declined variants.

      As for the allegedly related words, I'm only willing to consider the most similarly sounding one, which is 'ibhar' (sea) (Berber).

      It is strikingly similar but it would seem easier to relate to Arabic 'bahr', which you also mention and has exactly the same meaning. Considering that IBAR is most likely a derivative of IBAI or at least of a common root IB(a)- (also IBON = creek) and that the -AR part is clearly a suffix of unknown original meaning, I think that we are before a false cognate.

      Notice that 'sea' in Basque is ITSASO (possibly related to Gr. THALASSOS), where the root is the fossil "water" particle *IZ- (-aso is clearly a respect suffix, used for ancestors: ARBASO, and grandparents: AITASO, AMASO, and hence ITSASO probably refers to a deity, the one the Greeks called Okeanós, which was not different from the actual body of water).

      The only way to discern if there is a connection with some of the other similar-sounding names (Ibra, Ipera are the only ones that may have the root IB-) would be to understand what meaning they have in the local languages.

      The IPPARI river in Sicily looks like a very good candidate for Vasconic origin, as it is clearly in the region where other similarly sounding river names (Iber-us, Tiber, Ibar, Hevros) appear. However I must insist that, at least judging on Basque, those names do not mean exactly river (the physical water course, which would be IBAI) but rather the lands around the river, its banks (IBAR).

    2. "... I'm more inclined to opt for an entry into Europe via Iberia and/or Sardinia than via the Balcans".

      Well, I have shown that the term IBAR (river bank) as fossil river name is present in all the Mediterranean, from Iberia to the Balcans (or vice versa):
      → Iberus (Iber originally, now Ebro by Indoeuropean deformation of Iberus: loss of median vowel and regular -us → -o Romance change) in Iberia
      → Tiber in Italy (probably T + Iber, where T- may well be an archaic IE article (like Eng. "the") or, else, like the French particle D', as in D'Artagnan, where Artagnan is the Basque surname Artiñan(o), Dordogne, where Ordogne is the Basque toponym Urduina, probably "beautiful water": ur duina or "who/what that has water": ur duena.
      → Ippari in Sicily (your mention, surely Ibar + -i, Italic plural nom., hence "the river banks" plausibly).
      → Ibar in Kosovo, oddly enough the best preserved form.
      → Hevros (Greek name, known in Bulgarian as Maritza), which seems oddly similar in its IE-induced evolution to Iber → Iberus → Ebro, hence presumably: Iber/Ibar → *Iveros/*Ivaros → Hevros.

      This is only the tip of the iceberg. I have already mentioned Serbo-Croat 'gore' (up), nearly identical to Basque 'gora' (up). I also mentioned Gr. 'thalassos' (sea) and Basque 'itsaso' (sea) but this one requires more extension:

      1. ITSASO, probable cognates GREEK THALASSA (not -os, my bad), Germanic SEA, ZEE, etc. (notice that the Germanic proto-word is probably incorrect and that the IE rooting is clearly forced and most unlikely). In the first case we notice the similitude mostly from the suffix -ASO (-os is just Greek), in the other it is the root *IZ-, probably in the basic nominative form of *IZA or *IZE, the connector, while the respect suffix -ASO is either lost or was never used.

      Back to the Greek THALASSA, there is clearly an extra element -al- when comparing with ITSASO. Maybe it was the word ahal = might, power, can (verb). If so the original Vasconic word would be *IZ-A(ha)L-ASO(-A), i.e. "the venerable power of water".

      .... (continues below)

    3. ... (continued from above)

      2. OIKOS (often pronounced EKOS, root of our common eco- particle, as in economics, ecology, etc.), meaning "house" or "home" is similar to Basque ETXE. The X is probably a diminutive or appreciative element so we can think of a root *ETZE instead, plausibly related to the verb ETZAN (to lay down).

      3. OKHI ("no") and NE ("yes") are totally non-IE Greek words. OKHI can be compared with Basque EZ, following the same kind of phonetic change as *ETZE. This seems totally regular, what is always a plus. NE is a bit more sloppy but I'd propose it is related to Basque BAI (by a separate derivation could also be related to Slavic and Romanian DA: "yes", which in turn might be related to Basque DA: "it is"). Greek is the only IE language AFAIK that does not use N- (ne, no) for the negative particle but does instead for the affirmative one, extremely strange and most likely a pre-IE legacy in any case.

      3. GAIA (earth, the goddess Earth, ultimate ancestor of all Greek gods) can only be understood using Basque, where GAI (nominative GAIA) means "substance, matter" but also has a second important meaning of "potential, capability": GAI IZAN = "to be able", -GAI = suffix with the approx. meaning of "will be" or "would be", i.e. EZKONGAI = bride or groom, fiancé/-ée (from EZKON-du = "to marry", in turn probably related to *ETZE and ETZAN as per above). IMO the documented name MARI for the equivalent Basque Goddess is a Christian-influenced pseudonym and the original name was surely GAIA. The presence of similar "gender-dual monotheism" beliefs in places as remote as India (Sakhtism, some Shaivite variants) and even East Asia (Taoism) strongly suggests that this type of cosmology was widespread among the Neolithic peoples of West Asia.

      I had another Greek word in mind when I began writing this but I totally forgot which one was it. If it comes back, I'll post an addendum.

  29. Again, please do not underestimate the probability of perceiving a resemblance between two words that entirely lack any real connection.

    Greek οἶκ- vs. Chinese 屋 (Mandarin wū, Cantonese ŋuk1 ~ uk1, Sino-Vietnamese ốc, Sino-Korean ok, Sino-Japanese woku > oku) "house; room; building, shelter" [In Japanese, this morpheme tends to mean more narrowly "roof," perhaps due to interference from the semantics of a native Japanese morpheme, ya, that is often transcribed with this Chinese character.]

    Ne also means "yes" in Korean. (Ye also means "yes" in that language, but there is historical precedent for a sound change *ny > y in Korean, whereas there is neither any Korean precedent nor any general phonetic motivation for a sound change of *y > n. Therefore, it is likely that Modern Korean ne and ye "yes" descend from an earlier ne or *nye. Please note that Korean also has an ~ ani "not," anio ~ aniyo ~ anyo "no," so there is no reason why /n/ may not appear in both affirmative and negative expressions even in one and the same language.)

    -gai ~ -gei > -gae ~ -ge is also a v. > n. derivational suffix in Korean: e.g. norigae "toy" < nor- "to play" (also norae "song" < *norghai < *nor-gai), nalgae ~ narae "wing" < nar- "to fly" + -gai, jige "a traditional Korean contraption to help with carrying objects on one's back" < ji- "to carry on one's back; to owe, to be in debt; to take responsibility." (The ai ~ ei > ae ~ e variation in vocalism of the suffix, which is now ignored in the pronunciation of most native speakers of Korean at least when they are speaking the "standard language" and not some local dialect, is a trace of so-called "vowel harmony.")

    1. "please do not underestimate the probability of perceiving a resemblance between two words that entirely lack any real connection".

      Neither should you. That's why I generally prefer to use mass lexical comparison, which provides statistical results that allow us to go beyond coincidences (too many coincidences are probably not any coincidence anymore but something real). In the case of Oriental languages I have already addressed the matter of the Sino-Caucasian hypothesis this way:

      The result is a clear "no way", at least for Basque - Sino-Tibetan.

      In general I remain extremely skeptical of macro-families spanning both East and West Eurasia, except where a Siberian connection is probable. For example a relation between Uralic and Altaic is still possible but inclusion of Indoeuropean instead seems harder to imagine and the relation should be rather understood as sprachbund (the Indo-Uralic one certainly but maybe also influence of IE into Altaic in the Chalcolithic and Bronze ages).

      A curious cultural difference between Western and Eastern Eurasian traditions I spotted some time ago, while following critically Roslyn Frank's studies on the alleged bear traditions she has proposed for North Eurasia, is that, while Western traditions do all consider Ursa Major to be a bear (mother bear), in Eastern traditions they tend to be explained in terms of water features (animals or others) instead. Uralic mythology fits best with the Eastern tradition, as for them it was traditionally a salmon weir, breaking apart of the nearly uniform Western identification of the constellation as a bear.

      I was telling Roz: well, you say that this bear mythology is a Northern Eurasian tradition but Uralics don't follow it and instead ancient (pre-Hellenistic) Hebrews did. It seems to me rather a West Eurasian tradition with nothing particularly "Northerner" about it.

      The reason why she thinks it is a Northern tradition is because it is also present among Native Americans, however this may be because they have partial Western ancestry via Altai.

  30. it may be a coincidence, since some of the words are similar to Tagalog (buhok/face, Hangin, wind, Lupa/dirt.

    But one could of course argue that this was the Spanish or Arabic/Semetic influence on the language...

    1. They don't look either Basque nor Spanish to me. I don't know where you see the similitude. Coincidences in sound between languages, much more striking than the ones you mention, do happen: there are only so many phonemes. The question is if they are many or just a few random ones: the latter case is normal and should be considered statistical noise.

  31. Without taking up your time (and space on your blog) , thanks for your post on similarities between Basque and Nubian. It inspired me to do some research. I was shocked to discover not just striking similarities between Vasconic languages and African languages, but actually far more striking similarities between Indo-European languages and African languages (Celtic, Italic, and Germanic groups in particular). I don't know if this is due to Vasconic substrate or if it is actually an inherant feature of those groups. Further research is warranted.

    1. My recommendation: try to stick to a pre-defined list (such as a minimally modified Swadesh list) and count the apportion. Low frequencies of matches mean nothing (or maybe some words are actually wanderworts but hard to demonstrate), you need high frequencies to feel reasonably certain about a relation.

  32. Some preliminary findings. (Note also that Celtic languages and Fula language have in common consonant mutations). Many of the Fula words have a suffix (eg. -gol, -ugol); so try to disregard the suffix and look at the main part of the word to observe the similarity with the Celtic words. Examples included from Proto-Celtic; Breton; Cornish; Welsh; Old Irish; Irish; Scots Gaelic; and Manx.

    Celtic Languages----------------------------------------------------------------------------Fula Language (English)
    *mi/me/my/mi/mé/mé/mi/mee----------------------------------------------------------mi (I)
    *kʷid/cid/cad e, céard/ciod--------------------------------------------------------------koɗun (what)
    *māros/meur/meur/mawr/mór/mór/mòr/mooar-----------------------------------mawɗo, maw- (big)
    *tegus/tiug/tiugh----------------------------------------------------------------------------tekki (thick)
    gwaz, gour/gwas, gour/gŵr-------------------------------------------------------------gorko (man - adult male)
    ben/bean chéile/bean-phòsda/ben, ben phoost, ben heshey--------------ɓeyngu (wife)
    gwaz/gour/gŵr-------------------------------------------------------------------------------gorko (husband)
    *tangʷat-/tengae/teanga/teanga/çhengey-----------------------------------------demngal (tongue)
    *mīlom/mil/mil/mil/anmandae, míl---------------------------------------------------mardi marle (animal - plural)
    *kū/ki/ki/ci/cú/cú/cù----------------------------------------------------------------------kutiru (dog)
    *angʷīnā/ingen/ionga/ìne/ingin--------------------------------------------------------fede ngo (fingernail)
    cos/cos/cas/cass-------------------------------------------------------------------------kosngal (foot)
    *koxsā/coes/cos/cos/cas/cass------------------------------------------------------kos ngal (leg)
    arogli/arogl-----------------------------------------------------------------------------------urgol (to smell - sense odor)
    mervel/merwel/marw---------------------------------------------------------------------maayugol (to die)
    do tuit/tit/tuit/tuitt--------------------------------------------------------------------------tettoygol (to fall)
    *louχsnā/loar/loer/lleuad, lloer--------------------------------------------------------lewru (moon)
    yen, yein/yeyn-----------------------------------------------------------------------------jangol (cold)
    *bānos/bán/bán/bane--------------------------------------------------------------------danejum (white)
    i fat/fada/fada/foddey---------------------------------------------------------------------ɓadaki (far)

    There were numerous others but I though these few stood out somewhat. More research is of course required.

    1. Is this part of a mass lexical comparison exercise? Seems so. How do other IE languages perform in contrast? Most of the words in that list are clearly IE and can easily be found cognates in Italic or Germanic. So are you implying that Fula, and by extansion all Niger-Congo maybe, is related to Indoeuropean?

      A few words like gwaz are arguably vasconic though (compare with gizon but some will argue form gŵr <> vir instead).

      In any case there are some apparent cognates that I find very unlikely like bán - danejum or mervel-maayugol.

  33. I am having a go at this in my limited spare time. I guess you can call it a mass lexical comparison exercise. I am comparing different Indo-European languages with different African languages. I haven't got very far with it so it is too early to draw any conclusions. However I can say that I found many cognates shared between Romance languages and Lingala/Kongo.

    1. I ask to find out what's the percentage. AFAIK under 10% or so it can only be considered noise. But if it's around 20% or more then it's probably relevant.

  34. I did a mass lexical comparison between the Romance languages of various regions of Italy and the Bantu languages of Africa. I found that the Italian Romance languages share around 20% of the words on the 207-word Swadesh list with plausible cognates in a small group of Bantu languages of Congo and Angola [source: Wiktionary Swadesh lists]. When I went all-out and compared the Italian Romance languages with ALL of the Bantu languages used on the Wiktionary Swadesh lists, the figure rose to around 40% of the 207 words [or 33% using a more conservative approach]. If you want to see my results I can show you here or can email them to you.

    1. Now, that looks intriguing, really. Could you share your list, so we can double-check? Upload it to Google Drive for instance and give us the link.

      Is this particular of Italian Romances or general of all Romances (i.e. did you compare with other romances like French, Portuguese, Sardinian or Romanian for control?, also what about classical Latin?) Also what mechanism would you suggest for such an unlikely connection?

  35. I'll share the list at the weekend when I have some free time again. I only started with the Italian Romance languages, my plan was to do the other Romances next, then Celtic, Germanic, Slavic, etc. as and when time permitted. Classical Latin is included in my set of Italian Romance languages. In terms of a mechanism: some scholars claim ties between Ancient Egyptian and Bantu languages [see: ]; Ancient Egyptian had some influence on Ancient Greek and Latin. Ancient Greek itself influenced Latin. Latin influenced swathes of Europe. Greek influenced southern Italian languages. Then of course you have identified a potential influence of Nubian languages on Basque/Vasconic; it may be the case that there could also have been a Bantu-like influence via the same route, which remains as a substrate in some European languages. Possibly some influence on Etruscan ['Southern Vasconic']? And if there was Bantu influence on Afro-Asiatic languages, then we could also consider, for example, Cardial Pottery culture, Phoenicians, Arabs etc. as other possible vectors of transmission. Here are some links to other relevant sources:- "The Syntax and Prosody of Focus: the Bantu-Italian Connection" [María Luisa Zubizarreta, University of Southern California]:
    "The Bantu-Romance Connection: A comparative investigation of verbal agreement, DPs, and information structure (Linguistik Aktuell / Linguistics Today)" [Cécile De Cat (Editor), Katherine Demuth (Editor)]:
    "DP in Bantu and Romance*" [Vicki Carstens, University of Missouri-Columbia]:

    1. Bantu expanded only since the Iron Age. In fact Bantu is a linguistic family much like Romance or Germanic, or Austronesian also, where there is still some mutual inteligibility across languages because the divergence only happened "yesterday". Afroasiatic on the other hand is probably the oldest identified linguistic family, going back to some 12 Ka ago, maybe more, a

      So it seems impossible that Bantu could influence Afroasiatic as such set because it would have to be an influence in a time when Bantu did not even exist as such. Actually the beginnings of Bantu expansion are not older than Rome, although Bantu is part of a larger family: Niger-Congo, which maybe you may want to study too.

    2. Also re. your first link: prosody is a most obscure subject, it basically reflects how you intonate (intonation, stress, rhythm and tempo), what may be affected by emotion but is also part of the characteristic accent, which varies a lot. I find hard to study the subject without a "musical" approach (i.e. sound files or sound graphs involved), what Zubizarreta does not use. Also it could be random convergence as her categorization only allows for a small number of possibilities.

      Gimme vocabulary, please.

  36. Here is a link to Google Books, so that you can see some of the "Bantu-Rommance Connection.." book without the hideous $173 price tag!

    1. A bit complex grammar analysis, some pages are missing anyhow. Gimme vocabulary, gimme, gimme!

  37. I have been inputting the data on to a Google Doc this weekend [was all written down on paper]. However it is taking rather longer than I had anticipated! [Not helped by endless distractions and the rather mind-numbing and repetitive nature of the task]. I am almost half-way through now so I didn't finish it this weekend as planned. As soon as is finished I will post the link on here.

    1. Ok, no hurries. I must admit I did not expect the work to be on paper, just copy-pasting all the words must have taken so long! You could always scan the original, (as long as your handwriting is readable enough, I guess).

  38. I finally had some time to get the Italian Romance languages/Bantu languages lexical comparison spreadsheet finished today now that I am not at work:-

    1. You should change format to public access, Chris. I cannot access the sheet (sent you a request for permission but surely better to put it open, so others may access it too).

  39. Thanks for pointing out.

    Link sharing is now on, here is the link:

    1. Got it thanks. First thing I noticed: chisto and shinto can't be cognates: chisto, pronounced /kisto/ is pronunciation variant of questo, which is even more clearly not the same. Both end in -to but otherwise no relation. Haven't got further yet.

    2. Thanks for your comments. That one probably ought to have been left off. Let me know when you've had a chance to read through the rest, cheers.

    3. I'm sorry that I have not got a proper criticism of your effort but I did try once and my impression was that all can be reduced to Latin (almost nothing new in the dialects, even if you may have not noticed) and the Latin-Bantu apparent cognates seem to be at most half of the ones you claim. That's my impression, not a too deep one admittedly.

    4. Thanks. Admittedly I threw everything in there, even the most 'oblique' cognates. I was looking for some rigorous, skeptical analysis. If half of the alleged cognates I included still stand, then that still equates to 20% of words on the 207-word Swadesh list between Latin and Bantu languages. There ought not to be any cognates, bar the odd one or two by pure coincidence. If the words in question are associated with Latin but not any other Italian dialects, then I need to consider other language groups which influenced Latin, eg Celtic dialects of N.Italy/Gaulish; Etruscan; and Greek.

    5. I'm gonna make a criticism on the fly of your list. First of all all or nearly all of the Italian dialectal words are directly derived from Latin, often in ways that negate your comparisons. For example "chisto" is obviously the same as Italian "questo", which comes from Latin "que istum" ~ "such as this". Spanish for example keeps the original form "esto" (this), without the added "que" (and, so) particle. Similarly tuttu, etc. are Lat. totum, óm is Lat. homo, and so on, and so on.

      So I'm only considering the Latin list in the follow up comment(s).

    6. Plausible: ego:ngai; ubi:wápi, tres:tháru, macer:móke, piscis:mbisi, flos:fulele; kutis:mukate; dens:dinu; unguis:ngala; genu:kinu; collum:umqala; pectus:petu; cor:ngoro; secare:seha, -sika; tergere:-giria; fluvius:umfula; terra:tiri; caelum:likulu; mons:mongo;

      Dubious: ligare (from "liga"=league, union):línga; tumere:-tumba; caelum:zulu (does not follow the logic applied elsewhere /k/=/k/ and not /k/=/z/); cineres:-cina (not same meaning, only one language); niger:nduma njiru; sinister:sinzohoto; cum:ka, kunye; si:te, se; nomen:nkombo

      Clearly or likely wrong: omnis: -onso; pauki: mokemoke; quinque: -sanu; homo: moto; folium: rifu, lisafu; manus:oma; olere:kutola; pugnare (from pugnus=fist):ku-ibiga, -pigana; lakus:li-chibi; sal:umucele; umidus:metsi,

      Wrong with explanation: "maritus" derives from "mas" (male) so NOT cognate of murume; mater: mama and pater: papa are "nursery words" and I systematically take them off from any comparison, as they are nearly universal; "animal":náma, "animal" derives from "anima" ("breath", "life", later also "soul"), so hardly appliable; oculus:ijisho (actually only plausible because it fits with Eng. "eye", both derived from the PIE root *h₃ekʷ-, not with oculus as such, compare also with Basque "begi" also); viridis (from "vireo": to sprout):-biriwira (green is not a basic color in many languages, blue, black, red, white are the most universal ones); bonus:-bwino (bonus derives from duenos → duonus → b(u)onus, looks similar but not if you consider the known and well documented etymology); propinquus:penepene (not just there is no clear sound similitude but prope = near, propinquus = "near hence");

    7. Total plausible: 19/207=9,2%

      I think it's below the threshold, particularly because they are cognates in ANY dialect, which in some cases (petu?) could even be Romance borrowings (from Portuguese). If you still think there's something to it, I'd recommend comparing Latin, PIE and Proto-Bantu (and/or proto-Niger-Congo, as Bantu is just a derived subfamily) directly in order to minimize noise and contrast if it's PIE or Latin the one with more Bantu apparent cognates. If you can't find such "proto-word" lists, choose a limited set of representative dialects rather than all of them (the more dialects compared, the greater the chance that a word sounds similar by mere fluke).

  40. Hi Maju, are you aware of this?

    1. No, I wasn't. I must say that linguistics is not my favorite field: too complex and full of uncertainties, not to mention the many amateurish rantings, sometimes even by people who claim to be titled linguists. But I'm bookmarking the link and will possible consider it at a later moment. Thank you.

    2. Linguistics doesn't appear to have a very rigorous, scientific, approach in comparison with, say, modern-day archaeology.

    3. It is a very slippery terrain indeed. There are some methods (comparative, internal reconstruction) that have more prestige but they are not perfect anyhow, and also require of long dedicated careers to produce quality results.

      The mass lexical comparison that I used above is often considered "unreliable" but can still serve as first exploration, because it is certainly much much faster. With speed you lose quality but it can serve to discard the less likely paths and suggest the most likely one, worth further research. Greenberg demonstrated its usefulness when grouping the, then ill-studied, African languages. Some of his proposed families have been revised and refined since then but the essentials are owed to his work with mass comparison. The most criticized issue is his proposed Amerind super-family, which mainstream linguists don't accept, however I am personally convinced that the family must exist (because archaeology and genetics oblige), just that it is too old (c. 17-20 Ka) to be reliably identified with the criteria that most linguists would be able to accept (the limit of reliable identification is around 10-12 Ka probably, which is the age of Afroasiatic). So I still think that mass lexical comparison is a valuable tool, when properly used and always subject to secondary revision, as everything is in science.

  41. Thanks for your honest review of my Romance/Bantu lexical comparison Maju.

    There was one word which I forgot to add to my list.

    English: walk

    Latin: ambulāre

    Lingala: -tambola
    Zaramo: ku-tembela
    Kituba: kutambula
    Luganda: okutambula

    [Zulu: -hamba]
    [Xhosa: -hamba]
    [Swahili: -tembea]

    I realise that this doesn't make any difference, but the resemblance is clearly there.

  42. Perhaps urum (black) - ilun (dark) is more clear than urdin, as beltz comes from beles, a likely loanword with PIE root *bhleg (to burn). You can see listed in Bronze of Ascoli (I BC) some Ebro valley hispanian riders with *beles names and surnames, all of them from iberian and vascon soils.

    There is a guy named AGIRNES BENNABELS F., ie, AGUIRNES (Aguirre?) son of BENNABELS (true black?).

    1. Arguably "ilun" might come from (h)il + une = zone of death (???) or more likely "il-du(e)n(-a)": who/what has the dead ones (ildun(-a) → ilun(-a)). And might also be root of Lat. "luna" (moon), via the basic nominative form "iluna". Basque "hile" and "hilargi" (moon) also seem to come from (h)il (to die or to kill). I don't see any strong relationship with "uri" or "urum" ("black" in Nubian). IMO "urdin" ("blue", anciently "gray" and "green" also in Basque) either comes from that or comes directly from "ur" = "water"; the logic is strong for this last derivation: "ur" (water) produces or is strongly associated to "urdin" (watery?) color (i.e. the "cold" palette).

      BTW I mentioned before something about "gorri" ("red", all warm colors originally) being related to gor (deaf) or gorroto (hatred) but I was told later that it's almost certainly related to English "gore" and they should derive from a root *gor- meaning "flesh" or "meat". There was a deeper etymological explanation for that (trust my veteran linguist friend) but I don't recall all the details right now.

      Beltz should come from "bela" or "bele", maybe originally "bel", and meaning "crow, raven". Bel(-e) → belez (proper of the crow, cf. also Spanish surname Vélez and related Velázquez, which are patronymics from Vela/Bela = crow, and Velasco/Belasko = little crow) and then belez → belz → beltz. Beles, found in some Aquitanian slabs, is probably already Belez/Vélez, i.e. "of the matter of Bele/the crow" or actually meaning "son of Bele/the crow". This -(e)z Basque suffix means "made of" (metal → metalez, plastiko → plastikoz, etc.) and has become a generalized Ibero-Romance method for patronymics, probably affecting first Castilian but reaching as far as Portuguese eventually.

      I can accept "Agirnes" as "Agirre" (but it could be something else and some have argued that Agirre is a medieval/early modern surname of former professional soldiers derived from French "guerre" or Spanish "aguerrido" = brave, ultimately both from Germanic "war" or "wer"). I certainly accept -BELS as "beltz", but what is BENNA? Double "n" could be the "nj" ("ñ", "nh", "ng", "ny") sound (NN → Ñ is a historical thing, the bar or tilde represented a double letter in monastic script, they had to save in parchment space), so I would think of "beina" or "behina". "Behin" is "always" but it sounds similar enough to "behi" (cow), so could it be just "black cow", just as AHERBELS is "black he-goat"? So I'd read it as Agirnez Behi(na)beltz. Alternatively it's not "beltz" (belz) but nabelz and "nabel" or "nabal" sounds like "nabar" (brown), so could it be behi-nabarr-ez, i.e. "son/daughter of Brown Cow"?

      This brings me back to Agirnez, what if it's not Agirre but something else? "Agiri" is "document", "agian" = "seemingly", "agirne" might be something like "scribe"? Again the ending in -s (-z) seems a patronymic (son/daughter of the scribe?) May surname Agirre of mysterious etymology mean just "scribe"? In any case it'd be Agirne → Agirre and the -s/-z ending is clearly a patronymic (at least to my eyes).

    2. Erratum: "hil-une" would rather be "zone of the dead (ones)", better translation (hildu = to die/kill, true, but hila = the dead (one)). Hilobi = nest of the dead = tomb (another example).

    3. Hi Maju, you forgot heart Biotz (basque) and life Bios (Greek)

    4. I did not "forget": it was not the matter under debate, was it?

      However you bring up a moist interesting likely Vasconic "infiltrator" via ancient Greek. I would however argue that it is not "bihotz" (heart) but rather "bizi" (to live) and maybe even a root particle *bi- (because the -os ending looks like standard Greek suffixation, which should be removed from analysis).

      One could extend this but getting quite speculative to the number two ("bi"), which fits with Basque permanent-creation-through-sex underlying mythology (not dissimilar from some deeper variants of Hinduism, which don't recognize the IE gods but only the Sakhti-Shiva couple, and not dissimilar from the Gaia-Eros beginning of everything in deep Greek myth either). Also some authors have noted that some key body parts' names suspiciously begin with the letter b-, notably "begi" (eye), "buru" (head) and the already mentioned "bihotz" (heart). These are not random body parts like fingers, lungs or the belly: these are the parts most directly associated to "soul" or "anima" (animal, animation) even in our modern non-vasconic languages.

  43. In this long brainstorming, that has gone from Nubia to Greece and the PIE, I remember that the basque Izar always brings to my memory the name of the Godess Ishtar, the babilonic Venus, represented by the morning star.
    Don't basques sing to Yirgin Mary: Goizeko Izarra, gustiz argia?

    My regards

    1. Of course "izar" (star) and Ishtar (Venus) might be related but I'd say that PIE *h₂stḗr (star, allegedly with a literal meaning of "shinier") is even closer, always assuming that the name refers to Venus the planet (as seems to be the case from iconography). So the matter is blurry and may well be a case of wanderwort in a context of international astronomy (and related astrology, religious syncretism, etc.) It's possible that pre-proto-IE (or part of the pre-steppe roots of PIE) may still have originated in West Asia (judging on genetics) so the notion may well be part of a very ancient group of shared roots maybe going all the way to Göbekli Tepe and such.

      "Don't basques sing to Yirgin Mary: Goizeko Izarra, gustiz argia?"

      Sounds vaguely familiar but I'm not into Christian religion enough to say. My impression is that this is a common Mediterranean concept and that you will probably find the same notions in Latin, Greek, etc. For example look up "Stella Maris", which is a mariner expression associated to Christian Mary but also to other older goddesses like Isis.

      Basque Mari (IMO originally Gaia, Mari would be a syncretic renaming under Christian influence) is not associated with anything celestial but rather with the underground, as is her male mate Sugaar, the Snake-God. They may "herd" clouds in the sky, they may manifest as meteorological phenomena, their mating in high mountain caves is said to produce the fertilizing (although sometimes scary) storms (their son Odei: the storm cloud) but in no legend they are related to astronomy (even the Sun and the Moon are said to be "daughters" of Earth or the Underworld, to which they return every day, once again following the chthonic principle). So with Mari/Gaia/Sakhti and her partner Sugar/Eros/Shiva we are before an Earthly religion in which the sky is only an epiphenomenon: they sneak into IE mythologies such as Hel & Jormungandir in Nordic one, Python and Demeter in Greek one, etc. but always as defeated, assimilated or antagonist characters (or in the case of Gaia as venerable but nearly irrelevant great-grandmother). Where they probably survived in somewhat better shape is in some variants of Hinduism (Sakhtism and some Shaivite variants that do not recognize Vishnu or other gods but as "aspects" of Shiva), which are quite clearly pre-IE in their tenants.

      Also for the record: Venus is the only planet that retains a native name in Basque and that one is Artizar (star of in-between, i.e. dawn/dusk star).

    2. PS- Re. Astronomy/Astrology, it's in Greek mythology associated to Uranus, son of Gaia and grand-father of Zeus, castrated by his son Kronos (event that produced the giants and Aphrodite-Venus). This may well be the same as Basque Urtzi (also Ortzi, Ost, recorded in the Middle Ages as "Urcia": "et Deus vocant Urcia"). Urtzi seems to be related to "ur" (water), just as "urte" (year) is (the exact meaning of the suffixes cannot be discerned clearly but maybe both from "urti" = prone or related to water).

      There's no mythology associated to Urtzi but it clearly had an influence in the name of many things meteorological (ortzadar = rainbow, oskabri = clear sky, oskorri = dawn, ostarri = lightning) and calendaric (osteguna = thursday, ostirala = friday). It seems to me than rather than referring to "a god" it meant originally just "the sky" and that was later "humanized" by influence of other religions such as Celtic, Roman or Christian ones. It's plausible that it was also the same with Uranus in the Greek mythology.

      Greek would be Ouranos, where -os is standard male suffix like Latin -us, so *Ouran- would be it originally. How well do Urtzi/Ortzi/Ost and *Ouran- fit? Etymologically I would not dare but conceptually they do seem related, as Uranus is the night sky and is directly associated with Astronomy, blending with Atlas (the Western or Iberian god that held the sky) in some obscure legends (and is depicted in rock art, make an image search for "Indalo", a modern pseudo-romantic renaming of the ancient icon).


Please, be reasonably respectful when making comments. I do not tolerate in particular sexism, racism nor homophobia. Personal attacks, manipulation and trolling are also very much unwelcome here.The author reserves the right to delete any abusive comment.

Preliminary comment moderation is... ON (your comment may take some time, maybe days or weeks to appear).