Recent Focused Series »

Indo-European Origins
Siberia
Northern California
The Caucasus
Imaginary Geography
Home » Featured, Indo-European Origins, Linguistic Geography

Do “Ultraconserved Words” Reveal Linguistic Macro-Families?

Submitted by on May 10, 2013 – 1:58 am 203 Comments |  
Today’s post takes on a recently published article by Mark Pagel, Quentin Atkinson, Andreea Calude, and Andrew Meade entitled “Ultraconserved words point to deep language ancestry across Eurasia”, published in PNAS. First, Asya Pereltsvaig examines the article from a linguistics point of view, and then Martin Lewis considers it from a cartographic perspective.

 

Part 1: Linguistic Critique

Can words remain recognizable across more than a dozen millennia, their meanings understandable by people speaking languages in diverse linguistic families? Mark Pagel, Quentin Atkinson and their co‑authors answer in the affirmative. Journalists and bloggers, moreover, have tended to interpret their study as indicating little change in the core vocabulary of a massive assemblage of languages, those found in the supposed “Eurasiatic super-family” tying together Indo-European, Uralic, Altaic, Kartvelian (Georgian), Inuit-Yupik (Eskimo) and Chukchee-Kamchatkan.* Mark Frauenfelder at Boingboing.net reported that “a research team led by Mark Pagel at the University of Reading in England has identified 23 ‘ultraconserved words’ that have remained largely unchanged for 15,000 years”. Tia Ghose, LiveScience Staff Writer echoed with “The researchers could predict what 23 words, including “I,” “ye,” “mother,” “male,” “fire,” “hand” and “to hear” might sound like in an ancestral language dating to 15,000 years ago”. Even venerated publications could not avoid sensationalism. In the Washington Post, David Brown, claims that the following passage consists largely of such “ultraconserved words: “You, hear me! Give this fire to that old man. Pull the black worm off the bark and give it to the mother. And no spitting in the ashes!”. Brown further contends that:

“… if you went back 15,000 years and spoke these words to hunter-gatherers in Asia in any one of hundreds of modern languages, there is a chance they would understand at least some of what you were saying.”

Sorry, but there is no such chance. If you go back a mere two thousand years in any preserved language, changes have been enormous. If you go back 15,000 years, all comprehension would vanish. Within any one “Eurasiatic” family, which diversified much more recently, speakers of languages in one branch are seldom able understand anything in Brown’s passage if expressed in a language of a different branch.

Consider, for example, a direct translation into Russian:

Vy, uslyshte menja! Dajte etot ogon’ tomu staromu muzhu. Potjanite chjornogo chervja s kory i dajte jego materi. I ne plevat’ v pepel!

Yet the authors claim that such ultraconserved words are found across all of the branches of the hypothesized “Eurasiatic” family. It goes without saying that the mutual comprehension of this passage would be nil between speakers of Georgian, Chukchi, Sakha, Tamil, and Udmurt, to name just a few languages in this hypothetical family.

Brown’s article in The Washington Post continues:

“That’s because all of the nouns, verbs, adjectives and adverbs in the four sentences are words that have descended largely unchanged from a language that died out as the glaciers retreated at the end of the last Ice Age. Those few words mean the same thing, and sound almost the same, as they did then.”

Just how wrong this claim is can be seen from the example of just one of the 23 “ultraconserved words: man. First consider its history merely within English over the past millennium and a half. Today, it is pronounced /mæn/ and means one of two things: either ‘an adult male person’ or ‘a person of either gender’. The latter meaning, however, is considered sexist by many, and is thus falling out of use. Words such as chairman, fisherman, and policeman are thus being replaced by such gender-neutral forms as chairperson, fisher, and police officer, just as mankind is yielding to humankind. But as the gender-neutral meaning of man is still evident in manslaughter and in the phrase no man’s land. As it turns out, the meaning of ‘an adult male’ is relatively new. In Old English (roughly, prior to the Norman invasion of 1066), this word—pronounced then with a vowel articulated further back in the mouth—did not mean a ‘male person’ but had only the gender-neutral sense of ‘a human being, person (male or female)’. The word acquired the sense of ‘adult male’ in Middle English. Prior to that time, an adult male was a wer, as distinguished from a wif, which then meant ‘woman (of any marital status)’, as it still does in idiomatic expressions like old wives’ tale and in the compound midwife, originally meaning ‘with woman (during labor)’. The word wer began to disappear in the late 13the century and was eventually replaced by man, which retained its old, more general meaning as it acquired the new, gender-specific one. (The term wer did survive, however, in such terms as “werewolf,” which make one wonder whether a female lycanthrope should be referred to as “wifwolf”.) Note also that the Old English man had additional meanings besides ‘person’, including ‘servant, vassal’, as in all the king’s horses and all the king’s men (we retain this meaning to this day). Thus, clearly the meanings of even “ultraconserved words” show considerable change over much shorter periods than 15,000 years.

Pronunciations of such core terms change too, as I indicated above with the shift in vowel articulation in man through the history of English. Within the Germanic branch of the Indo-European family, the reflexes of the reconstructed ancestral Proto-Germanic form *manwaz include Old Norse maðr, Danish mand, Gothic manna. In other Indo-European branches we find Sanskrit (Indic) manuh, Avestan (Iranian) manu-, Old Church Slavonic (Slavic) mozi. The latter is related to the Russian form muzh, found in the Russian version of the odd “Stone Age” passage above. This plethora of phonological forms in related languages is a result of sound changes, different in each family.

The list of the “ultraconserved words” in the PNAS article itself contains quite a few surprises, even we restrict ourselves to the 1,500-year long history of English rather than the supposed 15,000 years of shared “Eurasiatic” history. Among those oddities are thou and ye, both of which changed their meaning (and form, in the case of ye), switching from informal to formal. Another surprise is not, a word of recent pedigree as a negative particle (Pagel et al. incorrectly call it an “adverb”). Not began its career in the mid-13th century as an unstressed variant of the emphatic noht/naht ‘in no way’, not unlike pas in the modern French two-part ne… pas negation. In fact, both English and French are undergoing the so-called “Jespersen’s Cycle” (named after a Danish linguist and Anglicist Otto Jespersen). In the first stage of this cycle negation is expressed by a single preverbal element; in the second stage, a postverbal emphatic element is added and made obligatory; and in the third stage, this postverbal emphatic element replaces the preverbal element, making the latter optional or eliminating it altogether. Thus, in Old English negation was expressed by a preverbal ne, as in ic ne seah (literally ‘I not saw’). In Middle English, the same sentence was expressed as I ne saugh noht (literally ‘I not saw nothing’). Finally, in Early Modern English (around the time of Shakespeare), this sentence became I saw not (eventually, lexical verbs stopped inverting around negation and the so-called do-support was introduced to give us the modern I did not see). In a parallel development, Old French had only the preverbal negation, as in jeo ne dis (literally ‘I not say’). In Modern Standard French both a preverbal and a postverbal element are obligatory, as in je ne dis pas (literally, ‘I not say nothing’), while in colloquial French, which represents Stage 3 of Jespersen’s Cycle, the preverbal ne is optional, so that je dis pas is perfectly acceptable.

All of these subtleties escape the authors of the PNAS paper, who ignore grammatical patterns and changes as much as possible. Even their assignment of some of the 23 “ultraconserved words” to “parts of speech” is flawed. For instance, they call the demonstratives this and that “adjectives”, though these words exhibit neither adjectival morphology nor adjectival syntax. For example, demonstratives are in complementary distribution with articles, quantifiers, and possessors, resulting in the ungrammaticality of *the this book, *every this book, and *John’s this book; whereas adjectives are perfectly capable of co-occurring with these elements, as in the interesting book, every interesting book, and John’s interesting book. Note also that Pagel et al. give different labels to who and what: according to them, the former is a “pronoun”, while the latter is an “adverb”. Yet, these two words clearly share the same syntactic properties (except for the demonstrative-like use of what, as in What book?). Both who and what must appear in the beginning of a question (e.g. Who did you see? and What did you see?). But only one of them can occur in the beginning if both are present, as in Who brought what to the potluck party? and What was cooked by who? (ignoring the who/whom distinction).

While these issues may seem trivial or irrelevant to the larger considerations of the PNAS paper, they underscore the central issue, something repeatedly missed or consciously ignored by these authors and their collaborators (cf. Gray and Atkinson 2003, Bouckaert et al. 2012, and elsewhere): to wit, language is not merely words. The interchangeability of “words” and “language” is a neat conjuring trick, evident in the first sentence of the article’s abstract (highlighting mine):

“The search for ever deeper relationships among the World’s languages is bedeviled by the fact that most words evolve too rapidly to reserve evidence of their ancestry beyond 5,000 to 9,000 y.”

Pagel and Atkinson’s search for family relationships among languages is set off course at the onset by looking in the wrong place. It has been understood at least since Antoine Meillet’s work a hundred years ago that grammatical properties are more reliable than words as indicators of familial relationships. As Meillet (1908: 126) noted “Les coincidences de vocabulaire n’ont en general qu’une très petite valeur probante” (“Coincidences of vocabulary are in general of very little probative value”). In recent years, the searchlight has been focused—by bone fide linguists, not evolutionary biologists—on abstract syntactic properties, establishing formal grammar as a population science; see, for example, the work of Giuseppe Longobardi and Cristina Guardiano (e.g. Longobardi & Guardiano 2009). Just as the biological classification of species, originally based on externally accessible characteristics, underwent a revolution on the grounds of progress in theoretical biology, namely the rise of molecular genetics, so too progress in the phylogenetic classification of languages must be based on progress in theoretical linguistics. In order to push the research frontier, we linguists need to identify the basic building blocks of language, its “atoms”, in Mark Baker’s memorable metaphor, and examine carefully how they play out in linguistic evolution. Looking for “words that survived since the last Ice Age”, in contrast, is a seductive but ultimately a futile enterprise.

 

Part 2: Geographical Critique

The map used by MaAtkinson2Map.tiff copyrk Pagel, Quentin Atkinson, Andreea Calude, and Andrew Meade in “Ultraconserved Words Point to Deep Language Ancestry across Eurasia” is riddled with odd features and elementary errors. To begin with, the projection is inconsistent and distorted; a misplaced Alaska appears to be no larger than Kamchatka. Bizarre features include a massive white area in northwestern Russia. Is this supposed to be a “non-Eurasiatic” zone or a massive lake—Baikal substituted for Ladoga? Neither possibility makes any sense. Why is Kashmir mapped as a separate country; is this a deliberate political statement, designed to infuriate India and Pakistan in equal measure? The mapping of northern Borneo (Brunei and the Malaysian states of Sabah and Sarawak) as if it were an island in its own right is amusing, as is the mystery island off the coast of Yemen.

The map is also littered with minor errors in linguistic geography, such as the exclusion of Vasconic (Basque), the confinement of the Central European Uralic zone (Magyar) to Hungary, and the incorrect placement of the Middle Volga Altaic (Turkic) zone, which is located too far to the east. Such small mistakes are easily overlooked, however, especially as the authors have provided an honest disclaimer: “The color-shaded areas should be treated as suggestive only, as current language ranges will not necessarily correspond to original homelands, and language boundaries will often overlap.”

A number cartographic errors, however, are far more serious, and hence demand recognition. Intriguingly, some of the basic mistakes that characterized the Bouckaert et. al. Science article that we so harshly criticized appear yet again, such as the inexplicable exclusion of Moldova from the Indo-European realm. Considering the email exchanges that we had with Quentin Atkinson about this and related issues, we are surprised to see this oversight recurring. Note also that Macedonia is likewise excluded from the Indo-European zone, just as Estonia is left out of the Uralic zone. More problematic is the fact that Moldova, Estonia, and Macedonia are not even mapped within the supposed Eurasiatic macro-family. What kind of languages are we to imagine are spoken in these countries?

anlmap

Errors outside Europe are equally serious. Turkic Azeri, with its 23 million speakers spread across Azerbaijan and northwestern Iran, is classified as an Indo-European language. The NE Caucasian, NW Caucasian, Indo-European, and Turkic languages of the north Caucasus are all misclassified in the Kartvelian family. Kurdish appears in southeastern Turkey, but not in Iran, Iraq, and Syria. Indo-European languages do not appear in either Kashmir or Tajikistan, yet they do on far northeastern India and northern Nepal. The extent of Dravidian (Brahui) in Afghanistan is grossly exaggerated, yet the family is absent in northern Sri Lanka. The mapping of Inuit-Yupik is laughable, showing it as limited to, yet entirely encompassing, Alaska. In actuality, this language family extends across the Arctic to Greenland, yet does not extend into central or southeastern Alaska, which are instead Na-Dene- and English-speaking. I could go on, by the exercise would quickly become tedious.

To be blunt, such slapdash cartography has no place in serious scholarship. I doubt that any authors who would approve of such mapping have an adequate knowledge of linguistic geography to carry out such a research program. We should be able to expect much better from both the authors and the journal.

In short, “Ultraconserved words point to deep language ancestry across Eurasia” is premised on the notion that cutting-edge research in historical linguistics requires little knowledge of linguistic geography, linguistic history, or even linguistics itself. It is hardly surprising that such a research program would yield inadequate results.

 

________________

*Technically, the grouping proposed by Pagel et al. (2013) is different from the original extent of the Eurasiatic macro-family, as proposed by Joseph Greenberg, in that it does not include Nivkh, but does include Kartvelian and Dravidian families. Nor is this grouping co-extensive with the Nostratic macro-family, as proposed by Vladislav Illyč-Svityč and Aaron Dolgopolsky: Pagel et al.’s grouping includes Chukchee-Kamchatkan but does not include Afroasiatic. It is also noteworthy that most linguist find the Altaic family to be deeply problematic.

Previous Post
«
Next Post
»

Subscribe For Updates

It would be a pleasure to have you back on GeoCurrents in the future. You can sign up for email updates or follow our RSS Feed, Facebook, or Twitter for notifications of each new post:
        

Commenting Guidelines: GeoCurrents is a forum for the respectful exchange of ideas, and loaded political commentary can detract from that. We ask that you as a reader keep this in mind when sharing your thoughts in the comments below.

  • http://www.facebook.com/barbara.h.partee Barbara H Partee

    Sally Thomason has also just posted a highly critical post on Language Log – http://languagelog.ldc.upenn.edu/nll/?p=4612 . I am rather upset at PNAS. For self-submitted articles (as opposed to articles ‘sponsored’ by an NAS member), the editor-in-chief tries to find the most competent NAS member to serve as editor for that article. Once the member has accepted that role, his/her job is to identify 2 referes (using or ignoring suggestions made by the authors), and then when their comments are received, making the editorial decision.

    In this case I see that the NAS member (a “Foreign Associate”, not a regular member) was Colin Renfrew; I understand from Sally that he is on record in favor of this line of research.

    As long as that’s the system, it’s hard to imagine what to do. I will try to bring to the attention of the editor-in-chief the uproar this article has caused in the linguistic community. It’s embarrassing for PNAS and by extension for NAS.
    Anyway, nice article, thanks!!
    Barbara

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Hi Barbara, Thanks for sharing all this “behind the scenes” information. Indeed, it appears to be a total failure of the peer review article. While Colin Renfrew in his own work treads delicately when it comes to linguistics, he is no linguist. He is definitely in favor of this research as it supports some of his own analyses. Who the referees might have been, I don’t even want to think about. It is indeed embarassing for PNAS/NAS, but the same applies to Science, Nature and other “estimed” publications that have published similar stuff.
      And yes, I have read Sally Thomason’s critical post on LanguageLog. It’s good though somewhat too focused on the specifics ignoring the bigger issues.

    • alysdexia

      It’s not nescient; you are.

  • marie-lucie

    Great review, Asya! You explain the problems clearly without talking down to the reader, and you do give actual language examples. You correctly address the often overlooked point that “language” does not simply equal “words”, but that grammatical structure and elements (the most slowly evolving part of language) are more crucial to recognizing linguistic relationship than individual words. An obvious example is English vs. French: lots of words in common, but the grammatical skeleton is very different (verb structure, for instance). The basic vocabulary is quite different too: quoting just words from among the “23″: hand/main, man/homme, give/donner, spit/cracher, bark/écorce, that/ce, ashes/cendres, are hardly closer to each other than the words of either language are close to the Russian ones. Thank you!

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thanks, Marie-Lucie! I appreciate the praise coming from you especially. Good point about the English/French contrast. And yet, French is grammatically closer to English than Italian is to English, no? Requiring overt subjects for It rains/Il pleut., for example, where Italian Piove is just fine.

      • marie-lucie

        Thank you! By “grammatically” you seem to mean “syntactically”, and syntax is not very useful for showing genetic relationship. Some aspects of French syntax show Germanic influence on Old French, and this may be one of them. But French verb structure is definitely Romance, not Germanic, while English verb structure is the opposite.

        • http://www.pereltsvaig.com Asya Pereltsvaig

          By “grammatically” I mean morphosyntax rather than actual morphological pieces (i.e. bound morphemes). Indeed, requiring overt subjects is an aspect of French that has been influenced by Germanic (as is inversion in questions like Est-il… ? or the more frequent use of Adj-Noun order). See also:
          http://geocurrents.info/geonotes/new-city-with-mini-quiz

    • http://www.facebook.com/george.starostin George Starostin

      “Grammatical structure and elements (the most slowly evolving part of language) are more crucial to recognizing linguistic relationship than individual words”.

      Hi Marie-Lucie,

      I would think that this statement needs a correction: it is far from true that grammatical elements always evolve slower than the basic lexicon – in fact, there are quite a few examples to the contrary, starting from your own English / French example. Despite the heavy discrepancies in basic lexicon, English and French still share around 30 genuine lexical cognates on the Swadesh list – but how many Indo-European grammatical morphemes have they preserved in between? And this is far from a unique situation: there are multiple instances when grammar is powerless to determine the issue (see David’s note on East Asian languages below).

      Relying on basic lexicon is perfectly legitimate to whatever time depth we may take the comparative method in general – where basic lexicon fails, morphology also fails by definition (but not necessarily vice versa). The problem with Pagel’s paper is not that they are relying on words rather than something else, but that the work applies rigid methodology to very raw and preliminary data, above all else, without a proper understanding of the semantic aspects of linguistic reconstruction.

      • http://www.pereltsvaig.com Asya Pereltsvaig

        Thanks for sharing your thoughts, George! I think we simply don’t know how quickly “grammatical structures” change simply because they don’t have a good sense of what exactly “grammatical structures” refers or should refer. As Marie-Lucie has pointed out, there is a difference between bound morphology and syntax, and as David pointed out below, Meillet’s and Longobardi’s understanding of “grammatical structures” are widely different too.

      • marie-lucie

        Geroge, I do not mean that reliance on words is the only thing objectionable in the Patel paper, and I entirely agree with David’s note below about morphology. The fact that some languages have lost a lot of their (especially inflectional) morphology does not mean that morphology should not be of great importance when it does exist. But morphology does not just mean grammatical morphemes, but general word structure. For instance, ablaut is part of IE verb morphology, vowel “interweaving” and other processes affecting a “skeleton” consonantal root are part of Semitic morphology, and such processes might remain or leave traces whether or not they are associated with individual grammatical morphemes.

      • German Dziebel

        “without a proper understanding of the semantic aspects of linguistic reconstruction.”

        It seems to be a major weakness of traditional as well as long-range comparativist research, so this is not what sets Pagel apart from “good linguistic science.”

        • http://www.pereltsvaig.com Asya Pereltsvaig

          Why do you say so?

          • German Dziebel

            Sound laws depend on what you put into a cognate set. Traditional comparativist method depends on the assumption of the stability of word semantics. But if semantics changes (and the form changes, too), then the comparativist procedure will not be able to detect sound laws. My favorite example: on the basis of Lat frater, Slav bratru, Skrt bhraater, etc. we reconstruct PIE *bhraater (which is essentially the Old Indic form) or, with a laryngeal, *bhreH2ter ‘brother’. But if we look around and adduce Lat mariitus ‘husband’, Germ. bruudi ‘bride’, Lith merga ‘girl’, Gk meirax ‘boy, girl’, etc. supported by the perfect semantic and morphological match between Latv marsa ‘brother’s wife’ (from group 2) and Lat fratria ‘same’, we will arrive at PIE *mreH2ter, with the secondary origin of aspiration from laryngeal throwback. It’s very different protoform with very different semantics (a sort of affine, which is consistent with typological evidence for the historical priority of affinal over consanguineal semantics). And this is not the only example. Swadesh list is an interesting, speedy heuristic but it needs to be treated as only one input into the overall long-range reconstruction complemented by a more systematic exploration of whole lexico-semantic classes (kin terms, numerals, etc.) from which the words on Swadesh list are pulled.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            But sound laws of the classical comparative method are not deduced based on Swadesh material only!

            (Nor am I convinced that both Latin frater and mariitus are derived from the same proto-form—what evidence do you have for that?)

          • German Dziebel

            Asya, what in the world do sound laws have to do with Swadesh list?

            mr > br is a typical IE phonetic process attested in such branches as Indo-Aryan (Skrt. braviti ‘talk’ < ml-, Russ molvit'), Greek, Latin (Lat brevis, Gk braXus < *mreghos), Germanic (bruudi above), Celtic. The exact morphosemantic parallel between Latv marsa and Lat fratria seals the deal. What's interesting about *bhreH2ter < *mreH2ter is that this initial cluster transition preceded the aspiration throwback (for the latter see Indic evidence as in duhita *mromiHx- (Gk murmees) > *bromiHx- > *bhromi- > *fromii- > *formii-ca.

            And again this is just one example of how semantic and formal variation need to be analyzed in tandem.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Indeed. But that’s what classical comparative method does do, no?

            As for the Swadesh list, you said that it is insufficient and sort of implied that it is what classical comparative method uses. Or what was your point in the last sentences of the previous comment?

          • German Dziebel

            Not really. I detected the *mreH2ter situation by applying a novel method of comparing existing IE cognate sets with each other relaxing some semantic and phonetic constraints. I wrote extensively about these experiments in “Algebra rodstva” in the early 2000s. Classical comparative method telescopes the complexity of formal and semantic differentiation into simple “head-head” or “leg-leg” cognate sets. The fact that *bhreH2ter from *mreH2ter hasn’t been identified over the past 200 years, while the usually reconstructed form *bhraater has received a myriad of fanciful etymological interpretations (“fire carrier,” etc.) is a good illustration of the limitations of classical comparative method.

            Yes, the Swadesh list is insufficient. I just never said that sound laws are derived from the Swadesh list. IE sound laws well predate the Swadesh list. The Swadesh list is just a good illustration how comparativists view semantics – as a rigid and stable set of universal concepts. Sound laws derived from such cognate sets are likely to be partial and inaccurate. Each concept, I argue, should instead be seen as a field of variation that has to be captured and described in order for correct sound laws to be derived.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            You contradict yourself: you say (correctly) that Swadesh list postdates the comparative method, yet use it to criticize the comparative method. The problem is that even though semantic change happens and may result in cognate sets with divergent meanings, relaxing semantic (let alone phonetic) constraints needs to be restricted so as to make the method meaningful. Classic comparative method may be too conservative in this, according to your taste, but it’s better to err on the side of caution, no?

          • German Dziebel

            I don’t contradict myself. The Swadesh list exemplifies the same tendency to simplify semantics as can be found in any other application of traditional comparativist method. Semantic change doesn’t simply happen – it’s part and parcel of language evolution. Caution is always a good advice, hence I’m advising to treat “classical comparativist method” with caution. There’s a constant circular argument going on there with cognate set definition.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            No argument about semantic change being part of language evolution. But will you be happy to include any meanings in one cognate set? Or where will you draw the line?

          • German Dziebel

            Historical forms need to have a convincing etymology and the evolution of entire lexicosemantic classes (kin terms, numerals, somatonyms, zoonyms, pronouns, etc.) should be clarified. This is where the line should be drawn, I think. Widely-spread IE forms (e.g., *bhreH2ter or kwetwoor ‘four’) don’t have etymologies, although they are visibly morphologically complex. The evolution of IE kinship system or numeral system is still obscure. And the respective cognate sets are composed of exclusively forms with the same meaning. The semantic variation in these classical sets is close to zero. This is a sign that “true” cognate sets that contain all of the formal and semantic variation necessary to etymologize the derived forms and understand the evolution of a lexicosemantic class have been separated into “artificial” ones.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            “The evolution of IE kinship system or numeral system is still obscure.” — this is exactly why I and most historical linguists prefer to err on the side of caution and include only items with nearly identical meanings. But I wouldn’t say that even kinship terms that are commonly considered have exactly the same (range of) meanings in various IE languages, even if compare present-day ones. E.g. ‘father’ in German has (shades of) meaning that is not found in English or Russian or French.

          • German Dziebel

            “this is exactly why I and most historical linguists prefer to err on the side of caution and include only items with nearly identical meanings.”

            But this precisely why the evolution of those systems is obscure. It’s like trying to build a theory of phonological evolution and use only those cases when /p/ corresponds to /p/ and /a/ to /a/. This will mean no evolution and no science of language.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            “this precisely why the evolution of those systems is obscure” — not really. The two issues are only tangentially related as sound changes should be/are observable across semantic classes.

          • German Dziebel

            They absolutely should but unless one properly assigns cognates to a true cognate set, the full scope of sound changes will not be recoverable, the individual items will not have an etymology and the evolution of the whole class will not be elucidated.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Not really. Sound changes should be apparent even if one considers only a subset of what you call “true cognate sets”. To put it the other way around, if a certain sound change emerges only if you allow items with very different meanings to constitute a cognate set, this is a very suspicious sound change, wouldn’t you say?

          • German Dziebel

            This is the core of the issue. Classical comparativist method tacitly assumes that semantics is irrelevant to successful sound law recovery. You just captured this weakness very well: take the simplest, most obvious subset and language-wide regularities will follow.

            The IE *mreH2ter, however, suggests otherwise. The semantics of the true, complex cognate sets I’m talking about needs to be meaningful of course, just like the sound laws elicited on the basis of those true, complex cognate sets need to be phonetically plausible and systematic. But the methodology of cognate set composition needs to be deductive as well as inductive to be able to derive true phonetic laws. Without this level of analysis there’s no proto-language reconstruction in the first place because classical compartivist method results in privileging one attested form over other. In the case of *bhraater, the reconstruction replicates the Sanskrit form. The laryngeal adaptation *bhreH2ter evolved it a bit, but the stumbling block is not H2 vs. aa here, but the voiced aspirate in the onset which is a Sanskrit-specific feature. The form *mreH2ter makes the reconstruction truly PIE by moving away from language-specific phonetic features and by bringing together two IE-wide cognate sets into one.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Let me see that I understand you correctly: you seem to be saying that in addition to sound changes that are apparent from words that haven’t changed their meanings (or haven’t changed them much, so that they still pass the cognates definition of the classical comparative method) there are also sound changes that only occur in words that radically changed their meaning. This is an interesting hypothesis. But what does it suggest about the nature of language change? It appears to me that for this hypothesis to be true, certain sound changes are tied in with meaning changes, are not independent. I don’t know of any example of known sound change that is tied in with meaning change. Do you? I agree that certain sound changes may be so rare that we don’t “catch” them if we have a very tight constraint on cognates, but apart from frequency I don’t see how such a hypothesis can be sustained.

          • German Dziebel

            Again the *mreH2ter > *bhraater example illustrates well what I’m trying to say: yes, all IE languages outside of Skrt seem to have undergone shifts away from bh- into b- and f-. This likely happened independently in different branches. But prior to that 1) mr shifted to br in all dialects in mr- forms of the mer-/mre- paradigm; 2) the medial laryngeal aspirated the initial stop. All these earlier processes happened against a different semantic background: there was no separate term for ‘brother’ or ‘clan brother’ in those days, but there was the notion of an affine engaged in recurrent marital exchange between “clans.” Since both forms *mer- and *bhraater are found in IE languages, this is not a “Nostratic” situation but an ancient IE situation. Classical comparativist method stops short of detecting this cognate set and would proceed by searching for correspondences to both *mer- and *bhraater outside of Indo-European, while in reality the brother group is of relatively late origin semantically, morphologically and phonetically.

            In a word, there’s no objective “meaning changes sound changes” dependency (pace Saussure) in the natural life of languages, but there should be methodological dependency between semantic and formal variation in linguistic reconstructions.

          • Jaska

            But those words for ‘ant’ and ‘eyebrow’ were found already by traditional historical linguistics, if I got it right? Even though they were seen as irregular. So, what does your extension of the method gives, that wasn’t there already?

            For me it is not a restriction of method, but some linguists use more strict semantic criteria than others.

          • German Dziebel

            The cognate set ‘ant’ is a well-known one. It’s the phonetic variation within it that has remained unexplained. By applying a new method across sets I found a solution. The cognate set ‘edge; eyebrow’ was not detected by classical comparativist method. My extension of it identifies unobvious cognation, which helps formulate new sound laws that explain what otherwise remains unexplained.

            I don’t think it’s a matter of using more or less restrictive semantic criteria. It’s a matter of developing a method that’s more geared toward situations when forms have undergone both formal and semantic changes. Otherwise, the method piggybacks on an “ideal situation” when forms objectively haven’t had time to diverge much from each other and hence is not much of a method after all. (I think Claire Bowern said something to that effect.) We also need to be able to provide strong etymologies, which classical historical linguistics often fails to do. Instead, it produces semantic monstrosities such daughter as “milker” (and there are many of those and they keep coming in mainstream Indoeuropeanist publications) because it’s not capable of including semantic change into the study of formal variation.

          • Jaska

            OK, German, this new sound law sounds quite good. Still, I wish to see it written open in a paper.

            If I get it right, you blame the mainstream Indo-Europeanistics both for too strict semantic criteria (not detecting cognates with different meanings) and too wide semantic criteria (monstrosities). How is that possible?

          • German Dziebel

            Thanks Jaska. For this paper to be written, I need a co-author with a very detailed knowledge of IE linguistics and specifically phonology. I have quite a few solid hypotheses (some are quite surprising if proven true) but I need help putting them on a solid phonetic footing.

            With your question, you nail the contradiction on the head. Classical historical linguists don’t allow semantics to influence cognate set composition, but then they are still compelled to provide etymologies, which are semantic interpretations, for the reconstructed forms. So they are forced to go outside of the immediate meaning (e.g., ‘daughter’ for IE *dhugH2ter) and link it to an etymological meaning (‘milker’ because there’s a homophonous root *dheugh-). The link is more than tenuous, but scholars leave no other option for themselves because they had already misconstrued the original cognate set. Instead, I propose “baking” plausible semantic evolution into the cognate set composition (maintaining phonetic plausibility, of course), so the search for an etymological meaning becomes redundant, and we get both phonetic and semantic evolution in tandem.

            In a word, a true comparativist method involves the reversal of the place of semantic chnage in historical reconstructions: from an etymological leap of imagination to a productive logic of cognate set composition.

          • alysdexia

            daughter/filia < dough; dig/finge; doe “suckle”

          • Jaska

            German:
            “The IE *mreH2ter, however, suggests otherwise. The semantics of the true, complex cognate sets I’m talking about needs to be meaningful of course, just like the sound laws elicited on the basis of those true, complex cognate sets need to be phonetically plausible and systematic.”

            Very interesting. But if a sound correspondence is regular, it should be seen also with some words with less varying meanings. If it is seen in this one word only, maybe it is not a question about true sound correspondence at all?

          • German Dziebel

            I’m just using *mreH2ter as an example because it’s seems to be rather lucid. Of course there need to be several forms attesting to the sound shifts in question. My other examples illustrating the same phonetic development are: Lat formica ‘ant’ next to Gk murmees ‘ant’, etc. Classical comparativist method has no explanation for Lat f- here, although the cognate set has been identified correctly. Or, among potential oversights, is the connection between Lat friigus and Slav *merz- ‘freeze’ exemplifying the same aspiration throwback happening on the heels of a mr>br transition. (The challenge here is to explain Gk rhigo-, which is usually paired with Lat friigus, but it’s an odd pairing distributionally and phonetically anyhow.) IE *bhruu- ‘eyebrow’ (Skrt bruu-h, Slav *bry ‘eyebrow’, Lith briauna ‘edge’, Russ brevno ‘log’) looks promising next to IE *merg’- ‘edge, limit, mark’ (Lat margoo, Goth marka) and suggests PIE *mreH2- ‘edge’ and *mreH2u- ‘eyebrows’ (dual?) with the historic eyebrow words being semantically and phonetically derived from the more basic semantics and phonetics of ‘edge’.

            So this does seem to be a potentially regular process.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            What is your explanation, then, for the fact (which you claim to be
            true) that some sound changes are only apparent when semantic
            definitions are relaxed?

          • German Dziebel

            I think it’s the product of the compounding effect of semantic, morphological and phonological changes. Semantic change is constant and normal, and the assumption of semantic ultraconservatism is flawed. The radical morphological alternation mer-/mr- created two pathways for phonetic development. Changes in society made the ‘brother’ role prominent and fixed across early IE speech communities, while the earlier social role of an ‘affine’ associated with the older mer- form of the root disintegrated into a myriad of forms with diverse meanings ranging from ‘husband’ (Lat mariitus), to ‘bride’ (Germ. bruudi, Lith marti) to ‘boy; girl’ (Gk meirax, Lith merga), Avest mairya ‘yeoman’.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            This is all fascinating, and maybe even true, but you didn’t answer my question: why certain sound changes are dependent on semantic changes (beyond what classical comparative method would be happy with)? What do we learn about the nature of human language from this fact? (if it’s true, which I don’t necessarily believe).

          • German Dziebel

            I was trying to answer your question, Asya, and I can’t add much more to what I’ve already said. I’ll keep your question in mind, as I think it’s a good one, but for me it’s always been a matter of improving the methodology, rather than seeking objective reasons for the discrepancy. But you’ve made me think: what if there’s some form of “linguistic relativity” or a Bakhtinian chronotope at play here?

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Or perhaps your hypothesis that by relaxing semantic defitions we will be able to uncover some sound changes that are otherwise *in principle* uncoverable is simply wrong?

          • German Dziebel

            Asya, you need to look at the evidence and not try to resolve the problem in the abstract.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            So considering the methodological implications of what you propose is prohibited?

          • German Dziebel

            Not at all. Why? This is precisely what needs to be done. But I don’t think you can figure out if my extension of comparativist method is “wrong” without focusing on the actual etymological material. If it proves to yield systematic results and solve problems that otherwise remain unresolved, then it’s probably not “wrong.”

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Here’s where we disagree: I think that meta-theoretical considerations are as important as data-driven ones in evaluating a given theory/method.

          • German Dziebel

            I don’t disagree with that. I just can’t contribute much to this specific issue as I haven’t thought about it, but I don’t believe it should be positioned as a wrong or right situation on the metatheoretical level. There may be a statistical test in order once the number of “complex” cognate sets becomes large enough.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            We’ll have to agree to disagree here. Let’s wrap up this discussion.

          • German Dziebel

            Thanks for the interesting conversation, Asya.

  • http://www.facebook.com/amjaker Alessandro Michelangelo Jaker

    I agree that there are many problems with this latest Atkinson et al. paper, but I am also concerned about many of the comments I have been reading, i.e. “how did this get published?” and “this never should have been published,” which sound like calls for academic censorship. How can we distinguish genuinely sloppy work from work which, perhaps, merely makes non-standard assumptions? New hypotheses, when they first arise, are almost always incomplete and deficient in some way–it is only by repeated trial and error that they are refined into something more convincing. Similarly, while I agree there is a great deal of garbage-in, garbage-out going on here, I am not convinced that their model is entirely unfixable. New hypotheses are almost always sloppy when they are first proposed.

    Personally, I would rather err on the side of allowing a crackpot theory here and there to slip through, rather than risk shutting down a genuinely interesting but incomplete proposal. Is there really any a priori way to tell the two apart?

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thank you for your comment, Alessandro.

      While in principle I agree with you that there shouldn’t be external censorship in science, I think here we are on solid ground. You say: “New hypotheses, when they first arise, are almost always incomplete and
      deficient in some way–it is only by repeated trial and error that they
      are refined into something more convincing.”—well, in this case this is not really a new hypothesis at all, as the authors have been at it, doing similar things and making the same major blunders for a decade, and their repeated trial and error became indeed convincing… that they are not on the right path at all, as I said in my part of the post.

      Moreover, I am all for original, yet sensible, assumptions/hypotheses, but to make such assumptions/hypotheses, one needs some solid basis. To deny the basics of at least three disciplines—linguistics, history, and geography—is not the same as to make an original breakthrough. Progress in science cannot be made by always starting from scratch, which is what one is left with if one denies the accumulated experience of previous generations of scientists in a discipline. Or three, as is the case here.

    • http://geocurrents.info Martin W. Lewis

      I agree that censorship must be opposed, and I have no problem with the fact that this article was published, Indeed, I am rather perversely pleased to have material like this to criticize.

      But I also think that we must recognize the fact that scholarly journals “censor” themselves to a degree through the peer-review process, which is one reason why I am glad that the internet allows an open-entry, anarchic alternative to scholarly publication. It is not that I want to do away with academic journals, but I do highly value the blogosphere’s alternative.

      Wild theories can indeed have value, but it is significant that some kinds of crackpot ideas — those dressed up with complex statistics — often easily pass though peer-review in journals such as PNAS, whereas others do not.

      • http://www.pereltsvaig.com Asya Pereltsvaig

        I agree with Martin that peer-review serves is censorship of sorts and that it is necessary if scientific progress is to be made. That’s why respectable scholarly journals have the process in the first place. Open-entry, free-for-all blogosphere has an important role to play too. The key, however, is the wisdom of knowing what sorts of things should be published where. “Render unto Ceasar the things that are Ceasar’s” etc. In blogosphere, a piece such as Pagel et al. would raise few eyebrows, but in a supposedly peer-reviewed journal? As one of our readers commented on Facebook, “peer-review” done by known advocates of the approach taken in the paper to be reviewed is like a jury consisting of people suspected of the same crime as the accused. Here we seem to have an instance of exactly that.

        Also, it seems to me that much of the outcry from (historical) linguists is not about how “crackpot” the ideas of the paper are, but how lacking in basic knowledge of linguistics (or history, or geography) its authors are. When Marie Curie discovered polonium, it’s not that she didn’t know about the periodic table…

    • Gman79

      I disagree with academic censorship too … but I’m not sure this is either a “crackpot theory” or a “genuinely interesting but incomplete” proposal. It seems too fundamentally flawed to be either of those.

      • http://www.pereltsvaig.com Asya Pereltsvaig

        Well-said!

  • http://www.facebook.com/david.erschler David Erschler

    Meillet apparently meant something drastically different from what Longobardi and his collaborators are trying to do. Namely, what he argued for was the importance of morphology (meaning functional morphemes and various idiosyncratic alternations, like the ones occurring in IE verbs) in establishing genetic relationship. This indeed remains one of the main techniques in the classical comparative methods, although of limited applicability when the morphology is impoverished, like, e.g., in many languages of SE Asia.

    Longobardi, on the other hand, works with highly abstract (and theory-internal) syntactic parameters, a rather controversial technique, which needs to draw on theoretical analyses of languages under study. Needless to say, these analyses are simply non-existent for the vast majority of the world’s languages, nor is there any hope to obtain such in any foreseeable future.

    The situation with dictionaries and, to some extent, with morphological descriptions, is incomparably better. This is what largely motivates “word-based” approaches to language classification.

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thank you for your comment, David. Indeed, Meillet could not have referred to the same abstract parameters that Longobardi refers to simply because they are a much more recent notion. I guess, my point was that Meillet understood the unreliability of individual lexical items as evidence of historical relationships a long time ago, but now the research can focus on the real underlying aspects of language. Whether the parameters chosen by Longobardi are valid is another question, not unlike the question of which items (and how many) should be included in the “Swadesh list”.

      Moreover, I agree with you that “word-based” approaches are easier to pursue at the moment simply because the progress in theoretical linguistics that I mentioned in the post is not completed yet, as it were. If we had an agreed upon understanding of the parametric theory, it would be a no-brainer, but as long as we don’t have it yet, does it still make sense to focus so much effort on “word-based” studies? I don’t think so.

  • http://www.eveningoflight.nl/subspecie Oscar Strik

    I was a bit disappointed by the article as well. I think there are a lot of interesting prospects in the statistical analysis of the relationship between frequency and conservation of words, and I am open to the idea that some words may have extremely long half-lives, in the sense of not being lexically replaced. But the argument put forward in Pagel et al. is far from convincing if you take all factors into account. In the end, it boils down to a conclusion that is plausible but by no means proven.

    In comparison, I was much more convinced by the 2007 paper by Lieberman et al. (http://www.nature.com/nature/journal/v449/n7163/abs/nature06137.html), which appeared alongside the earlier article by Pagel et al., precisely because they confine their analysis to a shorter period and using attested forms.

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thanks for sharing your thoughts on the matter, Oscar, and for the link to the related paper.

  • http://blog.zolnai.ca/ Andrew Zolnai

    Great reviews from both of you as usual, you seem ot be able to pull no punches w a smile, and that IS a talent. As one of Hungarian origin I cannot understand Finnish despite Finno-Ugric roots

    [My ancestor Bela Zolnai worked on that, um sure you have an opinion on that too. I think he wrote that words relating to steppes have common roots when both peeps were together, but then the roots diverge on terms reflecting their separation, Finns to the Baltic Sea and Magyars to the Pannonian plains, so I think fishing and agricultural roots differ. Again I didn't read his works myself.]

    I have a question on ‘man’. I heard that chairman etc. refers not to ‘man’ the gender differentiator but to ‘manus’ the hand that does it as in workman. I use it to poke holes in political correctedness and runaway gender issue topics LOL

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thank you, Andrew! Great point on Finnish-Hungarian dissimilarities.

      As for man, my etymological dictionary says it comes from “chair” + “man” (not “manus”). Surprisingly, “chairwoman” was attested in 1752, while “chairperson” only goes back to 1971.

    • marie-lucie

      Chairman, fisherman, policeman, workman, statesman, sportsman, and many others, include man ‘adult male human’ (formerly simply ‘adult human’). All those words designate men characterized by their role or function. The word manus ‘hand’ is a Latin word which does not have a cognate in English. If the man in workman was a cognate of manus, the complete word would mean a kind of hand, not a kind of man. Words such as manufacture and manumission are borrowed and slightly adapted from Latin.

      • http://www.pereltsvaig.com Asya Pereltsvaig

        I don’t agree that policeman and the like refer exclusively to male people in these occupations. Can’t you way “Some policemen came to the house” if there were men and women? Of course, any fan of British mysteries, such as myself, would know that there are also WPCs — that’s Woman Police Constable. But there is no MPC (Man Police Constable) to go with it.

        As for the word manus meaning hands, yes, but don’t they say “All hands on deck!”, meaning actually whole sailors?

        • marie-lucie

          Policeman, etc started as exclusively male occupations, and there were also exclusively female occupations such as midwife or washerwoman. When the male occupations were open to women, many women were uncomfortable with the suggestion that the default was male, so in North America there were two solutions: either use woman instead of man ( hence policewoman, chairwoman, spokeswoman, and the like) or use a generic, unisex word (police officer, chairperson, flight attendant, etc). If several police officers came to the door, I think that most people would announce The police is at the door, perhaps adding one man and two women if appropriate. As for hand in deckhand and farmhand, the word is indeed used for people, usually men, but not for every man on a ship or farm: only for those actually doing the physical work, not for officers or supervisors. And even if hand can refer to a man in these contexts (an instance of metonymy), man cannot be used to refer to an actual hand.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Great points, thanks, Marie-Lucie!

          • Peter

            If you are talking about a group of female police officers (composed of, say, four women) you CANNOT use the plural noun “policemen”. This shows that the word applies only to a plurality that includes males.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            I think you can use “policemen” there if gender is unknown or not important. Same as with “lions” for a groups of female ones, etc.

          • Peter

            I can’t imagine a context where “policemen” can be used to refer to a group of female officers because their gender is ‘not important’. And the fact that one cannot _refer_ to a group of female officers shows that the noun carries gender conditions. It seems to me that if a speaker can’t tell the gender of the officers (because they are too far or something) then the default assumption is that they are probably males, because of world knowledge, and the term “policemen” can be used.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            This is exactly the sort of situation I had in mind. Whether it’s world knowledge or language (or language informed by world knowledge), I would be hard pressed to tell.

        • Gman79

          Oops … Marie-Luce just made the point I was going to make, below.

        • alysdexia

          wapmen and women
          wermen and wifmen

  • TenDeuChen

    What do you mean “a direct translation from Russian”? If you translate something from a language, you then put it into another language. All you have there is Russian written with Latin letters. That is called a “transliteraton”.

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Sorry, a typo: it should say “into Russian”. I’ll have that fixed now.

  • PeasePorridge

    Saw the article in the WaPo and knew that Geocurrents would be on it. Thanks for the post, Asya and Martin.

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thanks, PeasePorridge!

  • http://www.facebook.com/richard.compton Richard Compton

    In the first paragraph of Part 1 I believe “Upik” should be “Yupik”.

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Ah yes, another typo. Will fix it. Thanks!

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thanks, I’ve fixed the typo.

  • A.F

    Anyway there’s a clear methodological flaw.

    How can they claim that a word is stable when we do not have a reliable reconstruction of the shared ancestor of these language?

    The whole thing is nonsense.

  • http://www.facebook.com/fejes.laszlo Fejes László

    One more major mistake on the map: the Uralic languages of the Volga region (Erzya, Moksha, Mari, Udmurt) are missing…

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thank you for pointing this out, Fejes! We did discuss this error while working on the post, but somehow it didn’t get mentioned in the final version of the text… I am glad you brought it up!

  • Pingback: Stuff « Econstudentlog

  • http://www.facebook.com/rebecca.tun Rebecca Imogen Tun

    So was Borat tapping into medieval language change?
    My suit is grey………..naht. My suit is grey not.

  • Jaska
    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thank you, Jaska! I was wondering what your reaction would be… Great stuff!

  • Vladimir Mayakowski

    I am not a linguist and will not discuss linguistic things.

    I would like to speak about a way of critique when one thing what is criticized is done by the criticizer himself. In Part2 you wrote the following – “Why is Kashmir mapped as a separate country; is this a deliberate
    political statement, designed to infuriate India and Pakistan in equal
    measure?”. At the same time you wrote “Turkic Azeri, with its 23 million speakers spread across Azerbaijan and northwestern Iran, is classified as an Indo-European language”. And by this you actually infuriate Iran as there are some claims of Azerbaijanian politicians to join this northwestern Iranian provinces to their country based solely on the fact that people on that Iranian region speak Azeri language. In your text I see the following – if India or Pakistan then raging is prohibited, if Iran then please go ahead.

    One more thing. In Grugni et al 2012 paper (“Ancient Migratory Events in the Middle East… New Clues from the Y-Chromosome Variation of Modern Iranians”) the authors describe very well which linguistic shifts have been taken place in northwestern Iran. Namely, northwestern Iranians were speaking Azari language, what is a member of Indo-Iranian subfamily of Indo-European languages. This language survived at least up to 17th century, i.e. up to very recent times for the region. And taking into account that Pagel et al have written in the paper that “The color-shaded areas should be treated as suggestive only”, the yellow color in northwestern Iran is acceptable.

    I am not completely in favor of this study and others from the same authors. But I think the direction of the research, namely Bayesian glottochronology in historical linguistics, is correct. Of course, there need to be a critical selection of research materials, refinements and polishing of similar word sets, purification from borrowings, etc.

    • http://geocurrents.info Martin W. Lewis

      Thank you for taking the time to respond, but your criticism is not valid. The map in question shows two things: sovereign states (uncolored polygons enclosed by grey lines) and language families (shown by color). Kashmir is depicted as a sovereign state, whereas in actuality it is a disputed territory, the control of which is divided by India and Pakistan. The geopolitical mapping here is simply incorrect.

      It is an undeniable fact, however, that Azeri is spoken by as a first language by most people, numbering some 10-15 million, in northwestern Iran. Mentioning this fact has no bearing on issues of Azeri separatism in Iran, a minor movement covered in earlier GeoCurrents posts (see the post on the “Cartoon Cockroach Controversy). Note also that I called the language “Azeri” rather than “Azerbaijani,” even though “Azerbaijani” is the term used by the Ethnologue, the Wikipedia, and most other sources. I prefer “Azeri” precisely to avoid any possible insinuation of Azerbaijani irredentism.

      As far as the historical linguistic situation is concerned, your point is interesting but not relevant. The map show current-day distributional patterns, not those of earlier centuries. Any “suggestion” that northwest Iran belongs in the I-E zone today is simply misleading.

      • http://www.pereltsvaig.com Asya Pereltsvaig

        Here’s the post on the “Cartoon cockroach controversy”: http://geocurrents.info/geopolitics/iranian-azerbaijan-and-the-cartoon-cockroach-controversy

      • Vladimir Mayakowski

        It is worth to note that I agree with you that Kashmir must not be drawn within a separate polygon. That is clearly incorrect. But I spoke about “infuriation”. If one should avoid it in scientific literature than he must do it for all of the similar cases. And it does not matter what is the reason for “infuriation”, polygon or color or somewhat else.

        According to current-day distributional pattern west, southwest and southeast Azerbaijan must be under yellow color. Because in the west and southwest of what is currently recognized borders of Azerbaijan there is the de facto independent republic of Nagorno Karabakh what is inhabited with armenians, and they speak Armenian, what is Indo-European language. Southeastern part (Lenkoran, with its surroundings up to the Iranian border) is inhabited with talyshes, and they speak Indo-Iranian, what is also an Indo-European language. Moreover, eastern part of northwestern Iran is also inhabited with talyshes, and they also speak talysh (one part of the same people live in Azerbaijan, the next part in Iran). So this is the situation and it can’t be a “misleading suggestion”.

        • http://www.pereltsvaig.com Asya Pereltsvaig

          Good points about Nagorno-Karabakh and Talysh areas. However, there’s still a pretty big chunk of Azerbaijan that is Azeri (Turkic) speaking, no IE. Check out our linguistic map of the Caucasus:
          http://geocurrents.info/place/russia-ukraine-and-caucasus/caucasus-series/languages-of-the-caucasus-4-0

          • Vladimir Mayakowski

            Nagorno-Karabakh is not an area. It is already 20 years has its own government. It is a de facto independent republic.

            Official census data from Azerbaijan is very far from true values.
            http://en.wikipedia.org/wiki/Talysh_people

            They now do exactly the same as Turkey government did during the beginning of the 20th century in Turkey – Armenian, Assyrian and Greek genocides. In Turkey now all people are turks. They don’t accept even kurds as a separate ethnic group in Turkey. In reality there are many different ethnic groups there. But they forced to see themselves as turks.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Re: Nagorno-Karabakh, it’s a (linguistic) bracketing problem: I meant [NK] + [Talysh areas], not [NK + Talysh] areas.

    • http://www.pereltsvaig.com Asya Pereltsvaig

      “I am not a linguist and will not discuss linguistic things.”, “I think the direction of the research, namely Bayesian glottochronology in historical linguistics, is correct.” — isn’t it a contradiction here? More importantly, Gray and Atkinson are always careful to point out that what they are doing is NOT glottochronology. They do not assume a constant rate of change. This assumption has been proven to be wrong, so nobody’s really buying glottochronology as such. But there’s enough issues with their approach so we don’t need to criticise (or support) them for something they don’t do.

      • Vladimir Mayakowski

        Ok, I can substitute “Bayesian glottochronology” with “MCMC”.

        • http://www.pereltsvaig.com Asya Pereltsvaig

          That’s not the same thing at all. MCMC may be a valid algorithm for processing certain kinds of data but it doesn’t mean that it can/should be blindly applied to everything.

          • Vladimir Mayakowski

            No, no. Not blindly!

          • Jaska

            Vladimir, here you can see some of the many problems in the method which applies computational phylogenetics “blindly”, ignoring all the methods and results of historical linguistics:
            http://www.elisanet.fi/alkupera/Problems_of_phylogenetics.pdf

          • Vladimir Mayakowski

            Thank you for the link.

            I should say that the study of Bouckaert et al. (2012) can be wrong, maybe completely or maybe in some instances. I will not discuss it here. But your way of critique seems not very good. You refer too much to you own studies. For example many scholars agree that Samoyed is the first branch split off from proto-Uralic. You did your study, and made a suspicion to that view. But the issue is not solved yet. Nevertheless, you use this your consideration to make the study of Bouckaert et al. (2012) wrong. With this way of contradiction you are very similar to Gray school. They also believe themselves too much. You can be true about Uralic branching pattern, but you can be wrong as well. So you must not base so extensively on your own considerations.

            I would like to ask you one question. I have not seen that you discuss bootstrapping issue. In both studies (Bouckaert et al. (2012) and Gray et al. (2006)) Anatolian language or languages split off from other Indo-European languages first within all the bootstrap trees. Whether the shortcomings of that studies, which you discussed, are so much misleading that all of the bootstrap trees out of 100 or more have the same topology concerning the Anatolian languages – they always split off first?

          • Jaska

            Vladimir:
            “But your way of critique seems not very good. You refer too much to you own studies. For example many scholars agree that Samoyed is the first branch split off from proto-Uralic.”

            I refer to my own research, because during the recent year nobody else has studied the subject. All uralicists, who are aware of my research, agree with it – they just haven’t yet published anything where they would say it aloud. You cannot rely on the outdated views of those linguists who are not yet aware of the new results, can you? Nobody has disproved the new taxonomic status of Samoyed – some scholars just aren’t yet conscious about it.

            Vladimir:
            “You did your study, and made a suspicion to that view. But the issue is not solved yet. Nevertheless, you use this your consideration to make the study of Bouckaert et al. (2012) wrong. With this way of contradiction you are very similar to Gray school. They also believe themselves too much. You can be true about Uralic branching pattern, but you can be wrong as well. So you must not base so extensively on your own considerations.”

            What is the contradiction you see? I don’t see any. My critique shows that there are many problems and weaknesses in the phylogenetic method, which the Gray School has not even considered – that means I expand the viewpoint! I show them how many possible interpretations there are, even though they always see only one. Only if I was narrowing the viewpoint, you could blame that I am like them. Do you understand this?

            Vladimir:
            “I would like to ask you one question. I have not seen that you discuss bootstrapping issue. In both studies (Bouckaert et al. (2012) and Gray et al. (2006)) Anatolian language or languages split off from other Indo-European languages first within all the bootstrap trees. Whether the shortcomings of that studies, which you discussed, are so much misleading that all of the bootstrap trees out of 100 or more have the same topology concerning the Anatolian languages – they always split off first?”

            As I wrote, also the traditional historical linguists think that Anatolian was the first branch to split off. So the result itself may be correct. But my point is, that the result cannot be confirmed merely on the basis of the lexical level, because there are other possibilities like false divergence. The phonological and morphological level are needed to prove the status, and that is what Kloekhorst has done.

            Also the Gray School are very narrow-sighted interpreting their results: that Anatolian split off first still cannot prove that Anatolia was the Indo-European homeland.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Thank you for continuing the debate! One addition to what I already said above re: the Anatolian split — you are absolutely right that it doesn’t prove the Anatolian homeland. As J. P. Mallory recently pointed out, it actually disproves it.

          • Vladimir Mayakowski

            “I refer to my own research, because during the recent year nobody else
            has studied the subject. All uralicists, who are aware of my research,
            agree with it – they just haven’t yet published anything where they
            would say it aloud. You cannot rely on the outdated views of those
            linguists who are not yet aware of the new results, can you? Nobody has
            disproved the new taxonomic status of Samoyed – some scholars just
            aren’t yet conscious about it.”

            You must not still bring it as the well grounded example to disapprove another study.

            Contradiction is that you made the possibility of their conclusion as little as 3.125%. And this you did based on extensively on your own studies. I understand and agree with you that there are many weak points in that study, some of which I got here. In my first post I wrote that this strategy needs refinements. Moreover, I don’t want to “fight” instead of Gray and Atkinson. Let them to come and argue their study. And your critique can help so solve somehow many problems of their research. But in your critique you speak very sharply, not constructively, making their study just nothing.

            Not only the fact that Anatolian split off first, approves that Indo-European origin is in Anatolia. The splitting sequence also support to it – Anatolian languages first, and then Armenian-Tocharian branch in Bouckaert et al. (2012), and Hittite first and then Tocharian, and then Armenian-Greek branch in Gray et al. (2006). Here I see a trace of movement. In phylogeny the tree has a nested structure. And this is a strong argument, which favors the Anatolian origin. I don’t know about weak points of phylogeny in linguistics. And I don’t know what J. P. Mallory recently pointed out. Can you please give me a reference to that?

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Regarding using one’s own research to argue a point, I don’t see
            anything wrong with that per se. The question is whether one’s research
            thus being cited is methodologically sound or not. In this case, Jaska’s
            research is (a quick perusal of historical Uralic literature would show
            you as much). On the other hand, Bouckaert’s research is not
            methodologically sound, as many linguists have argued on many different
            grounds. Such “research” as theirs deserves the sharp critique that it’s
            been getting from various quarters.

            The Mallory reference I was
            talking about is to a recent article in Language Relatedness (or
            Relationships?). I don’t have the exact reference handy, but could look
            it up later if you are interested. His point is that if Anatolian is the
            first one to branch off AND as Gray/Atkinson argue the non-Anatolian
            languages moved both east and west from Anatolia, how did they maintain
            the unity that characterizes them as Nuclear IE (or non-Anatolian IE).
            It doesn’t make sense in the real world. So ultimately Bouckaert paper
            “deduces” (really assumes) that Anatolian is the first-order split, but
            the rest of their theory doesn’t add up with that point.

          • Vladimir Mayakowski

            I said that to criticize a study is better to do with well known facts. What is methodologically sound or silent is yet to be argued.
            Yes I am interested in that article. I think there is well possibility as after Anatolian split off there is more than 1500 years to diversify!

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Well, if you talk about Pagel et al. or Bouckaert et al., there’s plenty of FACTS that contradict their theories (for Bouckaert et al., you can read our series on IE origins to see some of those facts.

            The Mallory article I was referring to is:
            Mallory, James P. (2013)
            Twnety-first century clouds over Indo-European homelands. Journal
            of Language Relationship 9:
            145-154.

          • Vladimir Mayakowski

            It seems it its not possible to get that paper. They publish in a Russian journal. So I can’t speak about the content.

            But I would like to speak about the title. It is so funny!

            Why clouds???

            I like very much how accurately and carefully Mallory speaks about Indo-European homeland in many of his publications. At least in those papers what I have. But there is such an impression that the title was the following – “Twnety-first century clouds over my favorite theory of Indo-European homeland”, and further he edited the end.

            I read his papers very long time ago, and what I remember is that he said that all the hypotheses of the homeland have their shortcomings, but Steppe Hypothesis is that what has the least shortcomings. But still it has many of them. And that gaps make that hypothesis questionable. Now when Anatolian hypothesis got more support, why he writes clouds? Why not “more lights” instead of “clouds”? Proponents of Steppe Hypothesis must be ready one day to agree that they were wrong or partially wrong, taking into account that that expansion can well did happen, but secondarily after the initial expansion in Anatolia.

            The Anatolian homeland has also many gaps, but Gray and Atkinson studies make its weight more. I think this hypothesis is now most probable. Nevertheless, I would suggest to its proponents also be ready that this hypothesis can be wrong as well.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            I thought you had access to Russian journals… The title is explained in the article, but think about “тучи сгущаются”… And it’s not about challenges to one particular theory of Indo-European homeland (his favorite!), but about the challenges that any serious theory of IE homeland has to address. “All the hypotheses of the homeland have their shortcomings, but Steppe Hypothesis is that what has the least shortcomings” is a pretty accurate description of Mallory’s position, as it is of Martin and mine. Unfortunately for the advocates of the Anatolian theory, it cannot explain the “gaps” in the Steppe theory, but it has a lot of its own. Gray & Atkinson’s studies do not “make its weight more” because they provide no explanation for those problematic “gaps” in the Anatolian theory. Just repeating the same lie until enough people believe it may be good enough for a totalitarian propaganda machine, but it won’t work in science. But all that advocates of the Anatolian homeland are doing is claiming that their theory is the best, without even beginning to address any of the many problems and challenges that have been pointed out to them.

          • Vladimir Mayakowski

            “I thought you had access to Russian journals… The title is explained in the article, but think about “тучи сгущаются”… And it’s not about challenges to one particular theory of Indo-European homeland (his
            favorite!), but about the challenges that any serious theory of IE homeland has to address.”

            In my opinion Gray&Atkinson studies promote the Indo-European language research. And this is more light for sure! No matter that Gray&Atkinson studies are good or bad, but they bring new approach, new facts and new arguments. This can’t be “тучи сгущаются”.

            A theory that have least shortcomings can’t serve for a scientist very hard to hold on to this theory. Because the theory still have shortcomings! But you, Martin and maybe Mallory also cling to Steppe Hypothesis very hard. I can’t understand how such a big scientist as Mallory can have such a biased view. It seems you would never accept Anatolian Hypothesis whatever would happen in future studies. Can you answer tho the following question? Do you ready to accept Anatolian Hypothesis ever?

            Gray & Atkinson’s studies give answer to a big shortcoming of Steppe Hypothesis – the fact of the deep divergence of Anatolian languages from others! They answer many other smaller shortcomings of Steppe Hypothesis. Of course Anatolian Hypothesis brought a new questions. But further studies are necessary to answer them. Do you want Gray & Atkinson to answer all of that questions? That is impossible for one study.

            “Just repeating the same lie until enough people believe it may be good enough for a totalitarian propaganda machine, but it won’t work in science. But all that advocates of the Anatolian homeland are doing is claiming that their theory is the best, without even beginning to address any of the many problems and challenges that have been pointed out to them.”

            Who speaks about totalitarian propaganda machine to whom??? The Steppe Hypothesis is the first candidate for that! For which totalitarian propaganda Anatolian Hypothesis can be beneficial? Instead of the cart the driver makes noise!

          • http://geocurrents.info Martin W. Lewis

            It is not true that Asya and I “cling to the Steppe hypothesis very hard.” It is true that we find it to be the most convincing theory, but we do not consider the case to have been closed. Our arguments against Gray and Atkinson are largely methodological, and hence would still stand regardless of where the IE language family actually originated.

            The deep divergence of the Anatolian languages, by the way, is most easily explained by an early separation and movement away from a steppe homeland. The presence of Hattic and other non-IE languages in ancient Anatolia are a huge problem for the Anatolia model.

          • Vladimir Mayakowski

            I like your position more regarding the “cling”. But it is not very much clear whether some methodological things would have much effect. Maybe they are not very much influential.

            “The deep divergence of the Anatolian languages, by the way, is most easily explained by an early separation and movement away from a steppe homeland. The presence of Hattic and other non-IE languages in ancient Anatolia are a huge problem for the Anatolia model.”
            Too much straightforward! That is not easy explanation! Moreover, I think that is not explanation at all!

            How Basque remained up to now alone in very much IE surrounding, but Hattic or other non-IE languages couldn’t remain at neighborhood? Especially when they eventually were disappeared.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            It’s not a full explanation, but a hint of one. You’ll have to wait for the full explanation for our forthcoming book.;)

          • Vladimir Mayakowski

            We can speak only about that what is now available. Who knows what would be in the future? I am sure you don’t think that your opponents would do nothing more in the future!

          • http://www.pereltsvaig.com Asya Pereltsvaig

            If the past is any indication of the future, they’ll do more of the same… :(

          • Vladimir Mayakowski

            For sure that will add much to our understanding the issue.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            What makes you say so? In the 140+ comments here (and more on our other posts), no a single argument or fact in defense of their approach has been put forward, no a single way in which they advance knowledge identified. When the best defense is to repeat the same mantra, the same empty slogans, that means there is no good arguments. So can we not continue this empty conversation? Unless you have anything of content to offer, of course.

          • Vladimir Mayakowski

            I offered. And you can’t say anything against! If you would not continue just stop!

          • http://www.pereltsvaig.com Asya Pereltsvaig

            You offered no solid arguments that not flawed or misinformed. If you think you did, what are they?

          • Vladimir Mayakowski

            I will copy it for you!

            I am not a specialist in this
            field, but the following paragraph from their research says that first
            split off the Anatolian languages is most probable.

            “Our results incorporate phylogenetic uncertainty
            given our data and model and so are not
            contingent on any single phylogeny. However,
            phonological and morphological data have been
            interpreted to support an Indo-European branching
            structure that differs slightly from the pattern
            we find, particularly near the base of the tree
            (16). If we constrain our analysis to fit with this
            alternative pattern of diversification, we find even
            stronger support for an Anatolian origin (in terms
            of Bayes factors, BFSteppe I = 216; BFSteppe II =
            227) (15).”

            In reality I don’t know what communications you have with Gray, or
            any other person. And one side reference to a personal communication is
            not an argument yet.

            And answering to your nonsense question would be also nonsense.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            “the following paragraph from their research says that first
            split off the Anatolian languages is most probable” — haha, it does not. It says that their method “proves” the Anatolian origin, which is not the same as the Anatolian being the highest-order split. Maybe they assume that the Anatolian origin follows from the Anatolian being the highest-order split, but that assumption is itself wrong. Moreover, that the Anatolian being the highest-order split is what they ASSUME, not DEDUCE. So in fact their “results” weakly emulate those of the comparative method, without offering anything substantial. Only people who don’t know what they are talking about — such as yourself — are impressed. Why don’t you find yourself another forum to argue about something you actually know something about?

          • http://www.pereltsvaig.com Asya Pereltsvaig

            I’m comparing the Gray Atkinson PR machine to a totalitarian propaganda machine: no facts, only slogans, and the way people fall for it’s dazzling facade is similar.

          • Vladimir Mayakowski

            This is your opinion only that there are no facts! This your sentence could be very far from truth.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            The burden of proof is upon you to show that we are wrong. So far you said, you claimed it, you maintained it, b hut you’ve offered zero proof that it’s the case. What did they offer that’s new or that explains well known facts or offers insight?

          • Vladimir Mayakowski

            Upon me??? Are you sure? You are linguists, not me.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Yes, if you got yourself involved in this discussion, you should be able to offer arguments for your positions, not just empty slogans. Or admit that you don’t have any, and since we have solid arguments that they are wrong that they are wrong. Either play the game or admit defeat.

          • Vladimir Mayakowski

            What I claimed I gave arguments! That is your gap that you don’t understand yet!
            Solid arguments must be used properly! That are scientific category, but not a tool to fight!
            Whatever you don’t like you call “empty slogan”. Maybe you are very much clever, but your opinion is not the final decision!

          • http://www.pereltsvaig.com Asya Pereltsvaig

            You didn’t give arguments. If you did, please repeat them briefly here. I am not going to search through 150 lengthy comments for tiny grains of sensible thought. If you can’t summarize your position, then they get out of this discussion. Because the smartest thing you ever said here is that you are not a linguist and would not discuss linguistic matters.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Regarding the use of one’s studies to mount a critique, I don’t see a problem with that, as long as those studies are done in a methodologically sound way. I believe enough general arguments are provided in that critique to demolish the Gray/Atkinson argument.

            Regarding Anatolian being the first one to split off the tree: this is actually something that Gray/Atkinso ASSUME, not deduce. They use the consensus that the Anatolian branch is the first-order split to root their trees (the algorithm initially produces only unrooted trees). That’s why all their bootstrap trees (and the consensus tree) have Anatolian is the first split.

          • Vladimir Mayakowski

            You either don’t know the study of Bouckaert et al. 2012 or want mislead the readers. In Bouckaert et al. 2012 authors. The following is that part of their study.

            “The estimated posterior distribution for the
            location of the root of the Indo-European tree
            under the RRW model is shown in Fig. 1A. The
            distribution for the root location lies in the region
            of Anatolia in present-day Turkey. To quantify
            the strength of support for an Anatolian origin,
            we calculated the Bayes factors (21) comparing
            the posterior to prior odds ratio of a root location
            within the hypothesized Anatolian homeland (11)
            (Fig. 1, yellow polygon) with two versions of the
            steppe hypothesis—the initial proposed Kurgan
            steppe homeland (6) and a later refined hypothesis
            (7) (Table 1). Bayes factors show strong support
            for the Anatolian hypothesis under a RRW
            model.”

            So they definitely deduce!!!

          • Jaska

            Vladimir:
            “You must not still bring it as the well grounded example to disapprove another study.”

            Why not? It is the best grounded view at the moment. What could be better starting point than the best grounded view?

            Vladimir:
            “Contradiction is that you made the possibility of their conclusion as little as 3.125%. And this you did based on extensively on your own studies.”

            No, that one you got wrong. The possibility is that small because in every step of their conclusion chain there are more than one possibility, although they only saw one possibility. That is pure logic, and I only give examples to illuminate these other possibilities.

            Vladimir:
            “I understand and agree with you that there are many weak points in that study, some of which I got here. In my first post I wrote that this strategy needs refinements. Moreover, I don’t want to “fight” instead of Gray and Atkinson. Let them to come and argue their study. And your critique can help so solve somehow many problems of their research. But in your critique you speak very sharply, not constructively, making their study just nothing.”

            They have been criticized for a decade now, and still their knowledge on historical linguistics does not seem to increase, and they mainly continue to ignore that there even exist other possibilities than the one they like to see. It is a natural reaction to weak hearing to raise one’s voice and articulate more clearly – hopefully it helps the message to find its target.

            Vladimir:
            “Not only the fact that Anatolian split off first, approves that Indo-European origin is in Anatolia. The splitting sequence also support to it – Anatolian languages first, and then Armenian-Tocharian branch in Bouckaert et al (2012), and Hittite first and then Tocharian, and then Armenian-Greek branch in Gray et al.(2006). Here I see a trace of movement.”

            You forget that the lexical level provides very weak and unreliable evidence, and therefore the splitting order of the Gray School should not be taken as the true one. Historical linguists have presented different groupings: Armenian is not connected with Tocharian as much as it is with Greek. After the Anatolian branch Tocharian may well have been the second to split off, but then there are large groups of Northwestern IE and Southeastern IE (Graeco-Armeno-Aryan). This kind of T-shaped grouping of early IE-dialects does not support the Anatolian homeland.

            Remember, that wherever the other primary unit was outside Anatolia, it is always possible that IT remained in the original homeland and not the Anatolian branch.

            I suggest that you read my links once or twice more, because answers to your counter-arguments are there already… For example one reason why they get/deduce the Anatolian homeland, is their mistake to allow the Aryan languages spread straight to the east from Anatolia. This is of course impossible, because we know that the Aryan development occurred on the North Caspian steppes.

          • Vladimir Mayakowski

            “Why not? It is the best grounded view at the moment. What could be better starting point than the best grounded view?”

            Considering own study as the best grounded view is the same as they (Gray and Atkinson) does.

            “No, that one you got wrong. The possibility is that small because in every step of their conclusion chain there are more than one possibility, although they only saw one possibility. That is pure logic, and I only give examples to illuminate these other possibilities.”

            At least at one link the possibilities can be different from that you suggested. This can lead to another % at the end.

            “They have been criticized for a decade now, and still their knowledge on historical linguistics does not seem to increase, and they mainly continue to ignore that there even exist other possibilities than the one they like to see. It is a natural reaction to weak hearing to raise one’s voice and articulate more clearly – hopefully it helps the message to find its target.”

            I mean not necessarily about them. Another group can do such research which will use their methods with taking into account your critique and suggestions. I am sure that there are many people who can do that statistics.

            “You forget that the lexical level provides very weak and unreliable evidence, and therefore the splitting order of the Gray School should not be taken as the true one. Historical linguists have presented different groupings: Armenian is not connected with Tocharian as much as it is with Greek. After the Anatolian branch Tocharian may well have been the second to split off, but then there are large groups of
            Northwestern IE and Southeastern IE (Graeco-Armeno-Aryan). This kind of T-shaped grouping of early IE-dialects does not support the Anatolian
            homeland.

            Remember, that wherever the other primary unit was outside Anatolia, it is always possible that IT remained in the original homeland and not the Anatolian branch.”

            The branching pattern suggested with historical linguists also has many drawbacks. So your methods can bring to wrong topology as well. So why not to join these two approaches?

            “I suggest that you read my links once or twice more, because answers to your counter-arguments are there already… For example one reason why they get/deduce the Anatolian homeland, is their mistake to allow the Aryan languages spread straight to the east from Anatolia. This is of course impossible, because we know that the Aryan development occurred on the North Caspian steppes.”
            I am not going to study Indo-European language origin and timing. What I am interested in I have already gathered at this point. One thing I would like to say is – nothing is impossible in the field what is not yet enough studied!

          • Jaska

            Vladimir:
            “Considering own study as the best grounded view is the same as they (Gray and Atkinson) does.”

            Dear Vladimir, haven’t you ever learned, that in science only arguments count? Can’t you see the difference between the two cases:

            1. The Gray School claims that their results are correct, but it is easy to present tens of problems and weaknesses in their method, which just cannot be explained away. Critique is well grounded, and it makes their results very uncertain.

            2. I claim that Samoyed was not the first branch to split off from Proto-Uralic, and I argue why the earlier views were incorrect. There are no pointed weaknesses in my method, at least so great that they would make the new results more uncertain as the traditional view. Nobody has presented counter-arguments questioning my method.

            So you see, when the historical linguists see biased results based on an unreliable method, they say it aloud. And when they see valid results based on a reliable method, they say it aloud. That is exactly what I have done here. It is totally irrelevant whether my own research belongs to the first or the second group – relevant is only that I criticize the unreliable method and praise the reliable method.

            So no, I’m not doing like the Gray School does; that is, I don’t ignore all the critique for the decades and stubbornly stick in my own method, many unreliabilities of which has been exhaustively pointed out by numerous critics.

            Vladimir:
            “At least at one link the possibilities can be different from that you suggested. This can lead to another % at the end.”

            If you read my paper, you would know that I handle exactly that point in the text. It belongs to the basic politeness in a public discussion that you read, understand and memorize the points in the very writing which you talk about… It is frustrating for me to write it all many times for you, because we all would have it easier, if you just read my drafts thoroughly enough.

            There is nothing wrong if you ask something, but that you make claims to which I already answered there, that is frustrating.

            Vladimir:
            “The branching pattern suggested with historical linguists also has many drawbacks. So your methods can bring to wrong topology as well. So why not to join these two approaches?”

            I have argued why the phonological level is much more reliable than the lexical level, when we are trying to find out the taxonomic structure of a language family. Your task now is to show that that is not the case; that the critique is invalid; and that there are some uncertainties considering the phonological level. Can you do that? If you cannot, do not claim that there is something wrong with the methods of traditional historical linguistics, or that the two methods are equally reliable. At the moment all the arguments point to the result that the lexicostatistic computational phylogenetic method is far more unreliable than the traditional method based on phonological innovations (supported by with morphological level) compared to the reconstructed protolanguage.

            How could we join the two methods? Their results are contradicting, and one of them is full of flaws… There would be no point to find an average between the results in this case, when one method alone gives more credible results than two methods combined.

          • Vladimir Mayakowski

            No need to write anything! Just take the very recent paper of Honkola et al 2013 and read! “Cultural and climatic changes shape the evolutionary history of the Uralic languages”.

            I am sure you will say that that study is as well wrong.

            Every study is wrong, except yours :)

          • http://www.pereltsvaig.com Asya Pereltsvaig

            This article is in the pre-publication stage and hence not available electronically — do you have a copy? Then we can all judge for ourselves if it’s right or wrong!

          • Vladimir Mayakowski

            I have that study, but it would be better for you to ask the authors to send.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Good idea, but I don’t know who they are and don’t have their contact information to ask.

          • Vladimir Mayakowski

            http://files.mail.ru/2335C34922A34FFEAD581894201F05C3
            This is the link for two papers. One is what I mentioned. The other one is another paper from the same group.

            1. The software they used can deduce the root without indication. The software is the same as Gray group uses.

            2. For Samoyed first split off they have firm argument.
            After downloading please write here.

          • Jaska

            Vladimir:
            “I am sure you will say that that study is as well wrong. Every study is wrong, except yours :)”

            There is no need to start trolling, even though you run out of arguments. :-)

          • Vladimir Mayakowski

            Actually I don’t “fight” against you!!!

            My arguments you interpreted yourself, what in some cases is out of the meaning what I said.

            I have my own view to the problem (research using “Gray school” approach must continue, critique must be for construction not destruction, this is small list of my viewpoints). I can for sure answer each of your questions in the previous post. But I saw that that is useless. You can criticize them as much as you want, but the research is blossoming. Time after time often and often new publications emerge with that method, other scientists and groups involve in that research. And I hope that critique and data from historical linguists would polish their results!

          • Jaska

            Vladimir:
            “Actually I don’t “fight” against you!!!”

            No fight. You have defended the method of the Gray School, I criticized it. I have defended my arguments, you tried to question them.

            Vladimir:
            “And I hope that critique and data from historical linguists would polish their results!”

            Yes, they should come out from the vacuum where their method is perfect and always gives only one possible result = the right one. There are many possibilities, and they should be aware of these and not over-interpret their own results so blindly.

          • Vladimir Mayakowski

            I wrote “fight” not fight, meaning in this case debate!

            Not mostly the method itself, but the approach and the alternative way of the research.

            Ok, I agree.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Actually, I know that study very well. The claim that they use Anatolian as first-split to root the tree is from Russell Gray personally — I would assume he knows what they did or didn’t do?!

          • Vladimir Mayakowski

            Maybe further they use that strategy. But before it they did a study, which allow them to use Anatolian to root the tree.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            You are incorrect. MrBayes software that they use produces unrooted trees and something must be specified to root them. The “Anatolian = outgrouping” (based on what non-computational historical linguistics agree upon, not their own studies) is what Gray and Atkinson use. But you seem to know better than our computational phygenetics expert or even Russell Gray himself. A bit holier than though, aren’t you?

          • Vladimir Mayakowski

            I am not a specialist in this field, but the following paragraph from their research says that first split off the Anatolian languages is most probable.

            “Our results incorporate phylogenetic uncertainty
            given our data and model and so are not
            contingent on any single phylogeny. However,
            phonological and morphological data have been
            interpreted to support an Indo-European branching
            structure that differs slightly from the pattern
            we find, particularly near the base of the tree
            (16). If we constrain our analysis to fit with this
            alternative pattern of diversification, we find even
            stronger support for an Anatolian origin (in terms
            of Bayes factors, BFSteppe I = 216; BFSteppe II =
            227) (15).”

            In reality I don’t know what communications you have with Gray, or any other person. And one side reference to a personal communication is not an argument yet.

            And answering to your nonsense question would be also nonsense.

          • Jaska

            Vladimir:
            “The Anatolian homeland has also many gaps, but Gray and Atkinson studies make its weight more.”

            Errr, no. If a method is unreliable, the results do not strengthen any hypothesis. Only the support gained by reliable methods are relevant!

            Vladimir:
            “I think this hypothesis is now most probable.”

            You like to think so, because you haven’t understood what is wrong with their method. Try to read again the critique:
            http://www.elisanet.fi/alkupera/Problems_of_phylogenetics.pdf
            http://www.elisanet.fi/alkupera/Review_Pagel2013.pdf

            Vladimir:
            “But you, Martin and maybe Mallory also cling to Steppe Hypothesis very hard. I can’t understand how such a big scientist as Mallory can have such a biased view. “

            In science it is not a question of liking – it is a question of relevant arguments gained by valid methods. Mallory, Asya, me and many others do not LIKE the steppe theory, but we see it as the BEST ARGUED view. You would agree, if you just understood the many weaknesses and unreliabilities of the phylogenetic interpretations.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Well-said, Jaska! Thanks!

          • Vladimir Mayakowski

            Statistics can capture such a signals that other methods and approaches and people can’t! We see the results of it and it will see more in the future! You can find some minor things to criticize, but that things may not be much influential. We’ll see it in the future.

            “In science it is not a question of liking – it is a question of relevant arguments gained by valid methods. Mallory, Asya, me and many others do not LIKE the steppe theory, but we see it as the BEST ARGUED view. You would agree, if you just understood the many weaknesses and unreliabilities of the phylogenetic interpretations.”

            I agree with your first sentence. But what is a continuation – a critique, must be without liking too. Which critique I saw here is fulfilled with “liking”. You use not a scientific way. You don’t leave any possibility for other side. You can say that the possibility of the opposite side is very little. But little does not mean zero! So you must leave some room for that. I see very sharp critique here. That is beyond a good science!

            It is very surprising that you praise you and your supporters very much, but you have no any valuable publication during recent years to argue your hypothesis. They publish their studies in Nature, Science and PNAS, you publish yours in such journals that nobody properly knows. And the same time you speak about what is good science and how science should be???

          • http://www.pereltsvaig.com Asya Pereltsvaig

            “Statistics can capture such a signals that other methods and approaches and people can’t!” — such as???

            “They publish their studies in Nature, Science and PNAS, you publish yours in such journals that nobody properly knows. ” — yes and the problem is with journals, not with the researchers. The newspaper “Pravda” published that the USSR is the “workers’ paradise” and little samizdat publications showed the opposite. And who was right? We publish in serious linguistic journals, not general journals/magazines. That you don’t “properly know” serious linguistic journals suggests that you got into an argument about something you don’t know much about.

          • Vladimir Mayakowski

            “such as???” Such that you can’t imagine or infer with your methods!

            The problem is everywhere except you? Maybe the problem is in your research? You pub “Pravda” and Nature, Science and PNAS to the same line? What is “serious”? Nature, Science and PNAS are not serious? Go and publish in your “serious” linguistic journals! But remember that nobody reads trash!

          • http://www.pereltsvaig.com Asya Pereltsvaig

            “”such as???” Such that you can’t imagine or infer with your methods!” — give us one specific example.

            “Nature, Science and PNAS are not serious? ” — no they are not. Something that you not being a linguist don’t know. it’s a joke in the field. no serious linguist ever publishes there. and for a good reason. read this:

            http://languagelog.ldc.upenn.edu/nll/?p=4652

            “Go and publish in your “serious” linguistic journals!” — this is exactly what we do.

            Advice for the future: don’t argue about things you don’t know. You make yourself appear an idiot.

            Since you have nothing of value to offer in this discussion, your further comments will be deleted.

          • Vladimir Mayakowski

            Is your way of debate deleting those comments which are against you and which you can’t disprove?

            And also – keep quietness at least in blogs! All people can offend others in blogs, but if you think you are a civilized person, then put a mask of civilized person at least here and don’t offend me. I can say many things what I really think about you, and all of them would be real characteristics of your personality, but I don’t do that.

          • Jaska

            Vladimir:
            “Statistics can capture such a signals that other methods and approaches and people can’t! We see the results of it and it will see more in the future! You can find some minor things to criticize, but that things may not be much influential. We’ll see it in the future.”

            Statistics can hardly tell anything which couldn’t be calculated manually or observed visually, too (with much effort, of course). Statistics is a way to handle data, and the output depends on the input! There was statistics already long before modern complex phylogenetic softwares.

            Vladimir:
            “I agree with your first sentence. But what is a continuation – a critique, must be without liking too. Which critique I saw here is fulfilled with “liking”. You use not a scientific way. You don’t leave any possibility for other side. You can say that the possibility of the opposite side is very little. But little does not mean zero! So you must leave some room for that. I see very sharp critique here. That is beyond a good science!”

            No, darling, here you go wrong again.

            It is very scientific to point out all the weaknesses of the method. On the contrary, it is very unscientific to ignore all the weaknesses – it becomes close to religious fanatics. So you see, I’m the one being very scientific here, without liking or disliking. You should not shoot the messenger (= critic) if the phylogenetic method to interpret the linguistic data swarms with errors and uncertainties.

            I have all the time talked about “practically impossible” and similar relative measures; I’m very aware that nothing here can be 100.00 % certain or uncertain.

            Vladimir:
            “It is very surprising that you praise you and your supporters very much, but you have no any valuable publication during recent years to argue your hypothesis. They publish their studies in Nature, Science and PNAS, you publish yours in such journals that nobody properly knows. And the same time you speak about what is good science and how science should be???”

            As you can see, my writings are drafts – I reacted quickly to the very recent publications of Bouckaert et al. 2012 (last autumn) and Pagel et al. 2013 (this spring). Surely you must understand that my points couldn’t be published so quickly. I saw it more important to spread the critique quickly than let it bury in the lines of manuscripts waiting to be published.

            But because you are such an authority-believer, I will be glad to extend my critiques and
            publish them in the future. :-)

            You should already know on the basis of the first comments of this article that a “valued” paper is not a guarantee for the quality!

            You should also understand that in science only arguments count, not the forum or the writer’s status. Your view only shows that you are unable to understand the arguments presented by the critical reviewers. Very sad, indeed.

            Please show me wrong – show that you have understand the arguments we have presented against their method! And feel free to counter-argue, once you have understood them.

            And as Asya already said, serious linguists publish in the serious linguistic papers. Serious Uralicists/Finno-Ugrists publish in the publications like FUF and SUSA, as I have done (I’m a Uralist, if you didn’t know). Nature or Science are “scrap-of-everything” kind of papers.

            But by all means, be as stubborn and ignorant as you like and refuse to consider any argument which are not published in Nature or Science! :-D It tells more about you than about the arguments you don’t want to understand.

          • Vladimir Mayakowski

            Stereotypical thinking of you and your opponents kills me. Times change, but you stay there where you have been 100 years ago.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            This is exactly the sort of thing that I meant by “blindly”! Thanks!

    • http://www.facebook.com/rfmcdonald Randy McDonald

      “And by this you actually infuriate Iran as there are some claims of
      Azerbaijanian politicians to join this northwestern Iranian provinces to
      their country based solely on the fact that people on that Iranian
      region speak Azeri language.”

      The fact that the Azeri language is a Turkic language, not an Indo-European language like Persian, is an incontrovertible fact. Saying otherwise makes about as much sense as saying that Iran is on Mars, not Earth.

      Now, there are many things that can be drawn out from this linguistic fact. If one wanted, one could use the fact of the actual linguistic difference to justify (say) support for Azeri separatism in northwestern Iran. That’s not an inevitable conclusion, as far as I know.

      Claiming, as an effort to counter Azeri separatism, that Azeri is not a Turkic language because the language previously spoken there was Indo-European (is English not Germanic because most of England was once populated by Celts?), or because many of the neighbouring languages are Indo-European so therefore Azeri must be as well, is nonsensical. I’d argue that, if anything, making such ill-grounded claims for ideological reasons might well encourage Azeri separatism in Iran.

      • http://www.pereltsvaig.com Asya Pereltsvaig

        Well-said, Randy!

      • Vladimir Mayakowski

        I think you misunderstand what I have written.

        You don’t distinguish Azeri from Azari.

        Azeri for sure is a turkic language. That can be considered as a Caucasian dialect of Turkish. I don’t say that it is an Indo-European one. But Azari is Indo-European language as it is a West Iranian one. http://en.wikipedia.org/wiki/Azari_Language

        And the examples you bring are inadequate. Please read the references I give, and you would know better what is what.

        Azeri separatism is much more ill-grounded for many reasons. 1. Speaking with the same language does not mean that they are genetically as well the same ethnic groups. There are too many examples, especially in Turkic language family.

        2. In Azerbaijan Azeri speakers are much less than in Iran. So if they think they must join together then Azerbaijan must be joined to Iran.

        3. Who did ask Azeri speakers in Iran whether they want to be separated from Iran? Small number of those who wants is not important.

        • http://www.facebook.com/rfmcdonald Randy McDonald

          You took objection to Pereltsvaig’s sentence “Turkic Azeri, with its 23 million speakers spread across Azerbaijan and
          northwestern Iran, is classified as an Indo-European language”. You made no reference to any previous languages spoken in the areas of northwestern Iran where the Turkic Azeri language isn’t spoken, so neither did I.

          • Vladimir Mayakowski

            Reference is in the next paragraph as “Grugni et al 2012″

          • http://www.facebook.com/rfmcdonald Randy McDonald

            A citation of a paper dealing with the biological inheritances of northwestern Iran’s population, without making reference of its content, isn’t a very useful in-text citation, especially since we’re talking about the _cultural_ characteristics of the contemporary population of northwestern Iran. Which, as it happens, speaks a Turkic language called Azeri that is not closely related to Persian (or, for that matter, to the Indo-European Azari language once spoken there).

          • Vladimir Mayakowski

            Because there are only limited number of studies about the issue. This silence is in accord with USA policy against Iran. Fortunately there is a wikipedia page concerning that language. I wrote it in one of my previous posts: http://en.wikipedia.org/wiki/Azari_Language
            I brought this fact to say that when one criticizes another peoples’ study, must keep in mind that there could be another explanation as well.

          • http://www.facebook.com/rfmcdonald Randy McDonald

            The issue is irrelevant to the subject being discussed, namely, the fact that northwestern Iran is largely populated by speakers of a Turkic language who are not represented on a map that purports to show the distribution of language families in the contemporary world.

          • Vladimir Mayakowski

            It seems we are going into the cycle! Well I will say again what I have said already once.

            If we will draw the language areas by now living speakers, then we must draw with Indo-European color (yellow in the study) also Nagorno-Karabakh de facto Independent republic and Talysh areas!

            Decide yourself what is correct, but do drawing with the same way in each case, and don’t do one place according to one way, and the other case with another way! The same thing concerns critique as well.

            This was what I am saying since the beginning.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            If one wanted to be accurate, indeed NK and Talysh should be colored IE, but I doubt it’d be visible on a map of this scale. The Azeri area in Azerbaijan is larger… Plus, since they seem to draw language areas based on geopolitical boundaries, even consistency argument requires Azerbaijan to be colored Turkic.

          • Vladimir Mayakowski

            No matter that scale will make them small. They must do equally for all the map!

            “Plus, since they seem to draw language areas based on geopolitical boundaries, even consistency argument requires Azerbaijan to be colored Turkic.”
            Than they must fill whole Russian Federation with Indo-European color or not?

          • http://www.pereltsvaig.com Asya Pereltsvaig

            If that’s what they do for other countries, yes. My point (and Martin’s) is that their cartography is sloppy and they don’t do anything consistently.

          • Vladimir Mayakowski

            I agree!

          • http://www.facebook.com/rfmcdonald Randy McDonald

            We could do that, sure. It’s significantly more important, however, to represent the 13 to 18 million speakers of Azeri in Azerbaijan first, before we go on to represent the million speakers of Talysh on the Iranian-Azerbaijani border and less than a quarter-million in Nagorno Karabakh. Why correct small mistakes if glaring ones aren’t being tackled first?

          • http://www.pereltsvaig.com Asya Pereltsvaig

            That’s an excellent point, Randy! As Martin said in the post, there are more glaring problems, such as those concerning Moldova or Estonia…

          • Vladimir Mayakowski

            There are no 13-18 million Azeri speakers in Azerbaijan. But I suppose you wrongly wrote Azerbaijan instead of Iran.

            Mistake is mistake! Small or big, no matter, in scientific literature must not be mistakes!

          • http://www.facebook.com/rfmcdonald Randy McDonald

            In the cross-border area, then. There are substantially more Azeris in Iran alone than there are Nagorno-Karabakh Armenians or Talysh combined. Complaining that the latter two groups aren’t represented while saying that Azeris shouldn’t be shown as Turkic-speaking because of a language shift five centuries ago is … I don’t know what.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Good-point, Randy!

          • Vladimir Mayakowski

            I have replied already to this question!
            Please don’t make the discussion annoying.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Randy made a valid point that bigger errors should be addressed first—why is it making the discussion annoying? Because it doesn’t accord with your views?

          • Vladimir Mayakowski

            I told that there is no big or small error! Error is error, and must not be in a scientific literature! Do you against this?

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Yes, I think that there are different types of errors, and while some are unavoidable, others should be corrected. If we only allowed absolutely error free articles to be published (or blog posts for that matter), we’d see no scientific progress made at all. In fact, the best scientific theory is defined by its (possibly erroneous) predictions.

          • Vladimir Mayakowski

            Everybody does mistakes. But not everybody readily accepts right critique! If there is a 100% mistake, and somebody else point that, it would be better if that person would take right suggestion or critique into account. And for sure this process will not make harm to the scientific progress. And I have not spoken about predictions at all.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Predictions is how one evaluates a scientific theory. Which ties with your question about “clinging”: if it turns out that the Anatolian theory makes better predictions or describes facts better, we’ll advocate it. However, it does not. For example, it doesn’t explain the Anatolian being the highest split. More generally, there’s a lot of dazzling math in Gray and Atkinson work, but not much new beyond the pretty facade.

          • Vladimir Mayakowski

            In this part we discuss coloring issue and mostly technical mistakes!

          • http://www.pereltsvaig.com Asya Pereltsvaig

            ???

          • Vladimir Mayakowski

            With Randy we debate on the issue about how coloring must be done. All your last comments do not concern to the technical part such as coloring.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Well, let’s be a bit more gracious than that! Anybody can make a mistake or a typo, only people who don’t do anything don’t make mistakes. And if you want to require 100% error-proof comments on a blog, why don’t you stay with the sentence in which you so require, and make sure *it* is error-proof (as far as English goes).

          • Vladimir Mayakowski

            Not in blogs, but in critiques!
            I have not written a critique!

          • http://www.pereltsvaig.com Asya Pereltsvaig

            This is a blog! And you were pointing out an “error” in somebody’s comment, not the article and not our response.

          • Vladimir Mayakowski

            We are discussing your post in blog, what is a critique. I argue that there must not be an error there. Where am I pointing out “error” in somebody’s comment?

          • http://www.pereltsvaig.com Asya Pereltsvaig

            I believe you were pointing out an error in Randy’s comment.

          • Vladimir Mayakowski

            Randy’s comment is about my comment about your post.

          • alysdexia

            small or great
            lite or big

        • http://www.pereltsvaig.com Asya Pereltsvaig

          Like Randy, I don’t think the fact that some extinct language (Azari) was IE makes a difference as the map is trying to capture contemporary distribution of languages. That there is more speakers of Azeri in Iran than in Azerbaijan is quite true, but I don’t see how any of this is relevant to the point Martin made in the post, namely that Azerbaijan (or large part of it) and possible a chunk of Iran should be shown as Turkic/Altaic rather than IE speaking. That’s a fact on the ground, despite what one may think or want.

          • Vladimir Mayakowski

            Read my answer to Randy. If you will not get what I say, feel free to ask again!

  • http://www.facebook.com/narek.ghazazyan Narek Ghazazyan

    Excellent article debunking one more rather silly error-riddled pseudoscholarship (though I’m a biologist myself).

    Отличный анализ. С удовольствием читаю ваши материалы ;).

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thanks, Narek!

  • Chris

    Thank you for this very well done reply to the original article! While I don’t think we can rule out such broad and ancient language families, I do think the original claim as stated was a bit much! I couldn’t even understand the Russian, and that’s only separated from English by what, 5 to 7 thousand years or so? I have the feeling that no random wanderer in middle Asia 15000 years ago would understand either the Russian or the English sentence given!

    My only (very minor) complaint is over “to whit”. Whit is a niggling little thing; where I am sure you want to say “to wit”, i.e., to know.

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thanks for pointing out the typo, I’ll fix it now.

      As for the English/Russian split, it’s probably much more recent, I’d say 4,000 years ago or so… When and how Germanic split off the rest of the IE tree is a bit of a gray area, but it clearly wasn’t among the first branches of IE to split off.

  • Pingback: How old is Proto-Dravidian? | varnam

  • Pingback: Thoughts on Pagel et al. on ultra-conserved words | Diversity Linguistics Comment

  • Vladimir Mayakowski

    I know that you’ll remove this post also, but you lost! You start to do dishonest things! You never can answer to my questions honestly!

  • Pingback: Euraasialainen superkantakieli kitisyttää humanistia | Kielen kannoilla

  • Pingback: Livius Nieuwsbrief / juni | Mainzer Beobachter

  • Pingback: Aantekeningen bij de Bijbel · Livius Nieuwsbrief / juni

  • Language lover

    I’m not a linguist but I always had a feeling that Indo-European, Uralic, Turkic and possibly Caucasian/Basque all belonged to one super family known as Eurasiatic. It makes perfect sense to me. They all must have come from a common origin. We are all human beings afterall. Eventually all languages should go back to one proto-human language. But atleast in terms of Eurasia, I always had a feeling that Indo-European, Uralic and Turkic were closer to each other than what most mainstream linguists think.

    I want to know what Ms Pereltsvaig thinks of my observations. I have a better way to connect the Eurasiatic languages than the Reading University researchers. Rather than words, I look for other things.

    Firstly, consider this. In IE, Uralic and Turkic languages, singular words become plural by adding a suffix to the end of the word. For example in English, car becomes cars. In Turkish, araba becomes arabalar. In Farsi, otomobil becomes otomobilha. In Hungarian, auto becomes autok. So as you can see, in these three language families (Indo-European, Turkic, Uralic), plural words become singular by adding a suffix to the end of the singular word.

    The pluralization rule of IE, Uralic, Turkic are not found in Semitic, African, Austronesian or Chinese-Tibetan languages. For example in Semitic languages, singular words become plural by splitting the word in half and adding letters in the middle of the word instead of the end. So these pluralizing sufffixes are only found in the “Eurasiatic” languages. So the Eurasiatic languages share a similar rule for pluralizing words.

    ^That is one similarity. Here’s another similarity:

    Many IE languages can act left-branching like Uralic and Turkic, to a certain extent ofcourse (most of the time theyre right branching). What do I mean? Well let’s take a look at some phrases that cannot be constructed in other language families.

    For example, this phrase: “English-speaking countries”

    In Semitic, Austronesian, Chinese-Tibetan and African languages, you cannot literally say “English-speaking countries”. Instead, you must say it the long way: “Countries where English is spoken”.

    But in Turkic, Uralic, Slavic, Germanic, Caucasian and Basque, you can literally say “English-speaking countries”. These languages also allow great degrees of hyphenation. Again, hyphenation is uncommon in Arabic, Hebrew, Chinese, Filipino, Malay, Zulu, etc.

    So these are atleast 3 major similarities.

    Another thing is language history. Technically, Uralic, Indo-European, Turkic, Caucasian have pretty much originated from the same area, somewhere in the Eurasian steppes, in the vicinity of the Caspian sea. This should indicate atleast a shared geographic heritage.

    Anyway, my question to Ms Pereltsvaig is what do you think about my ideas? Im not a linguist but atleast I gave it a shot. Also, why is it taboo to discuss a relationship between Altaic, Indo-European and Uralic? Afterall, it’s only logical that these languages go back to an even greater macro-family. It’s only logical. Nothing about the Eurasiatic theory defies logic. Humans were one identity hundreds of thousands of yrs ago, but then they split into different major groups eons ago. Then these major groups split into smaller groups. And then these smaller groups split into even smaller ones. All that those researchers are doing is putting the pieces back together.

    It started when people saw similarities between German and English. Then people saw similarities between Germanic and Romance. Then between Indo-Aryan and European languages. The next step is the similarities between Indo-European, Turkic, Uralic, etc.

    I don’t understand why there’s a huge fuss. I think with time and a little bit of technology, the Eurasiatic language theory might become more realistic and cast less doubt than it does today.

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Language Lover:

      I’m not a linguist but I always had a feeling that Indo-European,
      Uralic, Turkic and possibly Caucasian/Basque all belonged to one super
      family known as Eurasiatic. It makes perfect sense to me.

      I don’t know what your area of expertise is (if any), but I am sure if I said this about it, you’d be might offended. Or let’s try this with some other science, say, astronomy. How about: “I am not an astronomer but I always had a feeling that…um the moon is a star like the sun. It makes perfect sense to me.” — does it work? I doubt it.

      Now, I am not against laymen learning about linguistics, this is exactly what we are trying to do here at GeoCurrents, but it would make sense that you learn first before you make stupid pronouncements, no?

      Indo-European, Uralic, Turkic and possibly Caucasian/Basque all must have come from a common origin. We are all human beings
      afterall. Eventually all languages should go back to one proto-human
      language.

      In that case, why not claim that Indo-European and Khoisan form a language family together. After all, we are all human beings, no?

      I always had a feeling that Indo-European, Uralic and Turkic were closer
      to each other than what most mainstream linguists think.

      Most people have a feeling that the Earth is flat. The whole point of science is that it allows us humans to discover things that we don’t necessarily have intuitions about. Same goes for linguistics.

      So here are some facts for you (and mostly for other readers who are actually interested in learning!):

      In IE, Uralic and Turkic languages, singular words become plural by adding a suffix to the end of the word.

      So do the majority of the world’s languages. As you can see from the WALS map and its full legend below, these languages are found on all continents and in various language families. In fact, of the 9 options of how to express plurality (and if to express it at all), 48% of languages (513 out of 1066) choose one option: plural suffix.

      The pluralization rule of IE, Uralic, Turkic are not found in Semitic, African, Austronesian or Chinese-Tibetan languages.

      FALSE! Plural suffix is the way to encode plurality in Hebrew and Tigrinya (Semitic), Ijo and Khoekhoe (“African”), Paulohi (Austronesian), Mandarin (Sino-Tibetan).

      Regarding your left/right-branching issue. This is not correctly formulated, but you are not a linguist, so it’s okay. What you are talking about the headedness in compounds. As is clear if we compare, say, English and French, this typological feature does not correlate with language families: compounds are right-headed in Germanic (“bottle opener” is an opener, not a bottle), but left-headed in Romance (“apribottiglie” in Italian, “abrebotellas” in Spanish, etc.).

      So these are atleast 3 major similarities

      Actually two, but who’s counting. Or that neither of them holds true…

      why is it taboo to discuss a relationship between Altaic, Indo-European and Uralic?

      It is not taboo, just a waste of time, trying to prove something that’s not true. Might as well waste time “proving” that the Earth is flat.

      • Language lover

        With all due respect but what’s wrong? You didn’t have to be mean and you didn’t have to imply that I’m stupid. Was I wrong for sharing my ideas with you? I learned alot about languages, more than a layman could ever do in his entire life, but ofcourse Im not an expert in detailed aspects of it, nor have I claimed to be. Which is why I shared this with you. I took my time with it and now I feel embarrassed for even starting this. The power rests in your hands to enlighten people who want to learn. Conversely, you can also use your knowledge in this field to put people down in a humiliating way, which you now just did.

        Im a doctor, if you were wondering. And Ive had many people ask me innocent questions of “something that feels like it’s related to so and so” but I never ridiculed them for saying it. Theres no problem with it. Afterall science is not black or white, its full of grey areas and confusions. You cant compare my hunch for a Eurasiatic language with someone who thinks a moon is a star. Thats not a fair analogy, it’s like you’re taking me for some kind of fool. I only had a hunch in the first place because of similarities that I saw and geographic history, etc. Prove me wrong by all means, I’m humbled by your better knowledge in linguistics, but please dont tear someone apart for asking/sharing.

        I was going to ask about possibility of Indo-Uralic but now Im too afraid to even ask.

        • http://www.pereltsvaig.com Asya Pereltsvaig

          I did not imply that you are stupid because you lack knowledge of linguistics (sorry if you understood it that way). But I did imply that it is offensive to express an opinion on a scientific matter that is based on ignorance of facts and personal intuition. Science cannot make progress if scientists have to keep proving the most basic things to people who think that their “hunch” is as valid as scientific proof.

          Re: Indo-Uralic, there are some proposals put forward to that effect (e.g. Kortland 1995), but most linguists think that the similarities between IE and Uralic are a result of borrowing, not common descent.

          Kortlandt,
          Frederik (1995) General
          linguistics and Indo-European reconstruction. Rask 2: 91-109.

    • alysdexia

      The fewest men (Homines sapientes) ever alive were about 10,000 in East Africa after Toba’s supereruption about 74,000 years ago. I do not believe that size at that time could keep one whole language, so there should hav been scores of language isolates. CT is their hýpothetical polýmorf but it has not been founden; the population were composite of DE and CF, and this split is amenabil from the rifts and dales that bar their thrubreeding.

      http://en.wikipedia.org/wiki/Haplogroup_L3_(mtDNA)
      http://en.wikipedia.org/wiki/Haplogroup_CT
      http://en.wikipedia.org/wiki/Afar_Triple_Junction
      http://en.wikipedia.org/wiki/Refugium_(population_biology)

  • John Mason

    Asya, you have disgraced yourself by the manner in which you replied to some of these comments. If your unprofessional behavior is anything to go by, then it’s no wonder why linguistic sciences came to a dead end. Perhaps your bitterness—towards the University of Reading team that conducted this research and received global attention for it—might explain why you’ve replied so distastefully to those who sadly did not know better.

    To anyone else who reads this and thinks he or she could start a peaceful conversation with Asya, think again. This website might have a comments section, but don’t let that fool you into believing that a two-sided dialogue is welcomed here.

    If anyone doesn’t know this by now, this website’s authors are strictly against grouping language families together to form macro-families. Having observed GeoCurrents articles for more than a year, it has become painfully obvious that not even the Altaic family is accepted, much less something larger than that in the form of Euraltaic (or Eurasiatic; whichever term you prefer).

    The linguistic community is neatly divided into two camps:

    One camp has hit a brick wall, rested on its laurels and is failing to inspire/innovate. This camp accepted the status quo because it’s not daring or brave enough to leave its comfort zone. According to this camp, the well-established major language families we have today are all we can ever know about. The rest are all impossible to uncover, therefore “we might as well leave it at that”. Then, of course, we have the issue of personal pride and ego coming in the way of open-mindedness. It’s hard for this camp to acknowledge its falterings; thus, it will stick by its ideas until they become a lost cause because there’s nothing more that it can offer to the scientific community.

    The other camp, on the other hand, is brave, bold and daring. This camp doesn’t have insecurities to overcome; it is willing to go to great depths in search for the truth. It’s ready to embark on uncharted waters. While this camp is a minority in the field of linguistics, the good news is that it’s growing and gaining strength. And with the help of other anthropological fields, as well as other sciences, this camp can establish a combined effort in its pursuit for further knowledge. Unlike the other side, these contributors do not think human history began 15,000 years ago.

    The future of linguistics no longer rests on the hands of people like Asya. Newer techniques will be introduced, more intensified cross-linguistic studies will be made, other sciences will be invited to form a combined effort to solve problems and analyze data, and of course the ever-improving technology will make the unthinkable thinkable (as someone else already mentioned).

    To anyone who might be reading this and was wondering about the validity of the Euraltaic family, I will say this… More research needs to be done. There needs to be a lot of patience because these studies sometimes depend on tools that might not be available today. But I will also say this… The Euraltaic theory is nothing new. It has been ongoing for many years. Eastern universities put more effort to research this subject than universities in the western world, let alone if you’re living in the US. There already exist some unaltered grammatical similarities between Uralic and Altaic that suggest a historical relationship. There already exist similar words between Altaic (Turkic & Mongolic) and even the furthest Indo-European languages. And these words aren’t loanwords; neither could they have been borrowed with thousands of miles separating them. There are a handful of exciting academic blogs I have read that made genuine comparisons between Euraltaic languages, otherwise not mentioned by western universities because their members either didn’t care or didn’t put enough effort to learn these languages themselves.

    Fortunately, a western university in the UK has reignited the quest. If only our schools cooperated with east European and south Asian schools earlier; we might have uncovered many things by now.

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thank you for sharing your thoughts. As this post alone has 193 comments to date, you can’t say we don’t encourage (and spend a lot of our time on!) two-sided dialogue on this site. This is just as offensive a comment as many others we’ve respectfully engaged on this site. What you fail to see is that a dialogue is possible only when BOTH sides have something informative to say and BOTH sides are respectful to each other, including the side you call “brave, bold, and daring” (and I would call “foolhardy”, and for a good reason). Another thing you’ve failed to notice is that this so-called “other camp” isn’t really a camp, but various separate research teams whose only common denominator is slinging mud at serious scholarship, because that’s what gets your name into the press, that’s what draws publicity and “laurels” as you call them. However, that “other camp” fails to prove anything or even to put forward strong evidence for their claims, evidence based on facts, not pure propaganda. Your comment, sadly, reads like more propaganda, with bold slogans, but not much more. This is not what this site is all about: we try to educate people about the facts that can be proven, no matter which “camp” proves them or discovers them or whatever. Here’s just one example: you claim that “There already exist similar words between Altaic (Turkic & Mongolic)
      and even the furthest Indo-European languages. And these words aren’t
      loanwords” — what is your proof that these words (what are they, by the way, so we are all on the same page?) are not loanwords? Just saying that languages are now spoken far apart doesn’t help, because we actually know how easy it is to move across the steppe zone. Nor is your attempt to see science as a competition between regional camps is very appealing, to say the least. Here I will not get into the ideological underpinnings of such regional research schools, but it’s something worth a separate post some day. Let me conclude by a quote whose authorship I don’t remember, alas: “Open-mindedness is good as long as your brains don’t slide out”.

      • John Mason

        Asya, if you felt offended by my comment, then perhaps you should take a closer look at yourself and recognize what you wrote to the other readers around here. Respect is a two-way street. In case you have forgotten, let me quote some of your own words and remind you of some of the condescending passages that you responded with, to your followers:

        “Now, I am not against laymen learning about linguistics, this is exactly what we are trying to do here at GeoCurrents, but it would make sense that you learn first before you make stupid pronouncements, no?”

        “So here are some facts for you (and mostly for other readers who are actually interested in learning!):”

        Now, I must admit, I wasn’t at all interested to reply to this article because I already knew about your ready-made objections to the whole premise of the Euraltaic family hypothesis. Frankly, I didn’t want to waste my time reading your repetitively outdated answers like a broken record. But I only felt inclined to comment on the article, myself, after becoming disgusted by your unprofessional behavior towards the other passersby.

        Perhaps if you had conducted yourself more maturely, I wouldn’t have suspected you of fostering bitterness for the subject at hand; neither would I have been so harsh in my criticism.

        As for the claim of “propaganda” goes, here’s what I can safely say… Unless Mark Pagel or anyone else in his team was unknowingly part of a pan-Eurasian political movement, or unless he was inspired by none other than Biblical texts to search for Japhetic tribal lineages, then I hardly think the University of Reading researchers would be so eager or even foolish enough to spread propaganda in their own respected institution. Alas, for someone such as yourself, who thinks offensive dialogue is only a one-way street, you sure have no problem diminishing the integrity of these universities by calling their teams “foolhardy” or by accusing them of spreading propaganda. And just what exactly is the ill-intentional purpose of the Euraltaic family hypothesis? None whatsoever, as far as I’m aware; except that it goes against your own ideas, and thus becomes a struggle of pride and ego for your own camp.

        “Nor is your attempt to see science as a competition between regional camps is very appealing, to say the least.”

        Likewise, your scholastic cherry-picking does no good to the advancement of linguistics into the unknown. Rather, it is characteristic of a person who is determined to stay in his or her comfort zone. Sadly, there aren’t enough pioneers in linguistics. If there had been, the languages of the Eurasian phylum would have been successfully traced back to the Euraltaic tree sooner rather than later. Nevertheless, there are institutions that have dedicated themselves to search deeper into language history. So apparently, and only according to you, these so-called ‘regional’ universities have an ideological underpinning. Yet again, everyone who opposes Asya’s views must be some kind of a propaganda-driven, pseudo-scientist with an agenda. I doubt your statements are even qualified, with regard to what is and what isn’t a propaganda. All these other researchers in the east, who actually dare to speak multiple languages while studying them, must be engaged in some kind of a pan-Eurasian conspiracy to undermine Asya’s camp.

        You have to wonder, with those kinds of statements, how on Earth could I even present you their arguments if I already know the predetermined outcome that you’ll bring about from such dialogue?

        You have constantly ignored the arguments for grammatical similarities, lexical similarities (that aren’t borrowings) and even genealogical/geographic history. I have seen your responses in other related articles and they’ve been nothing short of sidestepping. What should make me think this time will be any different?

        Thank goodness western universities are beginning to move away from this outdated mentality. Mark Pagel’s work is only the beginning of something new, and long may it continue.

  • alysdexia
  • Pingback: European and Northern Asian; Linguistic Connection - Historum - History Forums

  • kdammers11 .

    “What who read is not the issue here.” This is an example of “what” and “who” collating. *”The every man” is one example where a particular type of adjective (using the author’s restricted sense of adjective) cannot be used with an article. A standard dictionary, e.g. the “American Heritage Dictionary” on my book-shelf or the Merriam-Webster online dictionary, lists “not” as an adverb.

    But my point here is not to question the specifics of the article, but to note that, valid though the fine distinctions are for critical study, the real question is whether or not those distinctions need to be considered in a broad-brush study such as is being discussed here. Some-times it is better to gloss over fine points in order to see the ocean for the waves. Not having yet read the original research article, I can’t say; but it does seem that Asya Pereltsvaig is being overly harsh with-out substantiation.

    • http://www.pereltsvaig.com Asya Pereltsvaig

      “”What who read is not the issue here.” This is an example of “what” and “who” collating.” — Huh?

      “*”The every man” is one example where a particular type of adjective
      (using the author’s restricted sense of adjective) cannot be used with
      an article.” — Who calls “every” an adjective? I certainly do not. If you are able to read and process simple English, it should be obvious that I called it a quantifier, which is exactly what it is. While Pagel et al. call “this” and “that” adjectives, I argued that they are not.

      A standard dictionary might be a good source of information on the meaning, pronunciation and sometimes etymology, it is hardly a linguistic reference source. Dictionaries employ traditional parts-of-speech classification based on the meaning of words, with “noun”, “adjective” and “verb” as the main categories, and “adverb” as the throw-away category for everything that isn’t a noun, a verb or an adjective, pretty much. (Oh yeah there’s also prepositions.) Knowing what source of information to turn to is obviously important: would you turn to a driving manual to find out why your car engine isn’t starting? But Pagel et al. probably would. They clearly never went beyond such an outdated classification scheme, and that’s about as good as writing a (n allegedly) cutting-edge physics paper using pre-Newtonian notions. As my post underscores, there’s a lot more to language than words and what they mean.

      So it’s not “fine points” they didn’t get right, but THE very basics. The sort of stuff that we teach in Intro to Linguistics, first 10 minutes of the first class. Where have these guys been?! Am I being too harsh by expecting scholars to know the alpha and beta (forget the omega!) of the discipline they pretend to work in?