Kalasmaic, a New IE Language.

A report from Julius-Maximilians-Universität Würzburg:

An excavation in Turkey has brought to light an unknown Indo-European language. Professor Daniel Schwemer, an expert for the ancient Near East, is involved in investigating the discovery. The new language was discovered in the UNESCO World Heritage Site Boğazköy-Hattusha in north-central Turkey. This was once the capital of the Hittite Empire, one of the great powers of Western Asia during the Late Bronze Age (1650 to 1200 BC). […]

Yearly archaeological campaigns led by current site director Professor Andreas Schachner of the Istanbul Department of the German Archaeological Institute continue to add to the cuneiform finds. Most of the texts are written in Hittite, the oldest attested Indo-European language and the dominant language at the site. Yet the excavations of this year yielded a surprise: Hidden in a cultic ritual text written in Hittite is a recitation in a hitherto unknown language.

Professor Schwemer, head of the Chair of Ancient Near Eastern Studies at Julius-Maximilians-Universität (JMU) Würzburg in Germany, is working on the cuneiform finds from the excavation. He reports that the Hittite ritual text refers to the new idiom as the language of the land of Kalašma. This is an area on the north-western edge of the Hittite heartland, probably in the area of present-day Bolu or Gerede. The discovery of another language in the Boğazköy-Hattusha archives is not entirely unexpected, as Prof. Schwemer explains: “The Hittites were uniquely interested in recording rituals in foreign languages.” […]

Being written in a newly discovered language the Kalasmaic text is as yet largely incomprehensible. Prof. Schwemer’s colleague, Professor Elisabeth Rieken (Marburg University), a specialist in ancient Anatolian languages, has confirmed that the idiom belongs to the family of Anatolian-Indo-European languages. According to Rieken, despite its geographic proximity to the area where Palaic was spoken, the text seems to share more features with Luwian. How closely the language of Kalasma is related to the other Luwian dialects of Late Bronze Age Anatolia will be the subject of further investigation.

Thanks go to Trevor, Dmitry, and Trond, all of whom alerted me to this find.

Comments

  1. Since the Hittites collected the rituals (or centralized the cults) of the conquered peoples making up the empire, we can hope for more languages to come. We could wish for a larger corpus of each language, but you take what you get.

  2. Yes on all counts. This is an exciting time to be an Indo-Europeanist!

  3. I want to say that the more Anatolian languages they find, the older we should assume the Anatolian branch to be, but I don’t even believe that myself. More interesting may be if it’s becoming clearer that Anatolian languages can be divided in two subgroups that must have diverged for a considerable time.

  4. Somewhat frustrating that the article doesn’t indicate even approximately how long the text is, and without that information it’s hard to even speculate about how much could be gleaned from it.

    Another issue is that the Anatolian languages that are attested vary pretty considerably in the time of attestation, so there are of situations where the earliest known inscription in language X is from many centuries after the last known one in language Y. Which suggests to me that one possibility you can’t necessarily rule out is that this is an earlier or ancestral form of some language only hitherto known in inscriptions from many centuries later.

  5. David Eddyshaw says

    How much do the Anatolian languages differ from one another? Romance-level or Indo-Aryan level?

  6. Hmm. Interesting.

    Just a passing thought: If the similarities linking Kalasmaic with Luwian (excluding Palaic) are indeed 1) real and 2) shared innovations rather than shared conservative features, this might suggest that Palaic is intrusive to the area (From further North perhaps?).

    This in turn relates to something I have occasionally wondered: Could the scholars who have assumed/defended the theory that Anatolian came to Anatolia from the West (Balkans) and the scholars who have assumed/defended the theory that Anatolian arrived in Anatolia from the East (Caucasus) BOTH be right? Could Proto-Anatolian have been spoken somewhere in what is today the Southern Ukraine and expanded both eastwards and westwards, hugging the Black Sea Coast, until both halves of the sundered speech community met again in Anatolia?

    (Yes, Trond’s comment on two sharply divided subgroups of Anatolian is what triggered that particular train of thought).

    David Eddyshaw: I do not think anyone could answer your question on how different from one another Anatolian languages are: My understanding is that Hittite is the only one with enough documentation for us to claim that we have more or less complete knowledge of it, with all the other Anatolian languages being attested in such fragmentary corpora that huge chunks of their vocabularies and grammars remain totally unknown to us. The fact that they have been in close contact with one another makes the whole issue even more difficult, of course.

  7. Could Proto-Anatolian have been spoken somewhere in what is today the Southern Ukraine and expanded both eastwards and westwards, hugging the Black Sea Coast, until both halves of the sundered speech community met again in Anatolia?

    Now, there’s a basis for an epic poem if I ever heard one…

  8. David Marjanović says

    It seems that the Hittites collected the rituals (or centralized the cults) of the conquered peoples making up the empire.

    Like the Romans, the Hittites believed that all deities anyone worshipped were real, so they had to worship them too or risk pissing them off.

    Unlike the Romans, the Hittites believed the worship had to be done in the original language. Palaic is another languages that is known today exclusively from prayers in a Hittite context, and Hattic has that plus a myth with Hittite translation.

    How much do the Anatolian languages differ from one another? Romance-level or Indo-Aryan level?

    The latter, easily.

    If the similarities linking Kalasmaic with Luwian (excluding Palaic) are indeed 1) real and 2) shared innovations rather than shared conservative features

    A rather striking one is that the whole Luwian branch has the satəm shift (though not the satəm merger!), while Hittite and Palaic have the kentum merger at least graphically (but probably for real, given the lack of ki/ik spellings analogous to the ku/uk spellings used for the labialized velars in all Anatolian languages written in cuneiform).

    This in turn relates to something I have occasionally wondered: Could the scholars who have assumed/defended the theory that Anatolian came to Anatolia from the West (Balkans) and the scholars who have assumed/defended the theory that Anatolian arrived in Anatolia from the East (Caucasus) BOTH be right? Could Proto-Anatolian have been spoken somewhere in what is today the Southern Ukraine and expanded both eastwards and westwards, hugging the Black Sea Coast, until both halves of the sundered speech community met again in Anatolia?

    …yes, except… for the complete lack of Steppe ancestry in Bronze Age Anatolia. People buried by Mycenaeans have a bit, people buried by Hittites have none whatsoever.

    It does seem like Proto-Indo-Tocharian was spoken on the steppe, but Proto-Indo-Anatolian south of the Caucasus.

  9. Adiego compiled a dictionary of Carian, which is a Luwian descendent. And lots of hieroglyphic Luwian sometimes in Aramaic bilingual. I am optimistic with scholarly understanding of Hittite-era hieroglyphic Luwian (which also occasionally shows up in bilinguals here with Knesian).
    Admittedly our knowledge of Palaic is pathetic given our tiny corpus of that tongue.

  10. The fact that the Hittites appear to have adopted the name of the Hattians and their country when they took control of the area (as well as preserving some of the religious rituals in Hattic and according them high status) has always struck me as very interesting. The fact that, hundreds of years later, there were Canaanite, apparently Semitic hill tribes also known by a similar or identical name makes it seem all the more curious.

    I have asked myself in the past whether we would even be able to tell if the attested Hattic language was actually a Hittite conlang. It doesn’t sound very likely. Yet perhaps after another three and half millennia, scholars of Hattic will be pondering the mystifying origin of such secondarily attested Hattic words as scandicongo and fylosc.

  11. Why “Kalasmaic”?

    If the placa name is Kalašma, shouldn’t it be “Kalashman” or “Kalashmese”?

  12. David Eddyshaw says

    It’s calqued on the Welsh name for the language, Calasmaeg.

  13. David Eddyshaw says

    the Hittites appear to have adopted the name of the Hattians

    Isn’t that a modern mistake, rather than an autonym? The Hittites preserved the name of the land they conquered, but did they call themselves “Hittites”? I know that they didn’t call their own language “Hittite”, but then Americans (mostly) don’t call their own language “American”, so that may not mean much.

    I suppose that the English call themselves “British” …

    Melchert (in his Hittite grammar) says that the Biblical Hittites in fact spoke Luwian. (But then, so did the Hittite Hittites, apparently, after a bit. Life is so complicated.)

  14. David Eddyshaw says

    The fact that the Hittites appear to have adopted the name of the Hattians and their country when they took control of the area (as well as preserving some of the religious rituals in Hattic and according them high status) has always struck me as very interesting

    When the Red Hunter and his band of desperadoes/romantic heroes from the East founded what became the Mossi-Dagomba states, they imposed chiefs on the previously chiefless peoples they conquered. However, there was never any question of replacing the tendaannam “earth-priests”, who continued, of necessity, to be drawn from the indigenes. You can conquer the people, but the land itself cannot be conquered and requires its worship to continue as before. (The tendaannam decided all questions of land ownership up until colonial times; they still matter.)

  15. @David Eddyshaw: The Hittites had an older endonym Nesili, which continued in use after their conquest of central Anatolia. However, they also did, at least eventually, refer to the people of their kingdom as Hatt(i)li. The former name may have been more linguistic, the latter geographical or political. Moreover, there is a fair chance that Nesili had itself replaced an even earlier endonym, during the period when the most powerful Hittite polity was centered in the city of Nesha.

  16. The fact that, hundreds of years later, there were Canaanite, apparently Semitic hill tribes also known by a similar or identical name makes it seem all the more curious.

    Your own tribe once more exists in its ancestral region speaking a Semitic language and using Semitic personal names. But they are also to be found in a rather larger zone to the west, where they speak an Indo-European language and use Indo-European personal names meaning things like ‘inhabitant of the Island of the Mighty’. So the matter is not so curious as all that.

    If the place name is Kalašma, shouldn’t it be “Kalashman” or “Kalashmese”?

    There is a single uniform transliteration for Sumero-Akkadian, Hittite, Eblaite, Elamite, Hurrian, Luwian, and Urartian (but not Old Persian or Ugaritic, which are only cuneiform-shaped, cuneiformiform, if you like). When Old Akkadian split into Babylonian and Assyrian, the originally distinct sounds /s/ and /ʃ/ merged in Assyrian (but not Babylonian) as /s/, but were written with the syllabograms for /ʃ/. As an early IE language, Hittite had only /s/, so the transliteration š is read /s/. So “Kalasmaic” is correct. Some older names like Carchemish should really be -mis.

    I suppose that the English call themselves “British” …

    Indeed. Tolkien, “English and Welsh”:

    The modern Englishman […] has long read of British prowess in battle, and especially of British stubbornness in defeat in many imperial wars; so when he hears of Britons stubbornly (as is to be expected) opposing the landing of Julius Caesar or of Aulus Plautius, he is apt to suppose that the English (who meekly put themselves down as British in hotel-registers) were already there, facing the first of their long series of glorious defeats. A supposition far from uncommon even among those who offer themselves for ‘honours’ in the School of English.

    But in early times there was no such confusion. The Brettas and the Walas were the same. The use of the latter term, which was applied by the English, is thus of considerable importance in estimating the linguistic situation of the early period.

  17. January First-of-May says

    the originally distinct sounds /s/ and /ʃ/

    IIRC there’s some evidence that even in the original Old Akkadian, š was actually /s/, and s was actually /ts/ or something (and the traditional labels are inherited from West Semitic cognates which did have something closer to /ʃ/ and /s/ respectively). I’m not very sure of the details, though.

    There is a single uniform transliteration for Sumero-Akkadian, Hittite, Eblaite, Elamite, Hurrian, Luwian, and Urartian (but not Old Persian or Ugaritic, which are only cuneiform-shaped, cuneiformiform, if you like).

    Hittite cuneiform is indeed fundamentally Sumero-Akkadian, Eblaite ditto, and I think also Hurrian and its neighbour/close relative Urartian; don’t know enough about Cuneiform Luwian but I think it’s also the same script?

    Elamite though AFAIK used different signs with different values; i.e., very roughly, Sumero-Akkadian : Hittite : Elamite : Ugaritic ~= Chinese : Japanese : Tangut : Hangul.

  18. yes, except… for the complete lack of Steppe ancestry in Bronze Age Anatolia. People buried by Mycenaeans have a bit, people buried by Hittites have none whatsoever
    How many specimens do we have? If the IE element was a relatively small minority and the burials were all of people belonging to the non-IE majority, that could explain the lack of steppe ancestry even if the IE element came from the steppe. The bigger the sample is, the less likely that scenario would be.
    don’t know enough about Cuneiform Luwian but I think it’s also the same script?
    Yes, it is. WP for confirmation”: “Cuneiform Luwian (or Kizzuwatna Luwian) is the corpus of Luwian texts attested in the tablet archives of Hattusa; it is essentially the same cuneiform writing system used in Hittite.”

  19. Where was Kalasma? Cursory googling found sources saying NE, N, or NW Anatolia.

  20. David Marjanović says

    there’s some evidence that even in the original Old Akkadian, š was actually /s/, and s was actually /ts/ or something (and the traditional labels are inherited from West Semitic cognates which did have something closer to /ʃ/ and /s/ respectively)

    Basically, the use of the symbol š comes from interpreting Akkadian through a Hebrew lens in the 19th century. Hebrew is one of not many Semitic languages that really have turned that phoneme into [ʃ].

    Elamite though AFAIK used different signs with different values; i.e., very roughly, Sumero-Akkadian : Hittite : Elamite : Ugaritic ~= Chinese : Japanese : Tangut : Hangul.

    Wikipedia:

    Elamite cuneiform comes in two variants, the first, derived from Akkadian, was used during the 3rd to 2nd millennia BCE, and a simplified form used during the 1st millennium BCE.[1] The main difference between the two variants is the reduction of glyphs used in the simplified version.[3] At any one time, there would only be around 130 cuneiform signs in use. Throughout the script’s history, only 206 different signs were used in total.

    So it’s more like Simplified Chinese limited to the set of characters currently taught in Japan.

    How many specimens do we have? If the IE element was a relatively small minority and the burials were all of people belonging to the non-IE majority, that could explain the lack of steppe ancestry even if the IE element came from the steppe. The bigger the sample is, the less likely that scenario would be.

    You’d need remarkably strict endogamy over easily a thousand years for that to work.

  21. PlasticPaddy says

    @dm
    I seem to remember that in the large Hittite frescoes or carved monuments, the officers (or chariotmen) and the foot soldiers were of different racial types.

  22. In general, the Akkadian sound transliterated with ‘š’ is the result of a merger of PS s₁ (ʃ), s₂ (ɬ), and θ, while ‘s’ comes from s₃ (ts). Hebrew ש also comes from the same merger as Akkadian ‘š’ (although s₂ took a bit more time in Hebrew).

    Spellings of Sumerian loans into Akkadian suggest that /ʃ/ was the main realization of ‘š’ in Babylonian/Southern Akkadian. Internal and external evidence suggests that Akkadian ‘s’ was an affricate /ts/. My understanding of the current consensus is that the pronunciation of š later shifted to /s/.

    Most cuneiform-literate cultures got their writing from Akkadian, so the use of š- and s-signs probably depended on the phonetic inventory of the target language (Elamite seems to have had an s-ʃ contrast, Hittite did not) and the sibilant inventory of the Akkadian dialect it came into contact with.

  23. David Marjanović says

    different racial types

    Never seen that before.

    PS s₁ (ʃ)

    Isn’t that much more likely to have been a retracted [s̠], as found today e.g. in Greek, Dutch, Finnish and other languages without a /s/-/ʃ/ contrast?

    Spellings of Sumerian loans into Akkadian suggest

    Honest question: How does that work? I thought the basic difficulty in finding out how Sumerian was pronounced, e.g. whether there was an /o/, is that all the information we have is filtered through Akkadian? Is there enough internal evidence to be sure that, say, s wasn’t an affricate in Sumerian? Are there direct loans to or from Egyptian that didn’t go through Akkadian (and West Semitic)?

  24. David M.’s comment above on the whole Luwian branch having “the satəm shift” made me wonder: WHAT IF the Luwian branch entered Anatolia from the East, and the non-Luwian ones from the West?

    And…WHAT IF originally Proto-Anatolian was indeed located South (i.e. in the Southernmost, including coastal, parts of the Ukraine), as I wondered upthread, with core Indo-European (=non-Anatolian Indo-European) spoken further North? Both would have been dialect continua (i.e. Eastern and Western Proto-Anatolian would already have begun diverging before either branch entered Anatolia), and the same would be true of Core Indo-European. Hmm. Could there have likewise been a dialect continuum between Proto-Anatolian and Proto-(core) Indo-European, leading to areally shared innovations, such as…the satəm shift affecting, in Core Indo-European and Anatolian alike, the Eastern part of both dialect continua?

    Thoughts?

  25. Honest question: How does that work?
    It comes from the way loan words are spelled when they go from Sumerian to Akkadian or vice versa. Cuneiform has s-signs, š-signs, and z-signs. The relevant correspondences are:
    Sum – Akk
    z – s
    š – š
    s – š

    If we assume that Akkadian ‘s’ is an affricate, the usual interpretation is that Sumerian ‘z’ is a voiceless affricate, while ‘s’ and ‘š’ distinguish two sibilants that aren’t distinguished in Akkadian, namely /s/ and /ʃ/, while Akkadian speakers found their sibilant closer to the ‘š’ series.

  26. Etienne’s speculative comment reminds me that I have occasionally wondered if anyone in the circles where people like to speculate about the IE Urheimat has worked the https://en.wikipedia.org/wiki/Black_Sea_deluge_hypothesis into their speculations. A now-submerged Urheimat located in between “the steppe” and currently-dry-land Anatolia seems like it ought to offer an elegant solution to various difficulties. Do the conjectured timelines match up closely enough? Well, they’re probably speculative on both sides anyway …

  27. Lars Mathiesen (he/him/his) says

    Yes, I have seen speculations that the Urheimat is now on the bottom of the Black Sea and that’s why we haven’t found archaeological traces. We even talked about it here once, but very briefly. IIRC, it has been calculated that the shore line would have moved something like 50m per day — not quick enough to drown people, but certainly enough to cause an uncoordinated exodus and shatter what was a dialect continuum into language communities separated by hundreds or thousands of kilometers.

  28. David Marjanović says

    The relevant correspondences are:

    Oh, that’s interesting. I wouldn’t have guessed that the s series don’t even overlap between the two languages.

    If we assume that Akkadian ‘s’ is an affricate, the usual interpretation is that Sumerian ‘z’ is a voiceless affricate, while ‘s’ and ‘š’ distinguish two sibilants that aren’t distinguished in Akkadian, namely /s/ and /ʃ/, while Akkadian speakers found their sibilant closer to the ‘š’ series.

    Yes, but nowhere does it follow that the Akkadian š series was [ʃ]. [s̠] continues to make a lot more sense. The correspondences also don’t tell us which of the Sumerian s and š series was /s/ and which was /ʃ/ – or if perhaps one of them was [ɬ] for example.

    A now-submerged Urheimat located in between “the steppe” and currently-dry-land Anatolia seems like it ought to offer an elegant solution to various difficulties.

    …except that, as the map in the article shows, most of the Black Sea is really ocean in the geological sense, with an ocean floor some 3000 m under the surface. That kind of depth laughs at an ice age. Almost all the continental shelf that used to be dry is west & east of Crimea; the distance between Crimea and Anatolia has hardly changed at all.

    Do the conjectured timelines match up closely enough?

    Actually, maybe. Both the Khvalynsk culture and the mixture of EHG and CHG ancestry seem to have started around 5000 BCE or soon after, and the Black Sea Flood was some 6800 BCE according to the article linked to above. Khvalynsk is a plausible ancestor of Pit Grave/Yamna(ya), which succeeded it with spatial overlap, and Afanasyevo, which was contemporary with Pit Grave but far to the east… so it was plausibly the culture of the speakers of Proto-Indo-Tocharian. If we add another thousand years to get to Proto-Indo-Anatolian, a few rounding errors might get us there… though I have to doubt it.

    EHG ancestry is what’s lacking in Bronze Age Anatolia.

    And…WHAT IF originally Proto-Anatolian was indeed located South (i.e. in the Southernmost, including coastal, parts of the Ukraine), as I wondered upthread, with core Indo-European (=non-Anatolian Indo-European) spoken further North?

    Khvalynsk is indeed quite a bit north of the Black Sea. But the people buried by the Srednyy Stog/Seredniy Stih culture that began a bit later farther south had some 40% EHG and 10% WHG ancestry, and the preceding Dnipro-Donets culture buried people who were pure EHG apart from a bit of WHG admixture. Both EHG and WHG ancestry are lacking from Bronze Age Anatolia.

    Could there have likewise been a dialect continuum between Proto-Anatolian and Proto-(core) Indo-European, leading to areally shared innovations, such as…the satəm shift affecting, in Core Indo-European and Anatolian alike, the Eastern part of both dialect continua?

    The satəm shift actually seems to have played out subtly differently in Anatolian and Indo-Actually-European. The Luwian branch of the former seems to have had a few conditioned exceptions* that are poorly understood; the satəm branches of the latter (a minimum of three: Indo-Slavic, Armenian, Albanian) seems to have had another conditioned exception** that the Luwian branch seems to have lacked as far as I’ve noticed. This is underresearched, however.

    This is not to be confused with the satəm merger, the delabialization of the labialized velars. It was simple and unconditioned in Indo-Slavic, but apparently less so in Armenian and Albanian, and didn’t happen at all in the Luwian branch.

    Keep in mind, too, that the easternmost IE branch (before Sogdian reached Liáoníng) was Tocharian, which is kentum. The kentum merger seems to have been simple and unconditional in all branches that had it (a minimum of four: Hittite + Palaic, Tocharian, Greek, Germanic + Crotonian + Italo-Celtic), but it was followed by another merger, the one of the *Kw clusters into the *Kʷ series… and that one only happened in Greek after it had acquired phonemic consonant length: the intervocalic outcome there is not *Kʷ, but *Kʷː, judging from híppos and IIRC maybe a few others.

    Further thoughts: the apparent lack of loanwords between PIE and Proto-Uralic suggests some geographic distance. The oldest IE loans that may have been present in Proto-Uralic are specifically Pre-Indo-Iranian (and there’s one that may be Pre-Tocharian); likewise, Indo-Iranian contains a few Uralic loanwords, but none have been suggested for PIE. Instead, loans between PIE and Semitic have been suggested; “wine” in particular has an IE-internal etymology, yet shows up in Semitic anyway.

    * depalatalization by following *o or something…? but not by following *w…?
    ** Weise’s law: depalatalization by following resonants. If real, this must have created a lot of grammatical chaos as full-grade forms underwent the shift but zero-grade forms didn’t. That may explain why many such roots occur in a shifted and a depalatalized form, sometimes both within modern Lithuanian… but it also means a lot of evidence has been erased and is indistinguishable from, say, Pre-Germanic or Celtic loanwords, or Crotonian ones (good luck with those), or Pre-Tocharian ones perhaps on the other side.

  29. John Cowan says: “When Old Akkadian split into Babylonian and Assyrian, the originally distinct sounds /s/ and /ʃ/ merged in Assyrian (but not Babylonian) as /s/, but were written with the syllabograms for /ʃ/. As an early IE language, Hittite had only /s/, so the transliteration š is read /s/. So “Kalasmaic” is correct. Some older names like Carchemish should really be -mis.”

    Thanks JC & everyone else who commented on this.

    Sounds like it’s a moving feast. Which is why I love this field. It’s been a while since I learnt Akkadian, but we differentiated /s/ and /ʃ/.

    What’s the latest with s and š in Sumerian?

  30. David Marjanović: I had thought of Tocharian while typing out my scenario above, and I do not believe it is a counter-example. On the contrary: there is no reason why Tocharian could not have shared the innovations which make core Indo-European what it is and broken off from the rest of core Indo-European before satəmization as an Eastern areal feature (Within CIEMT -“Core Indo-European minus Tocharian”- and of Proto-Anatolian alike) had arisen and/or spread. That various later dialects of Eastern Anatolian and Eastern Core Indo-European differ in some details regarding the distribution of said satəmization is to my mind no counter-argument either.

    Now, all I need to do to make this notion of mine seem less implausible than it does now is find some other feature(s) shared by Luwian languages within Anatolian on the one hand and by (at least some) satəm languages within Indo-European on the other…

  31. David Marjanović says

    Fair enough. The strongest counterarguments remain the genetic ones.

  32. Mankowski’s Akkadian Loanwords in Biblical Hebrew (p. 156) says,

    Particulaly significant is the dialectal variance between Babylonian and Assyrian. The sibilants written with š-signs had the Babylonian pronunciation [š] and the Assyrian pronunciation [s]; conversely, sibilants written with s-signs had the Babylonian pronunciation [s] and the Assyrian pronunciation [š]. This is evident from the Aramaic treatment of loans and transliteration of PNs, as well as from the BH evidence, and confirmed by interdialectal borrowings in Akkadian. Further corroboration is given by words in which another phoneme in addition to the sibilant shows a characteristically Assyrian or Babylonian treatment.

    The latter refers in particular to intervocalic k voicing in Assyrian, and intervocalic *m>w in Neo- and Late Babylonian. Many examples are given.

  33. January First-of-May says

    the easternmost IE branch (before Sogdian reached Liáoníng) was Tocharian

    Technically, Kamarupa was (and its successor Assam is) some degrees to the east of Turfan (and many to the south), and apparently already Indic-speaking in the first millennium AD if not even earlier.

    Do we know when Tocharian arrived into the Tarim Basin? Depending on the relative chronology of settlement of Tarim vs. Assam, it seems possible that perhaps Tocharian was technically never the easternmost IE branch.

  34. David Marjanović: So on the basis of the genetics and the linguistics (and do not be modest, you may be the best-qualified person on the planet to evaluate both the genetic and the linguistic evidence), what is your take on the history of the spread of Anatolian (Was Proto-Anatolian spoken in Anatolia or not? If not, did it enter from the East or from the West? Why the absence of Steppe ancestry in Anatolia?) on the one hand and of the rest of Indo-European on the other?

    January First-of-May: There is a French scholar (I *might* be able to find the exact reference) who has argued that Common Tocharian must have been spoken to the northwest of the location of both Tocharian A and Tocharian B, so your point about Tocharian having possibly never been the Easternmost Indo-European language may well be true.

  35. The easternmost IE language (disregarding Russian colonization of Siberia) was probably whatever they spoke at Afanasievo… it being any form of pre-Tocharian may not however pan out, I’ve seen archeologists point out by now e.g that they lacked various domestic animals that do have surviving native names in Tocharian.

    The question I’d like to see more attention paid to re: spread-of-Anatolian is whether there even was a distinct Anatolian group, or if it might be an areal unit of Indo-European-that-didn’t-go-to-the-steppes. E.g. Kloekhorst’s recent proposal that *h₂ was *q is already suggestive instead of a family tree {Luwic, {Hittite, Core IE}}.

  36. Oh, this is going exactly where I hoped!

    David M.: …yes, except… for the complete lack of Steppe ancestry in Bronze Age Anatolia. People buried by Mycenaeans have a bit, people buried by Hittites have none whatsoever.

    It does seem like Proto-Indo-Tocharian was spoken on the steppe, but Proto-Indo-Anatolian south of the Caucasus.

    Ref. Lazaridis et al: The genetic history of the Southern Arc: A bridge between West Asia and Europe, Science (2022).

    Y linked to it in the Son of Yamnaya thread, but for some reason the discussion didn’t really take off. I think I needed more time with the supplementaries than I had. Anyway: short story: The Khvalynsk people were about half and half Steppe and Caucasus (the latter with some traces of even southerner genes). That northward spread is contemporary with a westward spread through Anatolia that was already known.

    How did the Caucasus element reach the steppe. “We” discussed this in 2020, starting with a reference to David Anthony and then speculating about a more southern homeland. Anthony argues that the Caucasus element on the steppe came from a population of settled harvesters that spread north along the western Caspian shore and up into the lower Volga. That population merged with the Samara people after the introduction of livestock. “Our” speculation was that the harvesters brought the language from their point of departure in a predecessor of the Kuro-Araxes culture in modern Azerbaijan or thereabouts.

  37. David Marjanović says

    The sibilants written with š-signs had the Babylonian pronunciation [š] and the Assyrian pronunciation [s]; conversely, sibilants written with s-signs had the Babylonian pronunciation [s] and the Assyrian pronunciation [š].

    Oh, great, like Etruscan.

    Welp, I should find that book somewhere.

    what is your take on the history of the spread of Anatolian

    My current guess is Proto-Anatolian (except see below…) was spoken somewhere around eastern Anatolia, and that Proto-Indo-Anatolian was spoken somewhere reasonably close to that – perhaps in Armenia or so as Gamq’relidze & Ivanov have been saying for forty years, though I haven’t read their works, and some of their many, many, many arguments are weak at best. I have to say, though, that I’m enough of an optimist to take Euphratic seriously. It would be the sister-group of Indo-Tocharian.

    January First-of-May: There is a French scholar (I *might* be able to find the exact reference) who has argued that Common Tocharian must have been spoken to the northwest of the location of both Tocharian A and Tocharian B, so your point about Tocharian having possibly never been the Easternmost Indo-European language may well be true.

    We talked about it in mid-August; it’s in open access here and argues, in a lot of detail, that a Pre-Proto-Tocharian stage was in enough contact with a Pre-Proto-Samoyedic stage that their sound systems converged, and that earlier the same had happened with Pre-Proto-Samoyedic and (Pre?)-Proto-Yeniseian. That’s evidence independent of archeology (and genetics) that Tocharian passed through southern Siberia or thereabouts at some point (and adds to the long-known fact that Tocharian noun morphology is typologically Uralic-like and for the most part not inherited from PIE).

    I’ve seen archeologists point out by now e.g that they lacked various domestic animals that do have surviving native names in Tocharian.

    Oh, that is interesting.

    The question I’d like to see more attention paid to re: spread-of-Anatolian is whether there even was a distinct Anatolian group, or if it might be an areal unit of Indo-European-that-didn’t-go-to-the-steppes.

    Of the many characters where Anatolian has one state and Indo-Tocharian has another (or Indo-Actually-European has another and there’s no data for Tocharian), there’s a bunch for which it seems quite a bit easier to assume that the Anatolian state is an innovation. Obviously they could probably all be areal; but the way to test that is to find more innovations that Hittite shares with Indo-Tocharian but not with Luwic. Otherwise, Kümmel’s paper from last year arguing that we can get away with interpreting [q] as a late western Luwic substitution of [χ] caused by an unknown (Tyrsenian…?) substrate remains in the running. …Oh, now I see you’re in the acknowledgments actually. 🙂

    Anyway, the paper points out that *h₂ was in Cuneiform Luwian.

    External comparison and typology certainly suggest that the most common source of *h₂ was an earlier **q at some point, but that’s another story…

  38. I’ve seen archeologists point out by now (…)

    Or make that “archeologist” singular; but none other than Jim Mallory at his plenary lecture at the Secondary Homelands of IE conference just last year. Alas I do not remember the full details. (Some of those talks at least were supposed to go up on YouTube at some point, but it beats me where one might find them.)

  39. The question I’d like to see more attention paid to re: spread-of-Anatolian is whether there even was a distinct Anatolian group, or if it might be an areal unit of Indo-European-that-didn’t-go-to-the-steppes. E.g. Kloekhorst’s recent proposal that *h₂ was *q is already suggestive instead of a family tree {Luwic, {Hittite, Core IE}}.
    Kloekhorst himself at least still counts Anatolian as a clade and asumes a basic Anatolian – Non-Anatolian split for PIE.

  40. David M.’s thoughts on the “Proto-Indo-Anatolian” Urheimat seem congruent with those in the Lazaridis et al. piece. Which maybe raises some questions of what we mean by “Urheimat” and why we want to know. Even if the language can be traced back another step or two further, it seems like on the Lazaridis-style view what happened with the Yamnayans was the ethnogenesis of the “Indo-Europeans” as we generally think of them – not just as speakers of a particular early language (or dialect continuum) but as a particular population group evolving out of a particular history of genetic admixtures and as the innovators/bearers of a particular culture reconstructed by Dumezil et al. which subsequently left traces (some fainter than others) all the way from Iceland/Ireland to the mouth(s) of the Ganges. If the Anatolian-speakers had a somewhat related language but not any of that other stuff, they’re kind of less interesting unless you have a particular enthusiasm for laryngeals.

    And it’s always good to be reminded that linguistic history often tracks genetic and cultural history, but sometimes it doesn’t.

  41. David Eddyshaw says

    Kloekhorst himself

    The whole book is actually open access:

    https://www.cambridge.org/core/books/indoeuropean-language-family/4B44B5ACF0D3BBA89B9408050F112A52

    Nice chapter by Don Ringe.

  42. David Marjanović says

    If the Anatolian-speakers had a somewhat related language but not any of that other stuff, they’re kind of less interesting unless you have a particular enthusiasm for laryngeals.

    At least some of that stuff is indeed missing from Anatolian. For example, there is a Hittite myth where the stormgod slays a dragon, but that’s a generic dragon-slaying myth no different from the one in Japan; the Indo-Actually-European details found in Beowulf, the Rgveda, that one poem by Lucretius and I forgot where else are not there.

    As usual, no data from Tocharian so far.

  43. Hans: [Link to Alwin Kloekhorst: Anatolian (2022)]

    David E.: The whole book is actually open access: [Link to The Indo-European Language Family: A Phylogenetic Perspective]

    Oh, this looks good. Thanks!

  44. David Marjanović says

    Nice chapter by Don Ringe.

    …well… it touches on all the topics it should and then some, I think, but it doesn’t do any of them justice.

  45. If it did, it would be a book. Or several books.

  46. David Marjanović says

    No. A chapter of about twice the length it has would be fine for present purposes.

    Generally the chapters I’ve looked at are all a bit short…

  47. Update: Rieken will be in Oxford giving a seminar on the Boğazköy-Hattusha tablet and its interpretation next Friday 2 February, 5:00 PM GMT.

    I don’t know how many Hatters will find it convenient to turn up at the old Radcliffe Infirmary, but it’s a hybrid event and the time isn’t terrible for leftpondians. (Advance registration is required, though.)

  48. Trond Engen says

    Oh, interesting! Unfortunately not a convenient time for me, so tlet’s hope they’ll tape it.

  49. If LH readers would like to the new language in transliteration, go to https://www.hethport.uni-wuerzburg.de/TLHdig/corpus.php and type

    KBo 71.145

    into the box labelled “Direct Search”.

    To see a drawing of the tablet itself, with the cuneiform signs, download the newest publication (KBo 71) of Hittite tablet fragment finds at https://www.hethport.uni-wuerzburg.de/kbo/pdfs/Schwemer_KBo_71_1_7.pdf and scroll down to number 145 on page 42. The drawing of the reverse side of the tablet is on page 43.

    The language of the first 9 lines (transliterated mostly in lowercase italic with some Sumerograms in uppercase roman and some Akkadian in uppercase italic) is run-of-the-mill Hittite. The ritual practitioner who speaks the words could be either a man or a woman: Hittite rituals are often performed by a wise woman (ᵐᵘⁿᵘˢŠU.GI, ‘old woman’). Also, I translate ‘god’ below, but the DINGIR could just as well be a goddess as a god. Hittite ritual and festival texts often include conjurations, prayers, and songs in local languages other than Hittite: Luwian, Palaic, Hurrian, and of course Hattic. (In another ritual text, a woman from Kaplawiya called Anna speaks a short conjuration in another unknown language. The conjuration is very short and there is not enough of this language to say anything definite about it at all. Kaplawiya was apparently in what the Hittites called ‘the Lower Land’, the plain of Konya and adjacent piedmont of the Taurus.) A bed of cut branches or leaves is often spread in Hittite rituals for the presentation of ritual offerings. The tick mark after the line number means that the actual number of lines that were on the unbroken tablet is unknown, and the numeration of lines in the transliteration simply begins with the first visible traces of a line on the fragment. Here is a quick translation of the part of the text in Hittite:

    1′  [
    2′  [ … ] but 1 pig … [
    3f′  they slaughter cattle and sheep on the foli[age].
    4-6′  S/he sets aside [some meat] from all (the slaughtered animals), from the both raw and the cooked, for the god.
    6′-8′   To the god for whom s/he breaks bread, and to the god for whom s/he libates from their cups,
    9′   s/he conjures in the language of Kalašma as follows:

    The text in Kalašmaic then follows under the dividing line. It is transliterated in roman type. (It is written in the same Mesopotamian cuneiform syllabary as the rest of the text. For LH readers who are unfamiliar with the transliteration of Hittite and Akkadian: hyphens are used to divided individual transliterations of the syllabic signs that make up a word—hence the many hyphens. The spaces represent the actual spaces dividing words on the tablet.) The final word QA-TI is Akkadian for ‘completed’ and is the standard scribal indication that the preceding text or tablet is complete.

  50. David Marjanović says

    Here’s an account of a presentation on the tablet and the language, with links to all the above and more.

  51. A detailed analysis of the text was just published by Elisabeth Rieken, Ilya Yakubovich and Daniel Schwemer

    Eine neue Sprache im Hethiterreich
    Der Fund der Kalašma-Tafel (Bo 2023/12 = KBo 71.145)

    https://doi.org/10.34780/v669-8p6f (free access)

  52. Thanks for that! Here are a couple of bits:

    Der nicht-hethitische Textteil verwendet keinerlei Determinative oder heterographische Schreibungen (Akkadogramme oder Sumerogramme), enthält allerdings auch keine Götter- oder Ortsnamen, bei denen man Determinative unbedingt erwarten würde. Das passt insofern zum bislang bekannten Gesamtbefund, als hethitische Schreiber in fremdsprachigen Textpassagen heterographische Schreibungen typischerweise ausgesprochen sparsam verwenden[2].
    […]

    Forlanini lokalisiert Kalašma im Nordwesten Anatoliens in der Region der heutigen türkischen Provinz Bolu[5]. Die wichtigste Quelle hierfür stellt eine Passage aus den Annalen Muršilis II. dar, derzufolge der Ḫutupiyanza als Verwalter der Länder Pala und Tumana einen Feldzug unternimmt und durch die Eroberung mehrerer Städte das Land von Kalašma wieder dem hethitischen Reich einverleibt[6]. Hieraus lässt sich auf eine Nachbarschaft von Kalašma und Pala, im Nordwesten des hethitischen Kernlandes, schließen. Dies erfährt eine Bestätigung durch eine Liste von abtrünnigen Ländern, in der Kalašma neben den Ländern Arawanna, Lukka und Petašša genannt wird, also Ländern im Westen und Süden von Ḫattuša[7].

Trackbacks

  1. […] found this news thanks to Language Hat Kalasmaic, a New IE Language. : languagehat.com and Language Log » A new Indo-European language (upenn.edu) They both quoted a report at New […]

Speak Your Mind

*