Talk:Indo-Hittite

Scientific evidence

The article regrettably lacks any arguments or evidence regarding the "Indo-Hittite Hypothesis", however contains a lot of science fiction (see e.g. my entry under Bronze Age below).HJJHolm (talk) 16:02, 24 November 2007 (UTC)

Dbachmann, is there any reason why you move this article back ? I thought the standard for wikipedia was to name articles "XXX language" or "XXX languages". Why should this article be an exception ? bogdan ʤjuʃkə | ^Talk 06:42, 7 Apr 2005 (UTC)

This article isn't exactly about a language or group of languages, it's really about a hypothesis within historical lniguistics. Naming it Indo-Hittite languages would imply that the content of the article should duplicate Indo-European languages, which of course it doesn't. But it might make sense to rename this Indo-Hittite hypothesis. --Angr/_comhrá 23:29, 24 May 2005 (UTC)

Of course, you do realize that even the Indo-European language group is a hypothesis, right ? They have different degrees of acceptance in the scientific world, but since none was directly attested, they are only hypotheses. bogdan ʤjuʃkə | ^Talk 13:56, 25 May 2005 (UTC)

Is Proto-Indo-European a hypothesis or a theory? Probably a hypothesis, because theories require more evidence. Decius 14:30, 25 May 2005 (UTC)

The hypothesis is : Proto-Indo-Hittite language, later branching into Anatolian languages and Proto-Indo-European language (<from which all Indo-European languages except the Anatolian languages come in this scenario). Decius 14:41, 25 May 2005 (UTC)

So naming it Proto-Indo-Hittite language is a possibility to be voted on if many people are interested. Decius 15:11, 25 May 2005 (UTC)

Some points:

1. Proto-Indo-European is definitely a theory, no longer a hypothesis, because of the overwhelming evidence that the Indo-European languages are descended from a common ancestor.
2. I think the point of this article should be to discuss the Indo-Hittite hypothesis (who proposed it, what the evidence in favor of it is, who argues against it, what the evidence against it is). Naming it Indo-Hittite hypothesis would make that intention clear. Renaming it Indo-Hittite language or Proto-Indo-Hittite language would imply that the article should focus on reconstructing the proto-language, and would largely duplicate Proto-Indo-European language.
3. It's customary to use "Proto-" in the names of proto-languages, if only to disambiguate between the name of the proto-language and a cover term for its daughter languages. Irish is a Celtic language, but it isn't the Proto-Celtic language.

--Angr/탉 07:03, 26 May 2005 (UTC)

I think the theory that all the Indo-European languages (including the Anatolian languages) descend from a single Proto-Indo-European language should be a hypothesis, simply because there is not enough evidence that the Anatolian languages are included in the same way. Decius 04:09, 27 May 2005 (UTC)

Though I agree with point 3, the Celtic/Proto-Celtic example is not parallel: in Indo-Hittite, 'Indo' stands for "Proto-Indo-European" and 'Hittite' stands for "Anatolian": so "Indo-Hittite language" could as well refer to the time when they were still one, before they split: so "Indo-Hittite language" could convey the same information as "Proto-Indo-Hittite language". That's true what you're saying though, and the convention is to use "Proto-" for the early language and to drop "proto-" when referring to the group. And I guess that such classification qualifies as a logical reason. Decius 04:41, 27 May 2005 (UTC)

Though I'm not necessarily advocating it be titled Proto-Indo-Hittite language, such a title would not violate accuracy: see Proto-Pontic language for a parallel example. The Proto-Indo-Hittite language hypothesis is about on the same level as the hypothesis that the Proto-Indo-European language includes the Anatolian languages. Decius 04:54, 27 May 2005 (UTC)

So you think there's equally good evidence in both directions, so both should be called hypotheses rather than theories? I'd say there's an Indo-European theory, which includes two sub-hypotheses: (1) Anatolian broke off from the others only a little earlier (say a couple of centuries) than the other broke up (Anatolian-as-daughter hypothesis), (2) Anatolian broke off from the others a lot earlier (say a couple of millennia), allowing the rest to develop independently before breaking up (Indo-Hittite hypothesis). --Angr/_{tɔk tə me} 05:33, 27 May 2005 (UTC)

That's good to. The reason I'm interested in this debate is because I feel that the Proto-Indo-Hittite hypothesis may well be the correct one, because the other scenario is trying to squeeze (from a preconceived notion) everything into one late proto-language through a short window of time. Probably they underestimate how much time it takes for such languages to develop. Decius 05:58, 28 May 2005 (UTC)

Bronze Age and Chalcolithic

<rant>An anon just added a paragraph about the question whether Anatolian broke off from the rest of PIE before the Bronze Age or even the Chalcolithic. The question that interests me is, why are people so intent to tie linguistics in with archaeology? The Indo-Hittite hypothesis is a linguistic question that can only be answered with linguistic data. If the archaeological evidence supports the hypothesis, that's great, but it doesn't strengthen the hypothesis. If the archaeological evidence does not support the hypothesis, that's okay too, because it doesn't weaken the hypothesis either. The same holds true for a lot of other questions in Indo-European linguistics. </rant> --Angr/_{tɔk tə mi} 20:27, 22 September 2005 (UTC)

There's been a bit of intermingling of historical linguistics with genetics and archaeology. Part of it is the link between language and culture. Interdisciplinary research into such matters can, on occasion, strengthen or weaken a hypothesis. AEuSoes1 06:19, 27 March 2006 (UTC)

There isn't only " a bit"...; BTW, the EBA (Early Bronze Age) starts in Anatolia at least with the beginning of the 3rd Millenium cal. BC., and NOT 3.300 in the Caucasus. See any modern archeological journal.HJJHolm (talk) 16:03, 24 November 2007 (UTC)

The EBA begins ca. 3000 BC. It is futile to debate when precisely it should be taken to begin. Christopher Edens, Transcaucasia at the End of the Early Bronze Age, Bulletin of the American Schools of Oriental Research (1995):

Transcaucasia had considerable impact on eastern Anatolian during the Early Bronze Age when the Transcaucasian style of black and red burnished pottery, architecture, and other artifacts appeared in eastern Anatolia and then in Syro-Palestine. The movement of this culture, variously called Kura-Araxes, Early Transcaucasian, and Khirbet Kerak, began late in the fourth millennium B.C., and its effects persisted into the second half of the third millennium. The incursion of the Kura-Araxes culture in the EB I coincided with the collapse of the Uruk-related trading system that had encouraged the emergence of towns and political complexity in eastern Anatolia.

in other words, EBI begins gradually, from about 3300, with the Kura-Araxes culture, brings about the end of the Uruk period in the 31st century, and is in full swing by the turn of the millennium. dab (𒁳) 10:23, 5 February 2008 (UTC)

Animate / inanimate distinction

"Notably, Anatolian doesn't have the IE gender system opposing masucline : feminine; instead we have a rudimentary noun class system based on an older animate : inanimate opposition reminiscent of noun class systems in non-Bantu Niger-Congo languages."

Well, the Russian language also has a animate/inanimate distinction in its nouns as well. No need to recur to Niger-Congo languages. —Preceding unsigned comment added by 82.241.221.24 (talk) 11:36, 9 October 2008 (UTC)

Russian has a masculine-feminine-neuter distinction in the area that Hittite shows an animate-inanimate distinction, so it doesn't really compare. Dutch is closer since its has merged the masculine-feminine of its old masculine-feminine-neuter system leaving two cases kind of like an animate-inanimate system. It probably still isn't a good comparison though.Ekwos (talk) 01:32, 16 December 2008 (UTC)

Before someone adds the Dutch comparison to the article, let me point out that it's quite different from the Hittite system (as I understand it). Dutch has largely merged masculine and feminine, but that gives two genders common and neuter, without much meaning attached to the genders: e.g. stoel ("chair"), is common, but all diminutives including those referring to people are neuter. QVVERTYVS (hm?) 16:24, 13 December 2013 (UTC)

good god the jargon

I'm sure undergraduate students of historical linguistics understand what "(SLR-D)" means, but we laymen don't.

Obviously one doesn't need to explain every technical term in detail on this particular page, but links to detailed information are necessary to make the article better than pseudo-academic wankery. 71.248.115.187 (talk) 02:47, 23 August 2009 (UTC)

Actually, SLR-D was described as "a method" in the same sentence and supplied with a reference. This is all laymen have to know without compromising WP:VERIFY.Rokus01 (talk) 13:07, 23 August 2009 (UTC)

Presentation of the matter seems inadequate

I don't think that the problem is quite appropriately explained here. For example, a tendency towards simplification of the noun case system within the history of Hittite (or other individual languages) is irrelevant; it's the Old Hittite stage that counts, if anything. (It is true, though, that in the plural, many case distinctions found in other branches seem to be attested nowhere in Anatolian and the overall case system in the plural looks much simpler, comparable to the dual.) Also, I fail to see how the vowel system in Anatolian in general is simpler than in PIE. It's rather the verbal (tense-aspect-mood) system of Hittite especially, which is unexpectedly simple compared to the traditionally reconstructed Brugmannian or Graeco-Aryan system, and more comparable to Germanic, lacking several categories usually reconstructed for PIE, which has given rise to this proposal. To be fair, however, the noun system is also simpler, considering the lack of the feminine and the dual, and the fewer case distinctions in the plural mentioned above. On the other hand, an additional case, the directive, has been reconstructed for Anatolian.

As far as I understand the issue, perhaps the decisive part of the problem is the overemphasis on Hittite, which is already reflected in the name of the proposal, "Indo-Hittite" instead of "Indo-Anatolian" or so. Yes, Hittite is strange in some ways, especially compared to other highly ancient (and presumably archaic) IE languages. But Hittite is only one out of several Anatolian languages; what should really matter is reconstructed Proto-Anatolian. As Hittite and its Anatolian sister languages were investigated more closely in recent decades, traces of categories which were previously considered absent in Anatolian turned up, or were directly found present in non-Hittite Anatolian languages. Therefore, reconstructed Proto-Anatolian turned out considerably less "strange" for an IE language than Hittite is. I can definitely affirm this for the phonology; the PA system as reconstructed by Melchert is surprisingly close to the PIE system (though still distinctive in some important points), especially the vocalism. Hittite and Luwian made it seem (at least superficially) as if *o had merged with *a, but Lycian hasn't. Therefore, even for Proto-Luwian, and a fortiori PA, a conventional five-vowel system has to be reconstructed.

The conclusion is that the "Indo-Hittite" proposal is simply misguided and Anatolian is essentially just another branch of IE on the same level as Indo-Iranian or Tocharian.

That doesn't mean, however, that everyone agrees that there is nothing special about Anatolian at all. Many scholars in the field do seem to think that Anatolian split from PIE first – with a crucial difference. The Indo-Hittite proposal implied a binary branching Proto-IE (sans Anatolian) vs. Proto-Anatolian; under the alternative view, there was never a "non-Anatolian" node (i. e., protolanguage). Instead, Anatolian separated from an already dialectally differentiated IE, a dialect continuum to be precise, not from PIE. Within the "non-Anatolian" continuum, some innovations could still spread, failing to reach Anatolian, which was already geographically detached and had lost contact with the remainder of IE, unable to exchange innovations with them. However, those innovations did not necessarily reach all non-Anatolian branches, either. Especially Greek and Indo-Iranian are often assumed to have shared innovations at a relatively late stage (but still before the Proto-Indo-Iranian stage). This idea is usually expressed as "Late PIE". (One innovation that is often attributed to "non-Anatolian" is the development of the perfect, whose counterpart in Anatolian, the hi-conjugation, does not seem to have a truly perfective meaning and may therefore represent an older stage. This is connected with the Urmedium proposal championed by Jasanoff, attempting to derive perfect and mediopassive from a single verbal category.)

I'm not sure how prevalent that view is, but it does seem to be relatively common in the field. There is also the idea, though perhaps somewhat less popular, that Tocharian was the second branch to split off and lose contact with the rest of IE. Italic and Celtic may have come next. One reason for this idea seems to be the mediopassive endings in *-r shared by all those branches (i. e., Anatolian, Tocharian, Italic and Celtic), which may therefore be an archaism that was replaced by the familiar endings in *-i in the other branches (however, Phrygian seems to have mediopassive endings in *-r as well, as in addaketor, vs. addaket, presumably cognate to Latin facitur and facit respectively with a preverb or perhaps augment). --Florian Blaschke (talk) 17:44, 1 March 2011 (UTC)

I don't think that the "bronze" argument is particularly strong, either, I'd even consider it inconclusive and obsolete. There is a word *aios reconstructed for PIE that must mean a metal, but in the attested languages it can mean "ore" and various metals and alloys: "bronze" as well as "copper" (as in Latin aes) and "brass" and (in Indo-Iranian) even "iron". This is similar to the state of affairs in Proto-Uralic, where we have only *waśke as a term for an unspecified type of metal. In both cases it is suspected that it was "copper", the prime working metal in the latest stages of the Neolithic. There are old words for "gold" and "silver", quite possibly of PIE date, as well, which would definitely exclude those metals. If we have only one term for "(working) metal" (not gold or silver) in a prehistoric European language, the reasoning goes, this does point towards copper, but not necessarily to bronze. --Florian Blaschke (talk) 03:30, 14 April 2011 (UTC)

@Florian Blaschke: The term Indo-Hittite was coined before the decipherment of the other Anatolian languages. --Tom 144 (talk) 16:54, 25 February 2018 (UTC)

Lack of innovations

From the point of view of phylogenetics, the Indo-Hittite hypothesis is strongly corroborated by the lack of any Indo-European innovations in Anatolian. Notably, Anatolian doesn't have the IE gender system opposing masculine : feminine; instead we have a rudimentary noun class system based on an older animate : inanimate opposition reminiscent of noun class systems in non-Bantu Niger-Congo languages.

The phrase "Indo-European innovations" makes no sense at all. Innovations of Indo-European as opposed to what else? We don't know any language or language family for certain to be related to Indo-European. Alternatively, "Indo-European" here could mean "non-Anatolian", but in that case, the choice of wording presupposes the Indo-Hittite hypothesis and the argument is thus circular. Moreover, the lack of some (outside of Anatolian) innovation in Anatolian as such does not prove anything, unless that innovation is common to all branches except Anatolian, and moreover cannot have spread via diffusion within an already differentiated dialect continuum (admittedly, the possibility of diffusion is difficult to exclude, though the spread of some instance of deep grammatical restructuring would be difficult to argue for – the feminine, however, is not sufficient in my opinion, especially given apparent traces of the older system outside Anatolian, too). In his Etymological Dictionary of the Hittite Inherited Lexicon (2007), chapter 0.6, pp. 22–26, Alwin Kloekhorst lists seven possible common innovations of non-Anatolian Indo-European, which would seem to validate the "weak" version of the hypothesis I've already explained above, but since that one is widely accepted, it says nothing for the Indo-Hittite hypothesis proper. --Florian Blaschke (talk) 21:08, 26 May 2013 (UTC)

Worse, as pointed out in Talk:Anatolian languages#Masculine/feminine noun class system ("gender") as an example of Hittite archaism., in this case the old conundrum "archaism or loss?" is especially unclear, and Anatolian may simply present a case analogous to various Germanic languages which have merged masculine and feminine genders into a common gender. --Florian Blaschke (talk) 21:37, 26 May 2013 (UTC)

The paragraph you removed was perfectly fine, as I think dawned on you as you wrote the above paragraph, you just found "Indo-European" used in the sense of "non-Anatolian" confusing. This could just have been clarified if indeed it was misleading. It is fair enough to say this article has too much unreferenced hand-waving written by Wikipedians instead of a referenced summary of academic opinion, but you are not helping the situation if you selectively remove the hand-waving that happens to go against your preferred opinion while leaving untouched the hand-waving you feel sympathetic towards. This way, all you achieve is turning a fair but poorly referenced page into a biased and poorly referenced page. --dab (𒁳) 10:33, 5 September 2015 (UTC)

@Dbachmann: Well, I disagree with "perfectly fine", and the paragraph doesn't sound particularly NPOV nor valuable in general to me (the circularity still grates), but if you will, restore the paragraph; I won't resist.

@Mellsworthy: Considering that you're guardedly supportive of Indo-Hittite, your third opinion would be very welcome. --Florian Blaschke (talk) 05:51, 25 July 2016 (UTC)

@Florian Blaschke: I shouldn't be construed as supportive of Indo-Hittite in all its glory(?), but I do agree with User:Dbachmann that, in the context of the Indo-Hittite hypothesis, Indo-European really does refer to what some other non-Indo-Hittite-subscribing Indo-Europeanists call "nuclear"-PIE or "core"-PIE, vz., PIE without Hittite (+/- Tocharian, depending on who's talking). I don't see any problem with the paragraph in this context. It doesn't count as NPOV to use a theory's terminology. A page on Freud's psychoanalytics would be remiss to skip past the id, superego, and the ego, even though the last of these terms is used vastly differently in other psychological theories. This article should make clear somewhere, but probably only in one place, (I haven't checked whether it does) that a majority of historical linguists working on this sort of problem have chosen another terminology for what is more-or-less the same thing, other than emotional coloring and the emphasis on "basically the same" or "we can't reliably reconstruct anything really different" (standard view) vs. "fundamentally different and clearly separately reconstructable" (Indo-Hittite view). I hope we can all agree on what an article like this one should do: This article should help people out who want to read and understand statements and articles about Indo-Hittite, and not be confused by this article perversely avoiding the Indo-Hittite theory's terminology, and also not be confused by this article ignoring the relationship to the majority terminology. That's my two cents, anyway. --Mellsworthy (talk) 08:57, 25 July 2016 (UTC)

@Mellsworthy: It's not just the terminology I have a problem with, or find jarring. The first sentence is simply circular, and thus redundant, the way it is phrased, even if you replace "Indo-European" with "non-Anatolian" (which actually makes this clearer, I think).

Try to look at it from the point of view of a person who isn't already convinced of the existence of non-Anatolian innovations: then the assertion clearly begs the question, IMHO. It is not NPOV; it is clearly partisan (yes, dab, I think the paragraph in this form makes the article biased to begin with), but written in Wikipedia's voice. Moreover, the fact that the crucial question is if the apparent non-Anatolian innovations are really innovations at all appears to be suppressed here. And even if they are common non-Anatolian innovations, there are still those pesky complications I keep talking about that make it so hard for a slam-dunk for the Indo-Hittite camp (though admittedly for their opponents too). Which is, of course, why even you don't believe in the classic IH hypothesis, only a weak variation thereof. (I think the two main weak variations are: 1) "Non-Anatolian IE" is basically a sprachbund, with diffusion of innovations across already existing dialect boundaries, not strictly a node; 2) "Non-Anatolian IE" is a node, but the differences from "IE including Anatolian" are at least not radical. I have defended 1, Melchert and you appear to be defending 2.)

Even apart from that, I just think the paragraph isn't well written and doesn't add much to the article, if it is at all necessary. Like, is it really necessary to adduce Niger-Congo when there are much more familiar languages with a common gender such as Dutch or Swedish?

However, I'm not going to edit-war about the paragraph. I'm just not comfortable restoring it and I find that my opinion on the paragraph hasn't changed; I don't like it any better than I did at the time. But maybe I'm just too biased and can't think clearly. --Florian Blaschke (talk) 23:53, 27 July 2016 (UTC)

@Florian Blaschke: @Dbachmann: A potentially agreeable change to the wording:

From the point of view of phylogenetics, the Indo-Hittite hypothesis is strongly corroborated by the lack of some apparently Indo-European-specific innovations in Anatolian. Notably, Anatolian did not have the IE gender system opposing masculine : feminine : neuter; instead Anatolian had a rudimentary noun class system based on an older animate : inanimate opposition ; the elaboration of such a simple system is reminiscent of the elaboration of the two way noun class systems in non-Bantu Niger-Congo languages into the much more elaborated system of Bantu.

I think that addresses the circularity problem and some of the objectivity problems, and hopefully seems reasonable to all parties. It seems that you should probably also compare to the situation of Scandinavian and Dutch, per Florian's suggestion, perhaps in a following sentence such as this:

A competing explanation for the mismatch in the gender systems of Anatolian vs. most other Indo-European languages is a simple merger, on the Anatolian side, of the masculine and feminine opposition usually reconstructed for the protolanguage by most Indo-Europeanists; this too is plausible, as such a merger has occurred in modern Scandinavian languages and Dutch, among others. Evidence for this merger has been proposed by some Anatolian researchers, but it seems the communis opinio is shifting to disfavor the merger hypothesis again. (cite Melchert)

In general, I agree with Florian that this article needs a lot of work, but I do not have the time to sink into it right now. Dissertation! --Mellsworthy (talk) 00:52, 28 July 2016 (UTC)

That sounds a lot better. Go for it! --Florian Blaschke (talk) 01:04, 28 July 2016 (UTC)

@Dbachmann: What do you think about the wording I have above? Do you think it's reasonable to put the counter-arguments together in a separate section? --Mellsworthy (talk) 00:54, 30 July 2016 (UTC)

Kroonen (2018)

Kroonen (2018), Linguistic supplement to Damgaard et al. 2018: Early Indo-European languages, Anatolian, Tocharian and Indo-Iranian argues in favor of the Indo-Anatolian hypothesis:

Our findings corroborate the Indo-Anatolian Hypothesis, which claims that Anatolian Indo-European split off from Proto-Indo-European first and that Anatolian Indo-European represents a sister rather than a daughter language. Our findings call for the identification of the speakers of Proto-Indo-Anatolian as a population earlier that the Yamnaya and late Maykop culture.

Joshua Jonathan -Let's talk! 04:03, 16 May 2018 (UTC)

I still find it suspicious that (per Kümmel 2015, as already admitted by Melchert) there are still no uncontested examples of non-Anatolian shared innovations in the phonological realm. This is a major weakness haunting the hypothesis.

As for Damgaard et al. (2018), a problem I've seen mentioned on the Eurogenes blog is that there is apparently no confirmation that the samples used came from actual speakers of Hittite or other Indo-European Anatolian languages, and in fact the study does not claim this. They could equally have been Hattians or Hurrians, in which case the lack of steppe ancestry would be much less surprising. This illustrates nicely the pitfalls of archaeogenetic research, and why Wikipedia should not rely on single studies. --Florian Blaschke (talk) 01:58, 24 August 2018 (UTC)

Holm

I have removed the section on Holm's investigations. His research does not appear to notable and a dissenting voice who has been paid little notice and whose criticism has not been taken up by others in the field does not generally deserve an entire section in an article. Tristanjlroberts (talk) 20:53, 6 October 2023 (UTC)

The anonymous "Tristanjlroberts" descibes himself as a Theoretical Pphysicist and therefore is not in a position to criticise people who have by far more experience in linguistics, archaeology, Bayesian glottochronology, or genetics. This is an extremely difficult and still debated field. Even the paper of Jäntti (2016) cannot not really enlighten us. Unfortunately, the honourable outing of his mental shortcomings does not change this.HJJHolm (talk) 15:27, 11 November 2023 (UTC)