Reconstruction talk:Proto-Indo-European/-tr̥

This accentuation

@Rua, Tom 144, so this suffix and *h₁ítr̥ would be fine in the compositional model, but not in either the Leiden or Erlangen model. I"m inclined to move *h₁ítr̥ to *h₁éytr̥. Thoughts? —*i̯óh₁nC^[5] 03:32, 2 February 2018 (UTC)

Well we do work here with the Erlangen model. The full grade has no support though. The Erlangen equivalent would probably be a proterokinetic, since Hittite's reflex cannot come from *-tér. --Tom 144 (𒄩𒇻𒅗𒀸) 03:36, 2 February 2018 (UTC)

@Tom 144: sorry, yeah protero-. —*i̯óh₁nC^[5] 03:46, 2 February 2018 (UTC)

The full grade would (as with all Erlangen proterokinetics) be accounted for with a leveling from the weak grades. As I've mentioned before, the Erlangen model is probably wrong about proterokinesis, but consistency is a thing. —*i̯óh₁nC^[5] 03:49, 2 February 2018 (UTC)

What am I saying? I meant amphikinetic, which in the neuter go *R(é)-S(Ø)-Ø ~ *R(Ø)-S(Ø)-és (see *déh₂nu). Regardless the point about leveling stands. —*i̯óh₁nC^[5] 03:53, 2 February 2018 (UTC)

@JohnC5: Is the morphology here based solely on Ringe? Is there a policy that says that we cannot contradict his views? I don't mean it for the Erlangen model. This might sound a little off topic, but it really bugs me the fact that we are right in the middle on the bh/m issue. I mean, there is now language that shows both morphemes. It's a kind of absurd solution for such a difficult problem. Choosing arbitrarily one morpheme wouldn't be so bad, but choosing both?! And assume that both branches independently made the exact opposite innovation. It cannot be taken seriously. Jasanoff has an elegant solution, which doesn't deliberately ignore the Anatolian evidence, and does not mess up PIE's subgrouping. I think it would be great we we could consider his proposal. --Tom 144 (𒄩𒇻𒅗𒀸) 04:39, 2 February 2018 (UTC)

@Tom 144: So, a few things:

• The unofficial policy is named @Rua (no offense). She is the primary force behind the adoption of the Erlangen model, though it also corresponds to a dearth of belief on her and my part in the Leiden model. The compositional model is comparatively new, not fully worked out, and has little to no infrastructure on this project. If we added modules that could do it, then I'd be fine adding redirects from the appropriate compositional reconstructions.
• We also do not deal with the Anatolian issue. We are mostly reconstructing Proto-Nuclear-Indo-European because we have a feminine gender, a dual, the broad use of aspect stems, the Caland system, and a few other things, none of which are reconstructable for the Anatolian branch. If we wanted to really fix this, we'd probably move every page currently labelled as PIE to PNIE and then remake the appropriate PIE pages where attestable.
• The Jasanoff solution works fairly well. Again, you'd have to convince someone (namely Rua), to reprogram the declension (and maybe conjugation) modules to implement it.

The truth is that we are fairly out of date in regards to Anatolian and that sub-grouping. A such, many of our inflectional modules and “PIE” reconstructions are actively wrong and should be PNIE. We might need to bring this up somewhere else, but at the moment, the only way to fix this would be to reprogram a bunch of stuff and get bots to move all of the pages. @Mahagaja, AryamanA, Chuck Entz —*i̯óh₁nC^[5] 05:48, 2 February 2018 (UTC)

(Why was I pinged? I have only a very basic knowledge of PIE. —Aryaman^{A (मुझसे बात करें • योगदान)} 17:56, 2 February 2018 (UTC))

I would certainly support something like this. Other solution could be to adopt the Proto-Indo-Hittite terminology if redirecting every thing is too tedious. Although from previous experience with Rua I'm dubious we'll ever reach a consensus. --Tom 144 (𒄩𒇻𒅗𒀸) 06:03, 2 February 2018 (UTC)

I am not a fan of Proto-Indo-Hittite as a name, personally. —*i̯óh₁nC^[5] 06:16, 2 February 2018 (UTC)

I based my work primarily on Ringe's work, which is the only source I had access to for a while. I have no problem including the Leiden model inflections in entries, but only as long as it's made clear that it's not really the mainstream. We should probably also add "Erlangen model" to the existing inflection templates, although I'm not sure where that would and wouldn't be appropriate. The term "Erlangen model" is still pretty new to me, and I don't really know what it implies. As for modules, I would appreciate help from others as I'm not all that familiar with all the details of the Leiden model, in particular what differs and what doesn't. —Rua (mew) 11:48, 2 February 2018 (UTC)

Leiden modelists do not tend to agree in all terms. I think the bases if the Leiden model was set by Beekes, and now the principal adherents are Kloekhorst and Kortlandt, (Pooth maybe?). They have innovated the model a little. Their ideas are a little crazy but interesting nonetheless. Here is how I generally equate the different models:

• Hysterokinetic= hysterodynamic 1 (e.g páh₂tr̥, ph₂térm, ph₂trés)
• Amphininetic = hysterodynamic 2 (e.g ǵʰémon, ǵʰménm̥, ǵʰmnés)
• Acrostratic = Static (e.g mah₂tr̥, mah₂tr̥m, mah₂tr̥s)
• Acrostratic o/e = hysterodynamic 3 (e.g nékʷts, nokʷéts, nokʷtés)
• Acrostratic ē/e = Static?
• Kinetic root noun = Dynamic 1 (the nomenclature is the only difference)
• Static root noun o/e = dynamic 2 (e.g. pḗts, pédm̥, podés)

About patterns:

1. Hysterodynamic = nom.sg= (é)-(Ø)-(Ø), acc.sg= (Ø)-(é)-(Ø), gen.sg= (Ø)-(Ø)-(é)
2. Hysterodynamic = nom.sg= (é)-(o)-(Ø), acc.sg= (Ø)-(é)-(Ø), gen.sg= (Ø)-(Ø)-(é)
3. Hysterodynamic = nom.sg= (é)-(Ø)-(Ø), acc.sg= (o)-(é)-(Ø), gen.sg= (Ø)-(Ø)-(é)
4. Static= nom.sg= (é)-(Ø)-(Ø), acc.sg= (é)-(Ø)-(Ø), gen.sg= (é)-(Ø)-(Ø)
5. Dynamic = nom.sg= (é)-(Ø), acc.sg= (é)-(Ø), gen.sg= (Ø)-(é)
6. Dynamic = nom.sg= (é)-(Ø), acc.sg= (é)-(Ø), gen.sg= (o)-(é)

In anything else they agree. I personally think that they suffer from confirmation bias, since they reconstruct things so confidently with so little support. The paradigmatic model suited the Vedic evidence quite elegantly, although we cannot say the same concerning Anatolian. But even though the paradigmatic model isn't complete yet, adding it whenever possible would be great. @Rua: What do we do concerning the anachronistic morphology? --Tom 144 (𒄩𒇻𒅗𒀸) 14:34, 2 February 2018 (UTC)