Talk:Proto-Indo-European language

English as a descendant

The fact that English is a descendant of PIE does not establish the notability of PIE, since PIE would be notable even if English were not Indo-European. PIE's notability rests on the fact that it's the proto-language on which reconstruction was done first and reconstruction has been done most thoroughly. PIE's notability is basically beyond question, and tacking an utter irrelevance like the fact that English is one of its descendants does nothing to improve the lead. We might as well say "Divehi is one of the modern descendants of this language". It's equally true, and establishes PIE's notability just as well (i.e. not at all). —Angr 16:45, 12 November 2008 (UTC)

P.S. I'm very sorry to read that you think refraining from edit-warring to include material that has not found consensus somehow makes you "the bigger man". Actually, it's merely doing nothing more than is expected of you. —Angr 16:52, 12 November 2008 (UTC)

Thank you for your classy reply. I hope I'm wrong, but I can't help thinking that I hurt your feelings with my decision not to edit-war with you.

Now Angr, the subject of the article is a language, and it happens to be the ancestor of the language in which the article, as the encyclopedia, is written. That's notable information. This is not, unfortunately, the Divehi Wikipedia. But if the article ever becomes detailed enough, perhaps Divehi can be mentioned in a section. And again, Divehi was not an ancestor of English, last time I checked. So English is in a rather unique position here.

Just in case that you didn't read the MOS section I linked to in an edit summary, allow me to quote some of it:

• "The article should begin with a straightforward, declarative sentence that, as briefly as possible, provides the reader who knows nothing at all about the article's subject with the answer to two questions: "What (or who) is it?" and "Why is this subject notable?".

It goes on:

• "If the subject of the first sentence is amenable to definition then the first sentence should give a concise definition that puts the article in context and is as clear to the nonspecialist as the subject matter allows."

And surely the notability of PIE rests on more than mere chance! It is the ancestor of some of the world's most spoken languages, and in recent centuries, the dominant languages, English among them.

Goodbye, Angr. SamEV (talk) 17:53, 12 November 2008 (UTC)

The topic is relevant enough already without an overt reference to English in the lead section and making one may be too POV. It's not the case that, if English were not an Indo-European language, the article would not be relevant enough for Wikipedia.

I notice that the lead section links to Indo-European languages, which says that three billion people speak one IE language or other. If we mention that in the lead section here, that might highlight the notability a bit more clearly. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 19:52, 12 November 2008 (UTC)

What "decision not to edit-war"? You started edit-warring as soon as you reverted Aeusoes. But as Aeusoes points out, and as I pointed out, your addition does not bring the article closer to meeting the MOS requirements you quote, because the fact that English is derived from PIE is not what makes PIE notable. Proto-languages that English isn't derived from, like Proto-Afro-Asiatic and Proto-Algonquian, are equally notable. Also, the fact that this is the English Wikipedia doesn't give English any special status here. No matter what language we're writing in, English is not a more significant descendant of PIE than any of its hundreds of other descendants. —Angr 21:58, 12 November 2008 (UTC)

Aeusoes, and Angr, that's not how it works. A subject is notable before it gets an article. But its notableness doesn't mean that everyone is familiar with or has even heard of the subject, of course. Thus the guidelines recommend that this notability be explained to the non-specialist. It's not at all a case of trying to make the subject notable, or of "establishing" or increasing its notability. The quotes from the guidelines are clear enough, really.

Mention of the 3 billion speakers is in the same class as my edit: it helps explain the notabilibity/importance of the subject. So I'm for it.

Angr, I restored my edit twice. Not much of an edit war, now is it? SamEV (talk) 00:40, 13 November 2008 (UTC)

I'm not sure what it is about my or Angr's comments above that imply that we think that a subject becomes notable after it gets an article. I agree that explaining notability is important and I don't think that you're trying to do anything but explain notability. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 17:11, 13 November 2008 (UTC)

I'm not saying notability doesn't have to be shown; I'm saying that a sentence saying "English is descended from PIE" does not actually succeed in showing the notability of PIE. I accept that your edit was a good-faith attempt to improve the lead by showing the notability of the topic, but I believe the attempt failed, because the edit does not in fact show the notability of the topic. —Angr 17:24, 13 November 2008 (UTC)

Ah, so the backtracking has begun. Both of you made comments that clearly accuse me of trying to demonstrate the subject's notability as somehow derived from its being an ancestor of English. Really, they're blatant. Angr wrote: "The fact that English is a descendant of PIE does not establish the notability of PIE, since PIE would be notable even if English were not Indo-European." And "your addition does not bring the article closer to meeting the MOS requirements you quote, because the fact that English is derived from PIE is not what makes PIE notable." Both comments are on this page.

Aeusoes, in turn, wrote this on my talk page:[1] "Hey, I didn't want you to walk away feeling scorned over the edit conflict at Proto-Indo-European language. What Angr and I were trying to get across is that the topic is relevant enough already without an overt reference to English in the lead section and that making one may be too POV. Your reasoning implies that, if English were not an Indo-European language, the article would not be relevant enough for Wikipedia (which is false, of course)."

Those comments speak for themselves. I'm out. SamEV (talk) 02:38, 14 November 2008 (UTC)

Hmm, you're doing a pretty bad job at conversing in a civil manner. You're either stubbornly grasping at straws, trying to sow discord, or have really poor reading comprehension. This whole "gotcha" thing isn't working for you. Neither is disengenuously claiming that you're walking away and then coming back. Let's focus on the article itself, eh?

Angr, what do you think of my suggestion above? — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 08:34, 14 November 2008 (UTC)

Sam writes, "Both of you made comments that clearly accuse me of trying to demonstrate the subject's notability as somehow derived from its being an ancestor of English." Gee, maybe that's because you explicitly said that's what you were doing with this edit summary, this edit summary, this comment, and this comment. I'm not seeing any backtracking on the part of Aeusoes or myself, though. Aeusoes, your suggestion is okay, I guess, but still PIE would be no less notable if its descendants were spoken by only three hundred people rather than three billion. —Angr 09:02, 14 November 2008 (UTC)

While that last point is quite correct, part of the notability of PIE within linguistics are that (a) it is a construction based on a very large number of securely attested languages (b) it acted historically as a proof of concept for the comparative method. --Pfold (talk) 11:12, 14 November 2008 (UTC)

Yes, those two facts establish PIE's notability much more than the identities of its daughter languages or the number of people speaking them. —Angr 11:18, 14 November 2008 (UTC)

(outdent) Don't look now, but you guys just contradicted yourselves massively; yet again.

And the projecting by you two continues. Reading comprehension? Everything you quoted me saying supports what I've said: I was endeavoring to convey/explain/contextualize the notability of the subject to those who may not know about it, just as MOS says we should do.

Aeusoes, I did walk away; where have you been? Then someone (guess who) started this thread. Why? Had I not walked away? But I replied. Each time you guys chose to reply to that. And so on. You're just as "guilty" of discussing as I am. But so long as you keep replying to me, and trying to twist things in the process, I'll reply. How's that? As for civility: Look who's talking.

Angr, it is just amazing that you actually think that this comment of mine says what you've just claimed it does. In it I clearly explain that notability precedes article creation, and that what MOS asks is that we explain it to the non-specialist reader.

You two are embarrassing yourselves. Please ask someone who doesn't pull any punches and whose advice you trust. You'll see.

Yeah, yeah. You'll predictably come back by suggesting I do the same. Maybe I will. Unlike you, I have nothing to fear from the exercise. SamEV (talk) 18:26, 14 November 2008 (UTC)

Just stick to discussing the article, kid. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 10:33, 15 November 2008 (UTC)

With your tactics you two spoiled it for me. I lost interest for now. SamEV (talk) 19:05, 15 November 2008 (UTC)

Goodbye. —teb728 t c 19:28, 15 November 2008 (UTC)

I still disagree and believe that Angr and "Ƶ§œš¹" are being unreasonable. —Preceding unsigned comment added by 174.17.244.68 (talk) 07:11, 10 August 2009 (UTC)

Capitalisation of sound laws

The articles about sound laws have different capitalisation (Sievers' Law, Edgerton's Law, Bartholomae's law, Szemerényi's law, Stang's law, Siebs' law) -- cf. also Indo-European sound laws#Sound laws within PIE. Is there any policy about such cases? Are these proper names? --ἀνυπόδητος (talk) 20:47, 21 December 2008 (UTC)

That's not the sort of thing Wikipedia has policies about, but the Chicago Manual of Style says the word "law" should be lowercased in such cases. —Angr 20:55, 21 December 2008 (UTC)

Could you move Sievers' Law? I did the rest but I could't move this one because I'm not an admin. --ἀνυπόδητος (talk) 21:09, 21 December 2008 (UTC)

 Done —Angr 21:36, 21 December 2008 (UTC)

Subscript numbers

When viewing this page in Firefox, the subscript numbers after the h's representing the laryngeals (*h₁, *h₂, *h₃ ) are invisible. They are however visible in Internet Explorer (oddly, since IE is normally held up to be the difficult one). --rossb (talk) 19:09, 28 February 2009 (UTC)

I'm using Firefox, and they look fine to me. Does it help to use the {{PIE}} template? How do *h₁, *h₂, *h₃ look? —Angr 19:35, 28 February 2009 (UTC)

I notice sometimes that table cell and line shaping is altered with subscripting/superscripting. It may have more to do with that than anything else. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 00:52, 1 March 2009 (UTC)

The {{PIE}} template doesn't seem to help, and it doesn't just happen in tables. If I use the ordinary numbers with the html "sub" element rather than the special charactes, it looks fine, like this:(*h₁, *h₂, *h₃ ). Similarly the subscript characters in Aeusoes1's signature look OK. The same problem occurs by the way in other articles dealing with the laryngeals. If it helps I'm using Firefox 3.0.6 on Windows Vista. --rossb (talk) 09:48, 1 March 2009 (UTC)

Interestingly I've now tried it on another PC on Firefox 3.0.5 on Windows XP and it looks fine! --rossb (talk) 10:05, 1 March 2009 (UTC)

Vocalic "allophone"???

As I understand, an "allophone" is a non-pnonological variant of a phoneme. As soon as it acquires phonemic role, it is not an allophone any longer. So, I don't understand why "i" and "u" when they are prononunced as vowels, are called "vocalic allophones" in the article; they are vowels. I would rather say that "y" and "w" are the consonantic allophones of "i" and "u", as in Spanish: i and u are vowels; but they can also be semivowels when standing with another vowel. There is no reason to call the "u" a "vocalic allophone" in the word *tuH, for example, where it is a vowel. --El Mexicano (talk) 12:11, 17 March 2009 (UTC)

Not really, read the section further: *u and *i lack several important properties that "real" PIE vowels *e and *o exhibit. Thus they're underlyingly consonants, phonetically really just vocalic allphones of of underlying *y ([j]) and *w. That they're phonemicised as vowels in most of the direct reflexes in immediate daughters doesn't invalidate their consonantic character in the mother language. *u in *tuH is a vowel just phonetically, with respect to the syllabification rules, phonologically it's a vocalic allophone of underlying *w. Evidence for "real" PIE *i and *u is really thin and very much debatable. --Ivan Štambuk (talk) 13:09, 17 March 2009 (UTC)

Also, to clarify, an allophone is any variant of a phoneme. So, for example, Spanish /i/ may be [i], [j], [ei̯]. The symbol between the slashes is an abstraction. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 16:41, 17 March 2009 (UTC)

It's OK, but why should we call a phoneme consonant when it is really a vowel in most cases? Also a semivowel is rather a vowel than a consonant, a vowel that you pronounce very shortly. Logically, if *e and *o are vowels in *hegom, why isn't a vowel *u in *tū? Would you be able to pronounce that u as a consonant, like [tv:]??? I don't think. Moreover, if they can be also long, they must be vowels, at least for me. --El Mexicano (talk) 19:37, 17 March 2009 (UTC)

That sort of thing (calling it one thing when another thing is more frequent) happens all the time. Spanish "voiced stops" are much more often voiced approximants, English /t/ is rarely simply a "voiceless stop" since it's either aspirated, a flap, or a glottal stop more often. Most notably, we consider /r/ to be a consonant in English, though it is much more frequently part of a syllable nucleus (I am speaking of my own dialect and those like it).

The article says that "Long variants of these vocalic allophones may have appeared already in the proto-language by compensatory lengthening" meaning that it's arguable that the long variants were separate phonemes, but certainly that paved the way towards them becoming so. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 23:36, 17 March 2009 (UTC)

In any case, while there certainly is room for debate about the best characterisation of these sounds, they are generally treated as vocalic allophones in the field, so that's what this article should reflect. It's worth adding that consonants certainly can be long, although this isn't so obvious if you're used to them being treated as "doubled". Finally, it's far from unheard of for sonorants to form the nucleus of a syllable. Croatian, for example, has cases of syllabic [r]. garik (talk) 23:48, 17 March 2009 (UTC)

I understand that calling a sound consonant or vowel, is just conventional. But I think in this special case, the article should explain clearly why *i and *u are considered (vocalic allophones of) consonants and not vowels. Let's see the basic principles: what's called a vowel? A sound that you can pronounce on its own. And a consonant is what you can pronounce only with a vowel. Sonorants like l, m, n, r can really behave like vowels (as for making a syllabe), though they are commonly called consonants; the same way, i and u can behave like consonants, e.g. in Spanish cuando ['kwando] or hielo ['jelo], nevertheless, they are vocalic phonemes. What frequency is reconstructed for PIE *i and *u as vowels in words? It would be arguable to call them vocalic allophones of consonants only if they appeared very rarely as a nucleous of syllabes, and rather as a consonantic element (semivowel). --El Mexicano (talk) 07:52, 18 March 2009 (UTC)

Your definition of the distinction between a consonant and a vowel can be enlightening to readers, but is a bit oversimplistic because of sonorants like the ones you listed as well as "semivocalic" sonorants like [j ɥ ɰ w]. There are also languages with words that have no vowels. I've seen a number of phonology texts make a clear distinction between consonants and sonorants (as well as vowels) because of the different ways they behave.

I agree that the article should explain more regarding vowels. I'm not even sure how many vowel phonemes PIE had because the article just lists allophones without describint conditions. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 16:16, 18 March 2009 (UTC)

Exactly, you have understood the problem. Talking about "allophones" has no sense without defining the phonological context they can appear and determining why they are just allophones and no phonemes. The definition should contrast phonemes vs. allophones by presenting some examples. --El Mexicano (talk) 17:07, 18 March 2009 (UTC)

PIE *y & *i and *w & *u are the same phonemes. Together with *m, *n, *l, *r they're usually called "resonants" because their syllabicity is determined by a simple rule: 1) vocalic allophones (*i, *u, *l̥, *r̥, *m̥, *n̥) are found between 2 nonsyllabics 2) consonantic allophones (*y, *w, *l, *r, *m, *n) next to a syllable peak 3) when 2+ "resonants" are next to each other, the rightmost gets syllabified first. See here (4.2.2.2, per Schindler 1977) for a formulaic description. From a phonological viewpoint, there is thus absolutely no difference between *y & *i, and *w & *i. PIE phonology article provides more detailed treatment on the difference of "real" vowels *e and *o, and *i and *u: Such syllabic sonorants therefore surface as vowels, but are distinct from the real vowels by that fact that they participate in ablaut alternations with their nonsyllabic pairs, and by the distribution; while the vowels *e and *o can be positioned only in the syllable nucleus, sonorants can also make appearance in both syllable nucleus and onset. Whether *i and *u are vocalic allophones of underlying consonants, or vice versa, is determined by more subtle factor than which of them frequents most. The section where the disputed claim is made alread links to PIE phonology article, and detailing the conditions of allophony would be an overkill for it, and there is really no need for it in such short section.. --Ivan Štambuk (talk) 18:55, 18 March 2009 (UTC)

It's not unreasonable to provide a quick summary of information like what the actual phonemes of PIE are. I'd do it myself, but a lot of the information regarding vowels that we could transfer over to here is uncited. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 19:21, 18 March 2009 (UTC)

For Ivan Štambuk: In Spanish, the situation is the same. i and u can appear as a nucleous of syllabe, e.g. mundo and lindo, and also as semivowels, e.g. cuanto, deuda, aire, tiene, however, they are always considered vowels, even if the /i/ is represented by -y in words like ley, soy. --El Mexicano (talk) 20:28, 18 March 2009 (UTC)

I'm not sure where you're getting that information, but /i/ and /u/ are either considered semivowels in prevocalic contexts or are considered the first element of a diphthong (and therefore not part of the onset or coda). See Spanish phonology. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 23:11, 18 March 2009 (UTC)

See "semiconsonante" and "semivocal" in the DRAE. When they are in prevocalic position, are called "semiconsonants" and in post-vocalic position, "semivowels". Anyway, it does the same, because in both cases they are treated as vocales (vocalic phonemes) and not consonants. For instance, in the word hielo, "i" is always a vowel, though it behaves as a consonant; also depends on individual pronunciations. Some persons pronounce it as ['ɟelo], others as ['jelo]. Another example is the word recaudador, I've heard a pronunciation like [rekaβda'dor] and not [rekawda'dor]. --El Mexicano (talk) 14:31, 19 March 2009 (UTC)

The terms "semivowel" and "semiconsonant" demonstrate that the vowel/consonant distinction can be limiting. By the way, many Spanish dialects demonstrate a distinction between a non-vocalic/semivocalic/semiconsonantal allophone of /i/ and /u/ and actual consonantal approximants which are not allophones of these vowels. Spanish phonology talks about this. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 15:43, 19 March 2009 (UTC)

How ‘i’ and ‘u’ are pronounced in Spanish has nothing to do with this article. According to published reliable sources: In PIE, */y/ and */w/ behaved phonologically the same as the other sonorant consonants, */m/, */n/, */l/ and */r/. And they were quite unlike the vowels, */e/, */o/ and */a/, which were subject to ablaut. Unless El Mexicano has a reliable source that says otherwise, his musings do not merit discussing here. —teb728 t c 21:25, 19 March 2009 (UTC)

Dear TEB728, let me tell you that logical thinking doesn't require any source. It requires just thinking. If a vocalic sound is able to make syllabe and can be even long, then it's a vowel and not a consonant, not a vocalic allophone, not a sonorant, but a vowel. That's the point. Goodbye. --El Mexicano (talk) 08:59, 20 March 2009 (UTC)

Unfortunatelly, OR policy doesn't require us to think, but to quote reliable sources supporting dubious claims. All the standard PIE phonology handbooks refer to *i and *u as vocalic allophones of *y ([j]) and *w. Phonetically *i and *u are indeed vowels, but phonologically they're really sonorants, not different from *m, *n, *l, *r (and their respective syllabic allophones). They couldn't be "long" - that's the post-PIE development. --Ivan Štambuk (talk) 11:43, 20 March 2009 (UTC)

Then how do you explain tū in Latin, Celtic, and other Indo-European languages? Anyway, if you need a source (unfortunately it is not in English), according to the Hungarian language language encyclopedia, A világ nyelvei [The Languages of the World] (1999), PIE have had five vowels, /a, e, i, o, u/ and all of them could be short and long. The other problem is, how could you explain the feminine gender ending -a, also present in several languages, if there was no /a/ in PIE? It can be found in Greek, Latin, Romance and also Slavic languages... --El Mexicano (talk) 15:10, 20 March 2009 (UTC)

Ever heard of the laryngeal theory? Tū is from *tuH, and the feminine ending is *-eh₂. —Angr 16:48, 20 March 2009 (UTC)

Actually, when it comes to monosyllabic words (personal pronouns, adverbs and misc. particles) in PIE, phonetic lengthening could occur, yielding not only *ē and *ō, but also *ī and *ū of non-laryngeal origin. I.e., PIE had e.g. for "you" both *tu (> Latvian tu, Doric τὐ) and *tū (> Latin tū, OCS ty, Lithuanian tù, Old Irish tú, Albanian ti), for "now" both *nu (> Sanskrit nú, Ancient Greek νυ, Latin nu-nc, Hittite nu, Old Irish nu, no) and *nū (> Sanskrit nū́, Ancient Greek νῦν, OCS ny-ně, Lithuanian nù), for "not" both *ne (> Sanskrit ná, OCS ne, Latin ne-que, Gothic ni-h) and *nē (> Latin nē, Old Irish ní, Gothic ne, OCS ně-) etc. One cannot explain the length in cases such as this by laryngeals due to the existence of short variants. Hence there is no point in reconstructing *tuH, *neH or *nuH in cases such as these, because it is much more likely that these words originally had short vowel which could be phonetically (but not phonologically!) lengthened, which is typologically not so uncommon with monosyllabics. The "real" lengths in nouns such as *pṓds had no fluctuation with dual forms which would indicate secondary phonetic lengthening. The cases with dual reflexes in same branch, or even the same language, show that the phenomena certainly existed in the parent language. Also very interesting is Balto-Slavic accentuation evidence which shows that "real" ablaut lengths (morpho-phonologically conditioned ones, i.e. the usual lengthened grades) yield regularly Balto-Slavic circumflex, whilst these secondary lenghts always seem to yield Balto-Slavic acute tone.. For those more interested, see this paper, p. 146ff ^_^

So basically El Mexicano is prob. true about PIE *tū, but this kind of long *ū and *ī is sporadic and non-phonological for PIE: all the other instances of them can easily be reduced (or assumed without loss of generality) to *iH and *uH sequences.

El Mexicano: As for the number of vowels in PIE - it's already discussed in [[PIE phonology]] article, from various viewpoints (minimal and strict phonological to phonetic one), including on the status of marginal and disputed *a. We cannot just state "PIE had X vowels" without bringing in the full perspective. --Ivan Štambuk (talk) 01:08, 21 March 2009 (UTC)

A question (just for fun)

How would you say in PIE:

"This user is able to contribute with a basic level of PIE"?

I'd like to make a userbox in the Hungarian wiki for it. Thank you. :) --El Mexicano (talk) 07:10, 24 March 2009 (UTC)

It's impossible to reconstruct syntax so there's no way to make any reasonable guesses about PIE word order. Because of that, the userbox would also be a lie. Sorry to be a negative Nancy. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 21:17, 25 March 2009 (UTC)

bʰ, dʰ and gʰ

Dʰ, bʰ and gʰ are sounds which are impossible to pronounce.

As far as I am concerned, PIE did not use them.

However, PIE did use dʱ and gʱ (plus variants of gʱ). Spacevezon (talk) 20:43, 25 March 2009 (UTC)

We're using the common way linguists transcribe PIE, which is very closely related to IPA, but detours in some regards. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 21:15, 25 March 2009 (UTC)

References in popular culture

Maybe the Wikipedia article should have a section on Proto-Indo-European in pop culture and just American culture in general. There's the American Heritage Dictionary, which has had an appendix of PIE roots since the 1970s, a 1992 book "A garden of words" about flower name etymologies by Martha Barnette, that "Atlantean language" made by Mark Okrand and appearing in "Atlantis: The Lost Empire" in 2001, and does anybody know anything about the presumably made-up language spoken by the baddies in "10,000 BC" by Warner Bros 2008? That sounded similar to PIE. "Finnegan's Wake" by James Joyce is also another big one. I once read a commentary on that book that made reference to Joyce' usage of PIE roots and related studies to formulate his garbled masterpiece.

Mention of these works could help increase public awareness of academic thought, something Wikipedia was founded to do.

35.8.218.54 (talk) 00:16, 20 May 2009 (UTC)

Actually, I kind of thought it was a nice change to have a Wikipedia article that didn't have an "In popular culture" section for a change. +Angr 10:18, 20 May 2009 (UTC)

Proto-Slavic

I'm just wondering, should "proto-Slavic" be included in the "Daughter proto-languages". I know "Proto-Balto-Slavic language" is already there, but proto-Balto-Slavic and proto-slavic are different. Would proto-Slavic instead be a 'granddaughter language' or something.Hypershock (talk) 04:48, 10 November 2009 (UTC)

I think it's best just to list Proto-Balto-Slavic as it's the immediate descendant. If we start adding later derivatives, the list will get out of hand, with a "Proto-X" added for every conceivable subgrouping whether plausible ("Proto–Insular Celtic") or not ("Proto–Western Germanic"). +Angr 20:17, 10 November 2009 (UTC)

Given that Balto-Slavic is still a controversial hypothesis and that Slavic is a well-understood and distinct language group, it would make sense to make a separate node out of it. 78.49.239.15 (talk) 15:59, 21 June 2011 (UTC) Wojciech Żełaniec

Finno-Ugric

I still can not believe that the Finno-Urgic language tree does not fit in somewhere under the Proto-Indo-European hierarchy. Especially since the Corded Ware and Comb Ceramic cultures overlapped each other, and did so during concordant time horizons.

I just can't believe it.

--Atikokan (talk) 05:04, 29 November 2009 (UTC)

References:
http://en.wikipedia.org/wiki/Corded_Ware_horizon
http://en.wikipedia.org/wiki/Comb_Ceramic_Culture
http://upload.wikimedia.org/wikipedia/en/4/4f/IndoEuropeanTree.svg
http://en.wikipedia.org/wiki/Urheimat

See argument from personal incredulity --Pfold (talk) 15:50, 29 November 2009 (UTC)

You could fill a whole encyclopedia with a list of things some people just can't believe, yet which are true. garik (talk) 16:21, 29 November 2009 (UTC)

Also cf. Indo-Uralic languages. --Trɔpʏliʊm • blah 22:11, 6 July 2010 (UTC)

Ergative and Active hypotheses

I'm suprised this article has nothing on what amounts to basically a century of discussion on the syntax of early PIE with respect to its typological classification as nom/acc, erg/abs, or active/stative. A recent list of references on the ergative may be found here:http://versita.metapress.com/content/r26389132nk67172/fulltext.pdf. As for the active/stative hypothesis, recent lists of references and discussion may be found in Lehmann's 1993 Theoretical Bases of Indo-European Linguistics and Bauer's 2000 Archaic Syntax in Indo-European —Preceding unsigned comment added by AD Messing (talk • contribs) 17:16, 14 January 2010 (UTC)

glossary

add a large chart of core vocabulary with cognates in all major families (might be a good idea to move the full table to a separate article and include an excerpted table in this article)

Yeah, if a single new section triples the length of an article (other than a stub), my instinct would be to make a separate article instead. And then (in this case) nominate it for deletion because WP:DICT. —Tamfang (talk) 06:01, 21 May 2010 (UTC)

I'll happily move the table to a separate article, and take out some of items that are arguably not "core vocab", but I don't hardly think it qualifies as a dictionary. This table began as an expansion of a table in Armenian language that has 40 PIE items along with cognates in Armenian, Modern English, Old English, Latin, Greek, and Sanskrit. Meanwhile, there is a similar but shorter table in Tocharian languages that has 20 PIE items along with cognates in Modern English, Tocharian A and B, Old Irish, Latin, Greek, and Sanskrit, and a similar table in Lithuanian language that had 17 Lithuanian items along with cognates in Latin and Sanskrit. There seemed an obvious need felt for tables of this sort, and it seemed like something that was more useful in a PIE article rather than duplicated in various different daughter languages -- although I realize now it would go even better in a page linked off of Indo-European languages rather than this page. At 150 or so items this chart is somewhat larger than the 40 item Armenian chart, but hardly dictionary size, and its point is not to be a dictionary but a set of examples of how cognates evolved. Benwing (talk) 07:24, 21 May 2010 (UTC)

Vocabulary lists do not belong on Wikipedia, they belong on the Wiktionary sister project. See the existing content at wikt:Category:Proto-Indo-European language, wikt:Appendix:List of Proto-Indo-European roots. It is one thing to give a brief list of lexemes in a language article, it is another to embark on a major project of compiling a full-scale dictionary on-wiki. That project exists, and it is called Wiktionary.

What I can see are a greater number of examples at Indo-European sound laws, where we currently just give a table of phonemes plus a bunch of random examples.

Imho, the table at Armenian_language#Indo-European_linguistic_comparison is misguided. It should illustrate Armenian sound laws, and would perhaps be more at home at Proto-Armenian, but there is no conceivable reason why that table should be burdenend with random cognates from English, Greek, Latin or Sanskrit.

--dab (𒁳) 08:17, 21 May 2010 (UTC)

Exceptional claims require exceptional sources.

I believe potentially that a number of sections in this article require the above rule to be applied to them, some examples are highlighted below along with a sentiment I have regarding the use of "unattested".

"The Proto-Indo-European language (PIE) is the unattested,"

a good descriptor in one sense but I contend that the complexity "unattested" brings could be cleared with the inclusion of a more suitable adjective in its place. Use of 'Unrecorded' or extrapolation on the fact that the PIE was never physically stored would seem a better idea.

"reconstructed common ancestor of the Indo-European languages, spoken by the Proto-Indo-Europeans. The existence of such a language has been accepted by linguists for over a century, and reconstruction is far advanced and quite detailed

Scholars estimate that PIE may have been spoken as a single language (before divergence began) around 4000 BCE, though estimates by different authorities can vary by more than a millennium."

I think at a minimum the assertions I have bolded require evidence of deductive reasoning through referencing, but before putting this on the watchlist, I would appreciate hearing a more informed opinion on the subject to assist Wikipedia/Myself in deciding if this page would be a good use of time to tackle. BoredextraWorkvidid (talk) 18:06, 22 August 2010 (UTC)

None of these claims are in any way "exceptional". The proper guideline to point to is WP:LEAD. The lead section should be a coherent summary of the article body, and does not itself need any references. The content it summarizes should of course be referenced, in the article body where it is expounded in greater detail. None of the things you put in boldface raise an eyebrow and can likely be found in any introduction to the topic. If you intend to "tackle" this page, I strongly recommend you read at least one introduction to Indo-European studies before you proceed. Several good such introductions are listed in the "References" section. If you find any statement in this article that is in blatant contradiction with a statement you find in one of these books, it will be early enough to call WP:REDFLAG. --dab (𒁳) 12:24, 24 August 2010 (UTC)

Aorist

The article Aorist is in need of editors who can help develop it, both in general and particularly in an IE overview section. If there's anyone who watches this page who can spare some time, your input would be much appreciated. Cynwolfe (talk) 20:47, 16 September 2010 (UTC)

Future tense in PIE

from Talk:Shall_and_will#Future_tense_in_PIE:

According to an unsourced statement at Proto-Germanic#Verbs Proto-Indo-European had no future tense, contradicting this article. I think I have heard this from other places too. Count Truthstein (talk) 00:05, 4 January 2009 (UTC)

I added a reference that mentions PIE future tense based. Please refer to item 9 here. 192.102.209.29 (talk) 23:44, 13 April 2010 (UTC)

Answered at Talk:Shall and will#Future tense in PIE. --ἀνυπόδητος (talk) 20:16, 26 September 2010 (UTC)

clarification question

"As there is no direct evidence of Proto-Indo-European language, all knowledge of the language is derived by reconstruction from later languages using linguistic techniques such as the comparative method and the method of internal reconstruction. PIE is known to have had.... "

If there's no direct evidence for it then how can we 'know' anything? We can believe, confidently believe or wildly speculate, but know? —Preceding unsigned comment added by 92.29.202.215 (talk) 04:42, 8 November 2010 (UTC)

How do we know that the universe is 13,500,0000,000 years old, or that the dinosaurs existed, that medicine isn't actually poison, or that you were born of a woman? If you are a skeptic, then no explanation will ever satisfy you, and only a life of hypocrisy is open to you. If you are a scientist, and want to do the work to know whether historical linguistics is valid, don't ask some stranger at wikipedia, go to the library and study the topic yourself for half a decade or so. I suggest you begin with Mario Pei. After a few years' reading you can study Saussure's theory of the coefficient sonantique, and then the conformation of his predictions with the decipherment of Hittite. Then you can judge for yourself.μηδείς (talk) 06:20, 8 November 2010 (UTC)

Stang and the vowel repertoire

“ In fact, Stang's law, a phonological rule active at the Proto-Indo-European stage, produces *-ām from older *-eh₂m as the ending of the accusative singular of the stems in *-h₂ (and possibly *-ās from older *-eh₂ns in the respective accusative plural), and therefore compels the assumption that the phoneme *ā was present in the language. ”

This new paragraph could stand a bit more fleshing-out: the chain of reasoning is not as obvious as one might like. Doesn't Stang's law predict /eh₂m/ → /ēm/ ? —Tamfang (talk) 22:43, 28 February 2011 (UTC)

*h₂ is a-colouring, independent of Stang. For example, the acc pl of the feminine suffix is *-eh₂m > *-ām.--ἀνυπόδητος (talk) 09:32, 7 March 2011 (UTC)

If *h₂ is irrelevant to Stang, where does the lengthening come from? —Tamfang (talk) 18:33, 7 March 2011 (UTC)

Sorry for not being clearer: the lengthening comes from Stang's, while the colouring (a versus e) is an independent effect of the laryngeal. --ἀνυπόδητος (talk)

Proto Indo-Europeans were Turk!

Proto Indo-Europeans were Turk!

Since proto indo-europeans originated in Central Asia so they were Turk they were T U R K Humanbyrace (talk) 00:08, 31 May 2011 (UTC)

See Nostratic. μηδείς (talk) 00:17, 31 May 2011 (UTC)

Proto-Indo-European didn't arise in Central Asia, but on the Ukrainian steppes. But since a certain species of wild ass also arose in Central Asia... --Taivo (talk) 01:46, 31 May 2011 (UTC)

May need more research to include Dravidian languages with PIE

So many similar sounding English words in Tamil, I doubt these classification,. http://aruniyan.wordpress.com/2010/10/28/english-words-that-sounds-like-tamil-or-originated-from-tamil/ — Preceding unsigned comment added by Malarmisai (talk • contribs) 13:01, 21 August 2011 (UTC)

No amount of original research would be enough because Wikipedia has a firm policy against inclusion of original research. For inclusion of a connection there would have to be significant published acceptance by respected linguists, and such acceptance does not exist. See Dravidian languages#Relationship to other language families for the current state of research. —teb728 t c 00:18, 22 August 2011 (UTC)

Dravidian languages are firmly excluded from PIE. Alexandre8 (talk) 01:30, 24 September 2011 (UTC)

This article grossly fails to give Wikipedians a coorect and accurate synthesis of the Proto-Indo-European language propblem

When De Saussure first proposed the laryngeal theory many scholastic linguists rejected his claim (they were indeed imitating the story of dogmatic papal councellors who did hang Galileo for him stating the truth about earth being spherical) but own empirical proofs from Anatolic Indo-European showed that Saussure was right!

We are here confused by this scholastic artithmetico-algoritimical approach wich kidds us with unprunucable words such as bhedh (is bh standing for beh or bhe!!?) and leads us to an imaginary aberrated frozen unrealistic unreal and never spoken Proto-Indo-European.

It's in fact a relic or let me say an ARTEFACT of the ancient germano-nazocentristo-linguistical paradigm wich saw in Sanskrit the perfect model of the Proto-Indo-European language but now we know that most of the 19 th and pre-Hitler 20 th century Indo-Europeanist field was just RUBBISH

Indeed after the recent phylogenetical Bayesian study we know that:

1/Armenia/Anatolia is the homeland of proto-indo-european

2/The Larynegal theory is right

3/The glottalic theory is right

4/Anatolic then Armeno-Greek then Aryan then Tocharian are the real model of Proto-Indo-European language

5/Indic is Aryan spoken by local non Indo-European indians

6/Iranic is older (because more diverse see Sughni, Scythian, Pashto, Yaghnobi huuuge internal diversity ...) and more archaic than Indic

7/Proto-Indo-European language's glottalics correspond to Proto-Semitic empohatics

8/Proto-Indo-European language's larynegeals correspond to Proto-Semitic larynegals

9/Proto-Indo-European language and Proto-Semitic are closely related (phonetically, lexically, structurally and morphophonologically[ablaut]) and they are forming 2 branches of the Lislakh phylum and are diachronically connected with the Hassuna-Half-Natufian-Araxes cultural (sites that saw the oldest attested Swastika motifs) complex wich is associated with the J1 and J2 hg's and the southwestern+westernasiaitic autosomal components (see Behar 2010)

10/Finally the Nominal system of Proto-Indo-European language was initially regular and similar to the proto-Semitic one

indeed according to Edward Lipinsky Proto Semitic had 7 cases (I have added to his model 2 vocative cases retrieved from modern Arabic)

Proto-Semitic cases/singular/dual/plural

1/vocative1(no ending)/-/-/-

2/nominative/-u/-aa/-uu

3/genitive/-i/-ay/-ii

4/accusative/-a/-ay/-ee

5/locative/-um/-um/-um (present in Arabic "labbayka allahUM labbayk"=(we) pray-you toward you Allah (we) pray-you

6/benefactive/-ish/-ush/-uush (akkadian "zikaram daqitu sarrISH"=I killed the man for the king)

7/comitative (by, with)/-am/-am/-am (ancient Hebrew yodAM=by hand)

8/dative(toward)/-ah/-ishum/-ishuum (ancient Hebrew BabelAH=toward Babel)

9/vocative2/-aah/-ayh/-uuh (modern Arabic "abAH"=o father)

As you can notice the case endings are regular and similarly pre proto-Indo-European (before the case erosion due to migration of proto-Indo-Europeans from their homeland in Western Asia to Europe and India) case system should be as the below

Proto-Indo-European cases/singular/dual/plural

1/vocative/-h2/-ah2/-oh2

2/nominative/-s/-aas/-oos

3/genitive/-h1/-ah1/-oh1

4/accusative/-h1m/-ah1m/-oh1m

5/locative/-h3/-ah3m/-oh3m

6/dative/-ah/-ahm/-ohm

7/ablative/-eh3/-ah3/-oh3

As for the Proto-Indo-European instrumental case it's merely the post-agglutination of Lislakh (Semitic+Indo-European common) bh1 (Ensglih by, Semitic bi) to the nouns we need to make the instrumental case of

Humanbyrace (talk) 15:54, 9 October 2011 (UTC)