Tropylium

Welcome!

Hello, and welcome to Wiktionary. Thank you for your contributions. I hope you like the place and decide to stay. Here are a few good links for newcomers:

• Wiktionary Tutorial
• How to edit a page
• How to start a page
• Our layout policy (nicknamed “ELE”)
• Criteria for inclusion (nicknamed “CFI”)
• Wiktionary Sandbox (a safe place for testing syntax)
• What Wiktionary is not
• FAQ

I hope you enjoy editing here and being a Wiktionarian! By the way, you can sign your name on Talk (discussion) and vote pages using four tildes, like this: ~~~~, which automatically produces your name and the current date. If you have any questions, see the help pages, add a question to one of the discussion rooms or ask me on my Talk page.

Did you know that Wiktionary can be adapted to fit your style? Go have a look at your personal preference page (nicknamed “PREFS”)! In particular, you can choose the option to show hidden categories, which will be of great help if you are looking for work to do.

Again, welcome!

—Ruakh_TALK 20:48, 26 December 2008 (UTC)

mies

Thanks for clearing up those Finnish questions in the Etymology Scriptorium. Would you happen to have any information on the etymology of (and lack of a final vowel on) mies which would be pertinent to the discussion here? - -sche (discuss) 04:19, 21 March 2014 (UTC)

(Continuing this discussion here) When you mentioned Germanic *mēgaz I started wondering. Proto-Germanic *g was actually [ɣ] between vowels, so is it possible that this was borrowed as *mexas before the loss of *x? A sequence like *exa would give *ee I would think. And if that's not plausible because of the relative dating of changes and loaning, maybe this could still be applied to a native word. The explanation, then, would be that *mees was originally a two-syllable stem which contracted, and therefore did not necessarily have a final consonant. I don't know if the loss of *x must necessarily precede the apocope of *i after two syllables, though. If it does, then *mexVs must be the earliest known stage, but if it doesn't then *mexVsi > *mexVs > *mees is also a possibility. —CodeCat 23:22, 4 April 2014 (UTC)

Interesting idea. We don't know of any precedents for anything like *-exa-, so that's not entirely ruled out.

Still, I believe that there was a change *ŋ > *x fairly early on in the Proto-Finnic era; the two are vocalized completely indistinguishably, and unlike the other nasals but alike semivowels, *ŋ fails to condition primary long vowels (the change *a *ä > *oo *ee). Given *keŋäč > *kevät "spring", I'd expect an incoming *mexäs (disharmonic stems might be anachronistic this early) to similarly end up as **meväs; or, even if vocalized completely, as **möös, given *mexə- > *möö- "to sell".

Your second approach clearly won't work, I'm afraid. There are quite a few roots like *šiŋərə > *hiiri "mouse", not **hiir; *śäxərə > *sääri "thigh", not **säär.

What might be possible is to assume that the root was adopted after the rise of primary long vowels, yet as an *ə-stem (there are plenty of examples of Germanic *-az ~ Finnic *-eh or *-es). I.e. *mēɣas → *meexəs > *mees. Though that does seem chronologically difficult — sufficiently old Germanic loans only ever seem to show *k/*g/*x → *k, I think. And *-eexə- might have still yielded *-eeve- anyway.

Even better regularity might be attainable if this were an older loan still. I don't know the etymology of the Germanic word — but if this had an original palatovelar, something like Late PIE *mēǵʰos would be expected to be adapted as Pre-Finnic *mejəs, from which *mees would be entirely expected. --Tropylium (talk) 03:14, 5 April 2014 (UTC)

Language codes, templates, etc.

Our language codes are based on the ISO 639 ones, but are far from an exact match (you can look through them here). You may disagree with our codes, but if you use a language code that's not in our system, all that gets displayed is a module error. The same thing happens if you change the IPA-based spelling in a template to a transliteration, but don't replace it with something else. You should never make an edit like this (diff), and if somehow you accidentally do, you should either fix it right away, or revert it if you can't. Always check the results of your edit before you leave the page. Thanks! Chuck Entz (talk) 08:06, 17 June 2014 (UTC)

You should add this exception to the description at WT:KCA TR. The documentation of transliteration schemes is important. --Vahag (talk) 23:11, 11 July 2014 (UTC)

Noted, thanks. --Tropylium (talk) 23:30, 11 July 2014 (UTC)

Proto-Samic *nealjē

Does PU *ń- > PS *n-? And does PS *nea- > Northern Sami njea-? Or is something else going on? —CodeCat 22:33, 24 October 2014 (UTC)

The reconstruction is wrong, is going on. :ı Apparently there has been an assimilation development *n > *ń due to the word-internal /j/ in several Uralic languages that has led to many sources to reconstruct original *ń-. --Tropylium (talk) 22:40, 24 October 2014 (UTC)

Finnic verbs in -c-

I created , which seems to be well-attested. But I'm not sure about the reconstruction. Is the stem reconstructed with a single -c- or a geminate -cc-? If the former, then why did it not become -s- in Finnish? Furthermore, it appears that -ct- regularly becomes -tt- in Finnish, but it was apparently changed to the weak grade form -t- analogically. Are there other examples of this? Could you also check the conjugation? —CodeCat 01:07, 3 November 2014 (UTC)

Certainly a geminate. The contraction of the infinitive from *-ccet- to *-tt- seems to be only North Finnic. Contrast > Finnish veistää, Votic vessǟ, Estonian vestä. --Tropylium (talk) 14:15, 3 November 2014 (UTC)

For veistää it seems that the form was though. The form you gave would only have one "ä" in Finnish. But what it does demonstrate is that veitsi + -tää gives veistää.

Still, I wonder what the regular outcomes of *c, *cc and *c' (half-long) are and how one would tell them apart when reconstructing (that is, which languages distinguish them). I thought that *c would always become s in Finnish but the veitsi example seems to indicate that's not the case. Yet there are lots of examples in which *c does become s in all the Finnic languages. So what's going on here? —CodeCat 15:11, 3 November 2014 (UTC)

Veitsi goes back to *cc, but the derivative veistää goes to, yes right, *veictädäk < *veiccə-tä-, with the consonant stem and cluster shortening dating already to Proto-Finnic.

Loosely, the development has been:

• *cc (strong grade) > Finnish/Veps/Estonian/Livonian ts, Karelian čč, Votic tts.
• *c̆c (weak grade) > Fi/Vep/Et/Liv ts, Krl č, Vot ts.
• *c > Common Finnic s, in some positions (but not all) with a residual ts ~ ds in South Estonian.

i.e. if it looks like an affricate, it's from *cc. More often than not, reconstructing *c requires internal reconstruction, either due to paradigmatic alternation with *t, or by etymology. E.g. asia is from a Germanic *anθija and hence must come from PF *acja. Or since veitsi has *cc, we have to reconstruct *veictä- and not *veistä-. --Tropylium (talk) 15:38, 3 November 2014 (UTC)

Given that the -i- in the diphthong of *veitsi was originally a consonant *j, how is it possible that *veicci has three consonants in a row? —CodeCat 20:07, 3 November 2014 (UTC)

Clusters with geminates are treated slightly differently from clusters with three different consonants. I think the chronology in these types of words was roughly (but don't cite me on this):

1. CVVCːV becomes a possible word shape
2. Coda *j and *w are reinterpreted as vowels
3. Words like *veicci are loaned or gain their current shape
4. CVVC₁C₂V becomes a possible word shape
5. CVCCːV becomes a possible word shape
6. CVC₁C₂C₃ becomes a possible word shape
7. CVVCCCV becomes a possible word shape

Proto-Finnic was a language at the 4th stage; we can reconstruct also e.g. *mëëkka 'sword', *joukko 'group', *paikka 'spot, mark', *viit-tä partitive of 'five', *puu-sta elative of 'tree'. On the other hand, alternations like *purttu > *purtu 'bitten' and *oncca > *occa 'forehead' were still productive. Apocope presumably dates to stage 2 (before this we'd've had been something like *veńćə, *vijtə-tä, *puwə-sta). --Tropylium (talk) 20:39, 3 November 2014 (UTC)

The Finnic prolative

I've collected a few cognate suffixes at -tse, but I'm a bit puzzled by the distribution. Finnish -ts- and Karelian -čč- clearly imply a strong-grade -cc-. But Finnish has word-final -e here along with an assimilative final consonant, which implies a previously lost consonant like -k or -h. Yet such a final consonant is incompatible with the strong grade found in Karelian. The change of final -e in Finnish with -i in the other languages is also puzzling, and perhaps even stranger is that Estonian kept the -i. What can you make of this? —CodeCat 01:47, 5 November 2014 (UTC)

At face value I would assume that the Finnish form has simply been affixed with a second element, i.e. *-icce-k, while the other languages retain bare *-icci. In Estonian suffix-final consonants sometimes have better odds of survival than stem-final ones, though that might not be the whole story about it.

(Incidentally etymologically it's not a suffix added to the plural stem, but rather a suffix that contains *-i- for its own sake, much like -inen : -ise-. But I suppose the plural analysis comes rather naturally and might be preferrable for modern Finnish.)

Hakulinen in SKRK has a brief discussion of the Finnish form, but he does not touch on Estonian, Veps and Karelian, so it's not of too much help. --Tropylium (talk) 08:58, 5 November 2014 (UTC)

I created now. But I just found that Karelian also has alačči, with no -i-. Could this mean that the -i- is not part of the suffix after all? —CodeCat 15:23, 5 November 2014 (UTC)

I'd guess it's either a reanalysis by a similar interpretation as in Finnish about -i- being the plural marker; or a loan from Veps (as a part of the substrate in Olonetsian?), where *-jcc- > -(j)čč-, versus *-cc- > -cc-. Hard to say in the absense of an etymological dictionary of Karelian, though. --Tropylium (talk) 15:37, 6 November 2014 (UTC)

*hüvä

I noticed that Estonian and Võro have unrounding of the vowel and loss of the -v- here. Livonian also has ü > õ. Is this a regular process? —CodeCat 03:13, 5 November 2014 (UTC)

The Livonian development is regular, yes, and involves a breaking *i- > jõ- (following the loss of *h- and the unrounding of *ü). Estonian hea ~ hää is in origin a dialectal form (there has been a ton of dialect diversity in Estonian, and Standard Estonian is perhaps less consistent yet than Standard Finnish in what forms exactly have been adopted). Perhaps generalized from an inflected form, but I don't know the details. The expected uncontracted hüva ~ hüvä is also attested from both North and South Estonian, though. --Tropylium (talk) 08:45, 5 November 2014 (UTC)

Northern Sami attributive adjectives

Adjectives in Northern Sami have a separate attributive form that is used when the adjective modifies a noun rather than standing alone. If I'm not mistaken, this is actually the original situation in Uralic, and the Finnic concord of adjectives is an innovation. But I don't know where this form came from; it's often not identical to the nominative singular. Do you know anything about this? —CodeCat 22:52, 13 November 2014 (UTC)

linnu

Do some Finnic languages really preserve the plosive in -tn-?

Evidently so. Another example is 'lamb' > Finnish vuona, Estonian voon (apparently with no compensatory gemination after a long vowel?), but Veps vodn, Votic võdna.

What about in unstressed syllables? —CodeCat 20:07, 16 January 2015 (UTC)

I don't know how t-stem essives or participles (the type *olutna > arch. olunna 'as beer' (modern oluena); or *pelätnüt > pelännyt 'having feared') are formed in the key languages, but I would not be amazed if there had been an earlier assimilation. --Tropylium (talk) 20:32, 16 January 2015 (UTC)

The Finnic verb suffixes -tu- and -u-

I noticed that our entry for has its suffixal gradation applied wrong. This is because the module treats the hyphen as standing for two syllables, so in that respect it's working right. But it puzzled me why it's still -tua in all the Finnic languages, and I started looking for Finnish examples of this suffix. The examples I found were either attached to one syllable, three, or attached to a consonant which would inhibit suffixal gradation. So then I considered what the outcome of a suffix-gradated would be; that is, , etc. The first -d- would disappear in Finnish, so you'd get something like -auda, -euda, or possibly -uda (I'm not sure what process would create a monophtong here, though). And of course, this changes the syllabification, allowing the -d- to be dropped, resulting in -ua.

So my question is, am I on the right track here? Are -tua and -ua originally the same suffix? —CodeCat 18:57, 18 January 2015 (UTC)

Curious, but the details don't seem to work. Principal issue: *-tu- generally applies to nominal roots, not verbal ones like *-u- does, and it is often segmentable as *-ta- + *-u-. This is the case for all three words we currently have in this suffix category:

• jäätyä ← jäätää + -yä, not jää + -tyä
• kihartua ← kihartaa + -ua
• paljastua ← paljastaa + -ua

Your odd-syllable bias does seem to exist, but it goes back to the base suffix. The usual even-syllable verbalizer -ta generally has the reflective equivalent -utua, which is kind of double-marked. We'd indeed expect endings such as -au(d)a or -eu(d)a (which IIRC is attested dialectally). But instead we find e.g. pato (“dam”) → padota (“to dam”) → patoutua (“to be dammed”); or in the few, often adjectival cases where no intermediate causative verb exists, kapea (“narrow”) → kapeutua (“to become narrow”).

So I am now skeptical on if there are grounds to reconstruct an independent *-tudak for PF at all.

There is still a "strong" allomorph of -u-, but this is -pua, as applied in monosyllabic stems like juopua, syöpyä, saapua (and not ˣjouda, ˣsöydä, ˣsauda). This though probably involves an epenthetic consonant of some sort. --Tropylium (talk) 19:56, 18 January 2015 (UTC)

Rename Nenets to Tundra Nenets

I renamed the language, but we apparently have a few entries in "Nenets" already. I know nothing about this language, so could you go through Category:Tundra Nenets lemmas and rename the language sections, while also checking if they are indeed Tundra Nenets? —CodeCat 21:51, 30 January 2015 (UTC)

OK, I've looked thru everything currently at Category:Nenets lemmas and the thus far nonexistent Category:Tundra Nenets lemmas. They seem to be all indeed Tundra Nenets. I'm for now unable to verify тиртя (tyirtya) (looks like a derivative from тирць (tyircʹ°, “to fly”)?), сельбя (syelʹbya) and ӈылека (ŋilyeka), though phonologically none of these can possibly be Forest Nenets. --Tropylium (talk) 23:46, 30 January 2015 (UTC)

Thank you. I noticed that you left the header as "Nenets" though. Could you change that as well? —CodeCat 23:48, 30 January 2015 (UTC)

Certainly doable, with our current number of entries. I guess assuming those three entries to be TN as well will be safe enough. --Tropylium (talk) 23:53, 30 January 2015 (UTC)

Appendix:Proto-Samic/toalëtēk

According to the vowel shift at w:Proto-Samic, this form couldn't exist. oa becomes uo before ë. So is there something missing?

This just means it's not an inherited word. If you look at my wordlist in progress, there are dozens of words with other unetymological vowel combinations (*ā-ë, *oa-ë, *ea-ë, *ē-ē, *ē-ō).

Also, is this a cognate of tulla? —CodeCat 00:13, 1 February 2015 (UTC)

Probably not. IIRC Samic cognates of that are only known from Ter Sami. --Tropylium (talk) 00:23, 1 February 2015 (UTC)

Finnish se

This pronoun is very irregular, but I'm guessing that many of the irregularities may be archaisms. I'm struggling to figure out where all the irregularities come from, though. Could you shed some light on this, maybe?

1. The nominative singular has a single short vowel. That is unusual in itself, but being a pronoun, I'm guessing that this is inherited. It's curious though that there is no -e > -i change.
2. Most other case forms have the stem si-. This is rather puzzling to me. Why the e > i change?
3. Even more striking is that instead of -ssa, -sta, -hen in the interior cases, there is -inä, -itä, -ihen.
4. In the plural, it seems that there is the stem ni- with the regular plural infix -i- and the normal case suffix?

I'm also wondering about the nature of the "extra" cases to the right of the table. Several Finnish entries have these, although I don't know which exactly. They are apparently not true cases, but they do seem to have similar formations, so they might be of a similar nature to the -r of some of the English pronouns, like here, where, there. That is, a special ending used for pronominal suffixes alone.

The superessive and delative endings seem to reflect something like *-gellä, *-geltä, in which the g has disappeared. I have no material from other Finnic languages (Veps would be particularly useful) to compare it with. The sublative, if formed in the same way, could have *-k-na > *-nna?

What I'm also interested in is whether these endings, or at least part of them, have cognates in other languages, and what Uralic origins they may have. —CodeCat 00:18, 8 February 2015 (UTC)

Another interesting bundle of questions…

AFAIK most other Uralic languages indicate that the original shape of the root is *ći- or *śi-, and so oblique si- is probably more original. The sg.nom. se could have been be modelled after the pl.nom. ne, which is in turn probably modelled after the personal pronouns me, te, he, which come from earlier *mek, *tek, *hek or perhaps *met, *tet, *het. The vowel alternation formally goes back to at least Proto-Finnic.

The s/n alternation, then, looks like it would have been generalized from the alternations *to- : *no- > tuo : nuo and *tä- : *nä- > tämä : nämä. These two alternations are both also found in Mordvinic, and partly Mari and Samic, so they're definitely pre-Proto-Finnic in age.

The "elative" and "inessive" are actually the original Proto-Uralic locative *-na and ablative *-ta. (Other fossilized examples are the adverbs kotona, kotoa.) -i- in the plural is obviously an infix, yes; I guess it was added to the singular too to disambiguate between e.g. the partitive sitä and the (ab)lative siitä.

The "extra cases" are a very heterogeneous group, and I mostly think calling them inflected forms of se is not a very good analysis.

• siellä, sieltä are just regular local cases based on an extended stem √sikä-, also seen in sikäläinen (and moreover cf. tämä, stem tä- → täkä- → täkäläinen, täällä, täältä). These likely to go back to at least Proto-Finnic. Veps has indeed sigäl 'there'. (And there's even a possible exact cognate from Eastern Khanty: ťeɣəlä 'there', but this sounds very suspicious, especially since the L-case series is a Finnic innovation.)
• siis uses the same "lative" element *-s as appears in the adverbs alas, ulos, ylös, pois, edes, etc. Origin unknown, though it probably has something to do with the inessive/elative/illative case group.
• sinne uses an ending which regularly forms the terminative case in Savonian and Karelian, and seems to be related to the Estonian terminative -ni as well. This can be used on most pronoun roots: tänne, tuonne, minne, jonne etc.
• siten (also täten, joten, kuten, muuten) is probably in origin just the instructive. -t- could be from the 2nd infinitive, *-eden : *-ten, one of the most frequent places where the instructive is used.
• silloin (also tällöin, tuolloin, jolloin) should be segmented as two components -ll- (probably somehow from the adessive) and -Oin (cf. adverbs like muinoin, muutoin, vihdoin; probably somehow from the instructive).

--Tropylium (talk) 23:14, 8 February 2015 (UTC)

• Regarding silloin, it seems more likely that it's si-lla-in, with the regular rounding of suffixal a. After all, something like -ll- could hardly have existed as a word-final element, there was presumably a following vowel. That leaves me wondering what the -in element could be, but in any case it's some kind of extension of the adessive, referring to a particular place.
• Is there anything more you could say about the terminative case? Is it Proto-Finnic, and how was it formed? —CodeCat 00:00, 9 February 2015 (UTC)

Jäännöslopukkeen merkitseminen taivutuskaavoihin

Lienet huomannut, että joku - todennäköisesti CodeCat - on muuttanut verbien ja nominien taivutuskaavoja siten, että niissä näkyy jäännöslopuke pienellä yläviite-x:llä merkittynä. Katso esimerkiksi kirje, joutua. Mielestäni tämä on:

• hämmennystä aiheuttavaa, koska suurin osa käyttäjistä ei ymmärrä, mitä merkki tarkoittaa,
• turhaa, koska ne, jotka ymmärtävät, eivät tarvitse kyseisiä merkkejä,
• harhaanjohtavaa, koska joku saattaa ryhtyä luulemaan, että suomenkieliseen kirjoitukseen pitää sirotella pikku x:iä sinne tänne,
• raivostuttavaa, koska kävin CodeCatin kanssa asiasta keskustelun, mutta hän tapansa mukaan viis' veisaa muiden mielipiteistä,
• ehkä väärin, koska en ole vakuuttunut siitä, että jäännöslopukkeet voidaan läiskiä jollakin kaavalla universaalisti oikein. Sinä luultavasti tiedät tämän asian paremmin kuin meikäläinen, joka on koulutukseltaan insinööri.

Minusta jäännöslopukkeiden merkintä taivutuskaavoihin pitäisi siis lopettaa. Lausumiselle on oma kohtansa. Mitäpä itse olet mieltä? Jos olet samaa mieltä kuin minä, voisimme nostaa asiasta keskustelun Beer parlourissa ja keskusteluttaa yhteisöä siitä, kenen säännöillä täällä mennään, eli onko natiivieditorien mielipiteillä mitään merkitystä, kun kieltä totaalisesti ymmärtämätön häirikkö riehuu näppäimistönsä kanssa (tietysti asiallisin sanakääntein). Samalla voisi yrittää poistaa typerän kysymysmerkin nominien taivutuskaavan nominatiiviakkusatiivin perästä. --Hekaheka (talk) 12:54, 15 February 2015 (UTC)

Edelliseen vuodatukseen voisi vielä lisätä sen, että yhden ääntämyksen piirteen nostaminen taivutuskaavoihin vaikuttaa älyttömältä. Tätä ei tosin parane mainita CodeCatille, koska sitten sinne ilmestyvät esimerkiksi kysymysmerkit sellaisten sanojen kuin kuorma-auto tai vaa'an keskelle. --Hekaheka (talk) 13:07, 15 February 2015 (UTC)

Kas. Joo, tää on munkin mielestä kyllä paremmin ääntämysohjeisiin kuin taivutuskaavoihin sopiva asia, eteenkin kun kyse on perusmuodot poislukien täysin sijamuodosta eikä itse sanasta riippuvasta ilmiöstä, ja joissain muodoissa tosiaan vaihteluakin esiintyy (lähinnä kyllä omistusliitteissä, joista meillä ei edes ole taulukkoja). Toisaalta: meillä ei ole vielä ääntämystä tai edes sivua monille sellaisille peruspäätteille kuin -lle tai -utua, joten mistä loppukahdennustietojen pitäisi tällä hetkellä edes olla saatavilla?

Siitä olen kyllä vähintään samaa mieltä, että jos jokin tälläinen käytäntö luodaan, sen pitäisi olla selitettynäkin jossain. Appendix:Finnish pronunciation on tällä hetkellä, noh, aika ala-arvoisessa jamassa, eikä selitystä tarjoa myöskään Wiktionary:About Finnish.

Yhden "riehuva" on sitten tietysti toisen "rohkea". Voisin aloittaa tsekkaamalla, mitä mallineiden muutoshistoriassa & aikaisemmassa keskustelussanne todettiin, ja ehkä jatkamalla juttua siitä. --Tropylium (talk) 17:57, 16 February 2015 (UTC)

Tietääkseni missään kielessä ääntämistä ei merkitä taivutuskaavioihin, vaan "Pronunciation"-osaan, esim. -lle, ole, olko, liene, saa, ottaa. Lisäksi osa jäännöslopukkeista on tulkinnanvaraisia tai puuttuu kaavioista (-ksi, epäsäännölliset verbit). Voisin yrittää itse poistaa pikku äksät ja kysymysmerkin. Ei siihen tarvitse osata lua-kieltä, senkun lukee koodia rivi riviltä. Verbien koodissa on tosin 1432 riviä, mutta jos aiheutan sotkun, kumoan heti muokkaukseni. Haluaisin ensin varmistaa, että aiheesta on keskusteltu englanniksi. @Hekaheka: milloin ja missä olet keskustellut ylä-äksistä CodeCatin kanssa? Häntä voisi sen sijaan usuttaa muokkaamaan Appendix:Finnish pronunciation'ia tai Wiktionary:About Finnish'iä. --Makaokalani (talk) 10:58, 14 March 2015 (UTC)

Näyttää siltä, että sekoitan johonkin toiseen keskusteluun, koska löydän vain tämän: Thread:User_talk:CodeCat/sign_ˣ_in_IPA. Viidestä pointistani kuitenkin neljä on edelleen voimassa. Beer Parlour -keskustelu siis tarvitaan, jos asialle halutaan jotakin tehdä. Voin aloittaa sen, jos tiedän saavani tukea. --Hekaheka (talk) 07:14, 15 March 2015 (UTC)

Saat tukea minulta. --Makaokalani (talk) 09:44, 16 March 2015 (UTC)

Finnish nominal inflection types

Hekaheka suggested asking you about my question on her talk page. Can you help with this? I'm asking in part to help improve the current Appendix:Finnish nominal inflection, where I'm making a kind of "tree" of the different types so that it's easier for people to understand how they are related. Please answer here instead of on Hekaheka's talk page. —CodeCat 02:04, 19 February 2015 (UTC)

7 in Proto-Uralic

I tried to reconstruct 7 as *śäjćem, but I'm not sure if all the details add up. I don't know enough about the languages outside of Samic and Finnic to tell if this form fits the sound changes known for them. Could you have a look? —CodeCat 19:30, 2 March 2015 (UTC)

A+ for trying, though thus far no one has managed to reconstruct a coherent proto-form for this word. I know of some adjustments that would help, but they are so far unpublished. --Tropylium (talk) 23:11, 2 March 2015 (UTC)

I suspected it was that bad. Still, the forms are too similar for it to be a complete coincidence. I'd like to keep the page at least to list the descendant forms, but I don't know what to call the page. Something like *ś?ćem maybe, or some other kind of wildcard character? —CodeCat 23:31, 2 March 2015 (UTC)

@User:CodeCat I think you are on extremely thin ice, if not walking on the water. You are reconstructing words into a language you admit you know nothing of!!! Why don't you concentrate on something where you are on safer ground? --Hekaheka (talk) 01:27, 4 March 2015 (UTC)

Why do you think I asked someone who does know? Maybe you're the one who should stay on safer ground; you've clearly stated in the past that you have little interest or knowledge in linguistics and etymology, when I asked about things. —CodeCat 02:53, 4 March 2015 (UTC)

Võro orthography

I'm going through the few Võro entries we have and adding templates and a few basic inflectional forms to them. I'm using this dictionary as a reference, which is provided by the Võro language institute. But it's hard for me to understand everything as most information is written in Estonian or Võro, which I'm not exactly fluent in and I have to rely on an automatic translator. So I wonder if you could help me figure out some things?

Firstly, the dictionary indicates palatalisation with a following apostrophe, both in the dictionary entries and in the running text in the foreword. So this leads me to believe that the apostrophe is part of the standard orthography and not just a mark used in dictionaries. But the Võro Wikipedia doesn't seem to use it at all, so I'm wondering if it's actually used or if it's just a prescription that nobody follows. I've also seen sources that say palatalisation is indicated with an acute accent placed on top of the letter, so like ś, ń, t́ (or t´) etc. Of course that doesn't help to clarify the matter. So I'm wondering what representation should be used in Wiktionary entries.

What confused me even more is that in the dictionary, there is de'tsembri, with an apostrophe after a vowel. Vowels can't be palatalised, so I'm guessing that this is supposed to indicate the stress (the Estonian cognate detsember appears to have second-syllable stress, according to ÕS). But I wouldn't know how to verify it... and in any case it's strange that they use the same symbol in two meanings.

Then there's the letter y, which was introduced quite recently according to en.Wikipedia. The Võro Wikipedia has w:vro:Nõna for "nose", and our entry was formerly also spelled this way, but the dictionary indicates nyna so I moved it there. Presumably we should be using the new spelling, but it's strange if Võro Wikipedia is not using it, so again I wonder what is the general practice among Võro speakers. —CodeCat 01:50, 6 March 2015 (UTC)

I have no active familiarity with the development of the Võro literary language, really. You might have better luck looking up if we have anyone who speaks Võro. But the introduction to the Võro Instituut's Võro-Eesti synaraamat confirms some things:

• y and palatalization apostrophe are normative. No clear info on if it's supposed to be a plain apostrophe ' or perhaps a free-standing acute accent ´ (as in e.g. Skolt Sami).
• Apostrophes following vowels are not a part of the orthography and indeed mark stress, on the preceding syllable however.

--Tropylium (talk) 04:03, 6 March 2015 (UTC)

álom

Hi Tropylium, I've noticed your changes in the etymology. I am not a linguist, so I can rely only on the references I use. Both the Uralonet online database and my printed ety dictionary contain the *oδa-mɜ format. What was the reason you changed this to *adema? What would be the best way to keep the proto-language items consistent? Do you have a different source? I think it would be good to provide references when we can. Thanks. --Panda10 (talk) 14:54, 8 March 2015 (UTC)

Hello! Yes, the traditional reconstruction for the root "to sleep" (same as in alszik) is indeed *oda-. That this may have rather been *ade- is a recent proposal, first aired in 2013. The full details would probably be best discussed on the proto-root's appendix page, once one is around? Duplicating etymological references across a set of cognate words gets difficult to maintain quite fast. Since you've requested the refs, I guess I shall create those.

For future reference, you may wish to note that UraloNet mostly (entirely?) relies on matters of reconstruction on the Uralisches Etymologisches Wörterbuch, and since it was published in the late 80s, the system cannot take into account any newer research. Not all of the roughly contemporary research (late 70s on) was consistently accounted for in it, either. --Tropylium (talk) 05:04, 9 March 2015 (UTC)

Thank you. So what is your recommendation? To stop using Uralonet as a reference? Or to continue to use it until new research comes along? On the Appendix:Proto-Uralic/adema page one of the references is the Uralisches Etymologisches Wörterbuch, the other is just as old (1988). --Panda10 (talk) 14:43, 9 March 2015 (UTC)

It's certainly still usable, and perhaps even preferrable to UEW (since it can be accessed online). I just would not consider it the latest word on issues of reconstruction. After all we need references for other types of information as well, e.g. descendants and semantics. (Also on this matter make sure to see the bare root *ade-.) --Tropylium (talk) 14:56, 9 March 2015 (UTC)

Ok, thank you. This is very helpful. --Panda10 (talk) 15:22, 9 March 2015 (UTC)

j- and w- stems in Proto-Finnic

From what I've seen, there are stem-forming suffixes -j and -w in Proto-Uralic, or at least in some of the descendants, including Finnic. When preceded by a vowel, these commonly coalesce with it. This is how I believe it happens in Finnic (correct me if I'm wrong):

• -ew > -u
• -aw > -o
• -äw > ?
• -ej > -i
• -aj > -oi, -i
• -äj > -i

I noticed that in Finnish, the stems in -i have a plural in -ei, and it got me thinking about how the plural would have evolved. Clearly -ej- alone would have become -i-, but would the plural stem not have been -ejej-? It's conceivable that this became > -eji- > -ei- in Finnic. But if that's true, then what happened to the other possible stems in the plural? You'd expect -äjej-> -äji- > -äi-, -ajej- > -aji- > -ai-, thus implying that i-stems originally had several different subclasses based on the plural vowel. Of course analogical levelling eventually preserved only the -ej/-ejej- type in Finnish. I don't know what would happen to w-stems, though.

A lot depends on relative chronology. What cases like mätä + *-j → mäti or kota + *-j → koti show is that unstressed *-äj, *-aj had already become *-ej by Proto-Finnic. I don't know if these can be still distinguished from original *i-stems though, as there has been plenty of analogical levelling. What has survived at least are differing plural stems for a-stem words and ä/e-stem words in Votic (e.g. muna : munõi-, but lumi : lumi-). It might be possible to similarly reconstruct e.g. nom.sg. *koti *kotei : obl.sg. *kotei- : obl. pl. koteji-, but before having a couple different -j-derivatives of Proto-Finnic age and their reflexes in some key languages to compare with, I couldn't tell.

That's very interesting. That would mean that the plural of was still in Proto-Finnic, which was then backed to in the ancestor to Votic. But that, in turn, implies that the plural of j-stems can't also have been -ei-, because the diphthong is preserved there. It follows that i-stems must then indeed have had the plural in -eji- still, and the ji > i change happened in the dialectal period.

What is also interesting is what this implies for monosyllabic stems ending in -je-, like . If the ji > i is indeed dialectal Proto-Finnic, then it stands to reason that the Proto-Finnic form was still bisyllabic , stem , plural stem , assuming that ji > i and je > i happened concurrently. And if that in turn is true, then it has implications for the productivity of suffixal gradation, as a trisyllabic partitive would gradate to , while Finnish reflects with the gradation readjusted after the loss of the middle syllable. Thus it was still productive in the dialectal Proto-Finnic period.

That doesn't necessarily follow. E.g. the contraction of post-tonic *-Vji could be older than contraction under secondary stress, which the -i- we are assuming would have had. Or these diphthongs could have been formed primarily thru syncope rather than glide loss, as is suggested by examples involving *w, e.g. *käwe- > *käw- > käydä, *suwe- > *suw- > suu (followed by generalization from the oblique cases to the nominative, so e.g. *voji : *vojeta > *voji : *voita > *voi : voita).

That's a good point; syncope would have reduced the syllable anyway. But in the plural, it's not so clear whether contraction would have happened for . So perhaps the plural preserved the glide until much later, *vojei- > *voji- > *voi-. On the other hand, that would probably mean that in part of Finnic it would actually be *vojëi- with a back vowel, which would resist the ei > i change. I have no idea if such a form is attested.

However, if Proto-Finnic indeed still had , and southern Finnic evidence with õ(i) suggests that it did, then we can't get around the fact that there is a clear distinction in reflexes between -i/õ(i)- < -ei- and -ei- < -ejei-. These two sequences did not fall together. Thus, I don't think there is any other possible conclusion than that the plural stem of was not but must have been a longer sequence. What sequence that was, I don't know. —CodeCat 15:42, 9 March 2015 (UTC)

Your latter paragraph seems clear, yes. But to clarify: only earlier *-Ai- yields demonstrable late Proto-Finnic *-ei- (or *-ëi-). Original *-ei- was contracted to -i- early enough to trigger *t > *c (and hence e.g. vesi : vesiä; yet setä : setiä, not ˣsesiä). That is, there are two chronological changes from *ei to *i:

• older layer (Proto-Finnic): unstressed *ei > *i, and *ai, *äi > *ei
• newer layer (general only in parts of Northern Finnic, absent even from some Western Finnish dialects): secondary unstressed *ei > *i

The former oblique pl. stem of voi would have simplified to *-i- already in the 1st round, if it was still around by then. --Tropylium (talk) 05:56, 10 March 2015 (UTC)

I never even noticed that a/ä stems never have assibilation. Thank you for clarifying this.

Well, they do; cf. lentää : lensi, uurtaa : uursi. This has been considered analogical (Finnish dialects have also imperfects like lenti, uurti), though according to a recent proposal these would be rather due to an early vowel reduction in heavy stems: *lentäj- > *lentəi- > lenci- vs. *setäj- > *setei-

It's also possible that the opposite happened. a/ä stems might have had assibilation originally, which was levelled out in nouns and many verbs, but what kept it from being levelled out in e-nouns was the nominative singular, which has -si. —CodeCat 16:39, 10 March 2015 (UTC)

If you assumed *ai >> i universally & early enough to cause assibilation, you again have no way to explain why Votic would have the declension type munõi- or Estonian mune-, but neither has anything like ˣlumõi- / ˣlume- for 'snow', or ˣkõrvõi- / kõrve- for 'ear'. --Tropylium (talk) 10:37, 11 March 2015 (UTC)

In any case, if we assume ei > i stage 1 happened already, then the plural of was , and remained so at least until the second stage, which prevented it from undergoing it. However, that makes me think that the stem of was also still . Or is there a particular reason why the contraction > could have happened earlier than ei > i stage 2 and > ? —CodeCat 13:41, 10 March 2015 (UTC)

I notice that you propose *koti for the nominative singular. Is there a reason why it's not *kotei? Did ei > i happen word-finally before it happened medially? —CodeCat 13:48, 9 March 2015 (UTC)

My bad, that's inconsistent. Upon checking, e.g. Veps indeed still has some nominatives ending in -ei. --Tropylium (talk) 15:26, 9 March 2015 (UTC)

It does? Can you give examples? —CodeCat 15:42, 9 March 2015 (UTC)

First example I spotted was ńäńei corresponding to Finnish nänni (both < *nännei < *nännä-j). --Tropylium (talk) 05:56, 10 March 2015 (UTC)

This does make things trickier because it becomes nearly impossible to distinguish original -ej stems from original -aj/äj stems, unless there is a Veps attestation. Then again, the same happens between -o and -aj stems. —CodeCat 13:41, 10 March 2015 (UTC)

No? The above-mentioned Votic evidence still works, and so does e.g. dialectal Finnish evidence (say, *äitei 'mother' > Southern Ostrobothnia äitee, or similarly > kukkoo). Plus there is no such thing as entirely "original" -Vj-stems anyway; all cases are derivatives or loans, so generally it's possible to just compare with the underived root (e.g. in the previous case, nännä is attested dialectally in Finnish).

I guess if you want to be able to tell the stem type just from standard-language citation forms, that will be tricky (and often impossible), yes. --Tropylium (talk) 15:03, 10 March 2015 (UTC)

It's interesting that you picked äiti of all words. That's generally considered a Germanic loanword, but the Germanic source has long *ī, not *ei (which didn't exist in later Proto-Germanic). So it's strange that *-ei was used to substitute for this ending rather than *-i. —CodeCat 16:42, 10 March 2015 (UTC)

Also, I just checked and Veps has kodi, not *kodei. So if Veps does preserve -ei, I wonder why it was not preserved in this word. —CodeCat 23:43, 10 March 2015 (UTC)

If I had to guess, loaning as *äitei could have been to avoid an adaptation as **äici. As for *kotei, I'll probably need to look more into the Veps situation. There may be dialect differences involved, I know that e.g. several dialects shift *ei to /ii/ or /i/, some even in stressed syllables.--Tropylium (talk) 10:37, 11 March 2015 (UTC)

Estonian (both north and south) has several vowel changing plurals that don't seem to have any counterpart in Finnish, like a:õ (kubõl, kuld), o:õ (talo, kand, vari), o:a (talo, kand, vari), u:õ (kanarik), u:a (kanarik), ü:e (häbü), ü:ä (häbü) which are found in Võro. I wonder where they come from, they might be related to this for all I know. Am I on the right track here? —CodeCat 00:10, 9 March 2015 (UTC)

The "a-plurals" of Estonian involve mainly analogical processes (on the other hand at least your 1st example is a regular epenthesis development: *kupla > *kupl > kubõl). --Tropylium (talk) 06:02, 9 March 2015 (UTC)

*kupla > *kupl > kubõl is epenthetic, but what about the stems? Compare the allative forms: singular kublalõ, plural kublõlõ; there's an a > õ change here. Am I correct that the õ reflects Proto-Finnic , just like in the Votic example? That still leaves the others though. In particular häbü is puzzling with its ü > e change. —CodeCat 13:38, 9 March 2015 (UTC)

Yes, I believe the first reflects *-ëi-. Cases like häbü are probably similarly analogy; I would guess -e-/-õ- in these is probably a dialectal change from -ä-/-a-, which as said has been adopted as a general plural stem marker (though I am not sure where from). --Tropylium (talk) 15:26, 9 March 2015 (UTC)

---

Alright, I looked up some literature on koti : kotei-. This turns out to be entirely secondary rearrangement that has no repercussions for Proto-Finnic.

• As is shown by agent nouns like luke- → lukija, pese- → pesijä, teke- → tekijä, the later assimilation *ei > *i also included *ej > *i(j), and thus it's not possible to derive -ei- from *-eji-.
• This also means that it is not necessary to assign phonemic diphthong status for Proto-Finnic; *[ei] = */ej/ remains feasible, and indeed this better accords with how some Finnish dialects have stops in the weak grade before *-oi, *-ei.
• Instead, sometime after the shift of *ei to i, the now-awkward inflection type sg. *koti(-) : pl. *koti- was analogically extended to sg. koti(-) : kotii- in western Finnish dialects. (Eastern Finnish dialects have also fixed situations like this, but by introducing a new plural marker entirely: koti(-) : kotiloi-.)
• This new secondary unstressed ii was diphthongized (dissimilated?) to ei in Southwestern Finnish. The change was in older literary Finnish attested also in various other cases, e.g. *etsiisi > edzeis for etsisi (= etsi- + -isi) — but this went out of the fashion over the 19th century, and it has survived 'til modern Finnish only in the plural stems of i-stem nouns, where it fulfills the need of disambiguating from the singular stem.
• Loss of -h- in the illative was even later, and hence only forms like kotiin, kotihin have been attested, and not ˣkotein or ˣkotin.

--Tropylium (talk) 18:10, 11 March 2015 (UTC)

This doesn't explain why other Finnic languages (that I can see) seem to agree on the existence of -ei- in the plural of i-stem nominals. Such nominals appear in Estonian and Võro as well, while Votic has -ii- and I have no idea about Veps. It seems unlikely that Estonian, Võro and Finnish happened to innovate on the same -ii- > -ei- change separately, don't you think? Furthermore, I think it's a bit of a stretch to suggest that former -eji- is contracted straight to -i-. As you demonstrate with lukija, e > i before j as well, but then the result would still have only been -iji- which presumably contracted rather quickly to -ii-. So I think this -ii- would still need to be posited for Proto-Finnic in any case, not -i-. —CodeCat 18:32, 11 March 2015 (UTC)

For N/S Estonian it would seem to be simplest to root this in retention of original *-ei-, followed by various analogical developments, e.g. we need to assume word-final -ei > -i anyway, and this may have been generalized to the singular stem. Alternately, perhaps -e- for original *i-stems and *ei-stems has been taken from the a-stems. Votic -ii- meanwhile does seem to reflect the same kind of analogical extension as (dialectal) Finnish — but the shift *-ei > *-i probably went thru *-ii, so it does not seem possible to assume that this was Proto-Finnic, as much as later analogy.

Also I don't think we have established that there ever was an *-eji-. Symbol algebra starting from *A+j+j might suggest such a thing, but I haven't found much evidence that any of the *-ei-stem words existed that early. So it's kind of like asking "what was the Old English plural of moose, if not meese?" The two most widespread examples seem to be risti and pappi (both found everywhere in Finnic) which we can per the semantics date around 800 CE, several centuries later than Proto-Finnic. (The actual suffix is ancient, but it may have only emerged as a derivative element from attributive forms such as lehmi-.) --Tropylium (talk) 20:31, 11 March 2015 (UTC)

The trouble is that at least some cases of the derivational -j end up not as -ei, but as -oi, which then becomes -o in Finnish. Samic also has some nouns with -ōj I believe, I don't know if it's the same suffix but it might well be. If they're the same suffix, then it must have existed in Proto-Finnic in some form.

I don't know of the particular semantics associated with the -j suffix anyway. The vast majority of the words using it are loanwords where the vowel is just a filler. But at the same time, the declension pattern of i-stems with their -ei- plural can't have come out of nowhere, there must have been some motivation. Something about it must have had existing semantics that made it particularly suitable for this role. After all, why did they make a whole declension pattern up out of thin air instead of using an existing one? —CodeCat 21:07, 11 March 2015 (UTC)

Well, there are attributive cases with *-oi all the same (say, siko-). And plenty of cases with the "wrong" suffix, including front-harmonic cases like → > eno/onu. If there ever was a time when the derivative suffixes based on *-j had a 100% phonologically conditioned distribution, it was over by late Proto-Finnic.

Samic *-ōj mostly corresponds to Finnic *-o or *-ü, not *-oi, and is in most cases a root formant rather than a suffix. Its history has not been worked out to anyone's satisfaction. --Tropylium (talk) 07:37, 12 March 2015 (UTC)

Looking at the inflection generated by the template on , I noticed that the singular and plural stems ended up the same. Whether this is correct or not I don't know, but it does seem plausible that if they were, the plural stem had its -i- restored in some dialects to avoid homonymy. This probably happened in Finnish, although Finnish also has some other surviving cases of -oi- like in the -oida type verbs. But it didn't happen in both kinds of Estonian, which completely disallow unstressed diphthongs now and have, as a result, started to create new plural forms based on the genitive plural form, presumably because the singular forms are likewise based on the genitive singular.

In any case, if -oi- was indeed analogically restored, then the same would presumably happen to -ii-. In cases ending in an i-diphthong, like , this could not take place as the sequence -oii- was not allowed. But it would still be a bit strange if -ii- ended up as -ei-. On the other hand, this is presumably still in the dialectal Proto-Finnic period, and long vowels were probably still not allowed in unstressed syllables. So that's the only way I could think of that -ii- could be turned into -ei-. But if this change did happen, it must have happened early enough to affect Estonian before diphthongs became disallowed, as it was early enough to be included in the ei > e and õi > õ changes that took place. —CodeCat 18:54, 12 March 2015 (UTC)

The introduction of this *-ii- would have been in the dialectal Finnish period, around the Middle Ages at earliest, since this analogical development postdates *ei > *i, and as you say, *oi > *o, which (to reiterate once again) did not even occur in various Finnic 'lects. If you're unclear on the reason for the diphthongization, the main point is that *ii > *ei was a sound change, not any kind of analogy. (Cf. the English Great Vowel Shift: *ī > *əi > /aɪ/.)

We also still have no reason to think any of this ever happened in Estonian, since, again, it can have just retained original *-ei-, and generalized from there to *-i-stems. If anything, the SW Fi sound change could have even been triggered due to medieval influence from Estonian, with which the dialect group shares some other developments as well. --Tropylium (talk) 19:24, 12 March 2015 (UTC)

Does that mean that Estonian preserves the distinction between -ei and -i stems, which have fallen together in Finnish? There would have to be a class of -e/õ nouns then, but I can't say I've seen those yet. Or maybe just didn't recognise them. —CodeCat 20:55, 12 March 2015 (UTC)

kymmenys

Are you sure this is not just kymmen + -us or contraction of *kymmenennys (from kymmenent- + -us)? —CodeCat 02:28, 29 March 2015 (UTC)

The latter is exactly where the mentioned suffix -nnys comes from, i.e. -nte- + -ys; I guess it's a matter of taste if we want to analyze it as a separate suffix. Hakulinen appears to think that it's the ordinal suffix + -(U)s combo that specifically denotes a fraction, though equivalently we could add a new sense + usage note for -us. --Tropylium (talk) 02:58, 29 March 2015 (UTC)

Well, the suffix -us forms abstract nouns from adjectives generally, and ordinals are adjectives too. So I don't think there is really a need for any separate suffix, the semantics and morphology of ordinal + -us seem quite transparent to me. —CodeCat 03:02, 29 March 2015 (UTC)

Not quite, this is the vastaus-type -us in use here, and then in the standard Finnish fraction names like kolmannes we have just a plain -s (: -kse-).

Though the last-mentioned word might be a good case for not positing separate suffixes anyway: if this were analyzed as kolme + suffix, there is no reason to expect the irregular -a- of kolmas to be duplicated. --Tropylium (talk) 03:12, 29 March 2015 (UTC)

Appendix:Proto-Finnic/hibus

This is a noun of the "vastaus" type, but I haven't seen those before in Proto-Finnic. Is the suffix -ks- or -kc-? I can't tell from the descendants. —CodeCat 17:51, 31 March 2015 (UTC)

It's *ks. A fairly common suffix really, but we've so far been mainly adding root words. --Tropylium (talk) 21:01, 31 March 2015 (UTC)

Thank you. In the meantime, I added a second one, . It appears that some descendants have -e- instead of -i-, I'm not sure why that is but they surely must be related. —CodeCat 21:18, 31 March 2015 (UTC)

teräs

I don't think that Finnic people would have known about steel... —CodeCat 19:28, 8 April 2015 (UTC)

Why not? Steel has been known, if rarely used, already since antiquity (similarly, steel also derives already from Proto-Germanic). I don't know about the archeological record, but I'd guess it was originally an imported material known mainly from its use in swords (hence the derivation). --Tropylium (talk) 19:37, 8 April 2015 (UTC)

Appendix:Proto-Slavic/dikъ

This was just created. Could this or a related term possibly be the source of ? —CodeCat 12:48, 9 April 2015 (UTC)

Looks promising. Slavic loans proper are too young to have participated in *ti > *ci (e.g. Finnish dialectal kaatiot (“underwear”) is from the same source as gaće), but if there's been a Balto-Slavic root with this approximate meaning around, it could well have made its way to Finnic at an early date. Kaczyńska's source paper indeed even mentions something about Finnish sika, but I can't make head or tails of Polish though. --Tropylium (talk) 03:07, 13 April 2015 (UTC)

Category:Proto-Finnic terms derived from Proto-Baltic

Re: Appendix:Proto-Finnic/rakja. Did you really mean to add this category, or did you just forget to put a "-" in the etyl template? Chuck Entz (talk) 19:39, 23 May 2015 (UTC)

Yes, it's a category we're going to need (unless we want to start enforcing the "Baltic = Balto-Slavic" approach). Liukkonen's etymology is considered less well-established though, so it's possible we should leave {{etyl}} out entirely. --Tropylium (talk) 16:41, 24 May 2015 (UTC)

I thought we were enforcing it, though I may be remembering the consensus rather than an actual decision. While it's true that Proto-Balto-Slavic is different from Proto-Baltic in that they're theoretical constructs based on differently-circumscribed datasets, the prevailing consensus seems to be that the reality they're designed to model is the same for both. If we're going to use a different name for different understandings of the same thing, then we should have different names for Indo-European depending on whether Celtic, Albanian, Armenian, Tocharian, Hittite, Mycenaean Greek, etc. were included at the time, not to mention the presence of theoretical advances such as laryngeal theory, recognition of borrowing between various Uralic and Indo-European proto-languages, areal diffusion among dialects instead of binary branching, etc. Chuck Entz (talk) 21:53, 24 May 2015 (UTC)

OK. The latest understanding is anyway that the Baltic material in Finnic is substratal loans from an extinct "Northern Baltic", rather than from the ancestor of Lithuanian and Latvian. So I take it we are using the term "Balto-Slavic"? --Tropylium (talk) 00:10, 25 May 2015 (UTC)

vajáldahttit

This word does not appear exactly in your Proto-Samic list, but there's one very similar form that does, *oajāltëttētēk. Note the oa ~ va discrepancy. Is this a normal change? —CodeCat 21:58, 2 September 2015 (UTC)

Spotted another: vuoitit vs. *oajtētēk. —CodeCat 22:00, 2 September 2015 (UTC)

Lehtiranta lists the former as indeed a descendant, but yeah, it looks like some kind of an irregular change has happened there. The second word has evidently just been replaced with a different loan from Finnish voittaa in Northern Sami (ditto for Inari Sami vyeittiđ, Lule Sami vuojttet, all three suggesting rather pseudo-PS *vuojtē-).

(The alleged "*oajtē-" is a Finnic loanword as well, though; so I would ask if the more western and eastern varieties could just have loaned the word equally late, in shapes that merely coincidentally suggest a single common Proto-Samic form.) --Tropylium (talk) 06:10, 3 September 2015 (UTC)

alasti

The reconstructed form should be shouldn't it? The suffix has a stem in . —CodeCat 16:11, 6 September 2015 (UTC)

Ah, right, yes. Although I wonder if it would be more convenient for us to reconstruct *st and not *ct in these cases. I don't think anyone has stated anything explicit about the dating of this change in the literature (yet), but the similar changes *kst > *st and *pst > *st are usually considered already Proto-Finnic. --Tropylium (talk) 16:28, 6 September 2015 (UTC)

sorvâ and hirvi

These look like they are related, but the vowels don't match: Inari Sami seems to indicate *šurve, while Finnish has *širve. Any ideas? —CodeCat 18:16, 30 September 2015 (UTC)

Inari Sami o-â is the regular reflex of Proto-Samic *ë-ë (see also e.g.: moonnâđ (“to go”) < *mënëtēk < *mene-, nommâ (“name”) < *nëmë < *nime); from PS *o-ë you get instead u-â (as in e.g. tullâ (“fire”) < *tolë < *tule). --Tropylium (talk) 19:13, 30 September 2015 (UTC)

Inari Sami phonology is so annoying... Thank you! —CodeCat 19:16, 30 September 2015 (UTC)

The -ta adjective suffix

Both Finnish and Northern Sami have a fairly large group of adjectives with a suffix that goes back to *-(e)ta. But the suffix doesn't seem to have any meaning, it's just present on adjectives. Do you know more about this suffix, and do you perhaps also know whether the -e- was part of the suffix or not? Both languages show a curious absence of *-ata adjectives. —CodeCat 15:41, 26 October 2015 (UTC)

Yes, it's generally analyzed as *-eta, and it's noted to be more of an adjective marker than a derivational suffix (though there are various cases where a loan adjective has been "nativized" by switching to this ending: consider e.g. Fi. kalpea ~ kalvas from *h₂élbʰos). Cognates are found all across Uralic, mostly in the west. I'm not sure if you're asking for anything else in particular here though. In case you want to look into the topic in more detail, Ilona Rauhala's recent PhD on adjectives in Uralic (Uralilaiset adjektiivit: Sanaluokan historian hahmottelua) should make a good starting point.

If you want to get a PU entry *-eta started, it will be certainly possible to flesh it out in more detail later. --Tropylium (talk) 00:01, 27 October 2015 (UTC)

You told me what I wanted to know, even if you didn't realise. :) I was curious about the shape (e or not), function/meaning, and distribution. You answered all of that! I'll be creating entries for Uralic, Samic and Finnic now. —CodeCat 00:05, 27 October 2015 (UTC)

*loppu vs *lop'ettadak

The words and are clearly related, but they have different stem vowels. I think that both may be derived from a hypothetical , but this doesn't seem attested anywhere. What can you make of this? —CodeCat 19:22, 10 November 2015 (UTC)

Yes, that would be the obvious analysis. The assumption seems to carry out, too: SSA reports a basic noun loppi : loppe- 'end, point' as attested dialectally from Tavastian Finnish (in standard Finnish its genitive lopen has been fossilized as an adverb), and a coordinate basic verb loppe-/lõppe- 'to end' from Karelian, Ingrian and Estonian (and perhaps Livonian, but loppõ could be just as well from *loppu-). --Tropylium (talk) 20:25, 10 November 2015 (UTC)

-taa and -ttaa

Is there a relation between these? And which one was used to derive täyttää? I would think that a two-consonant suffix would necessitate a prop vowel (täydettää?) so I would go for the former. —CodeCat 00:01, 11 November 2015 (UTC)

The latter suffix goes partly back already to a separate Proto-Uralic suffix *-kta- or *-pta- (which is probably but not demonstrably derived from *-ta-), but partly also to *-t-ta-, as in e.g. herätä → herättää. So yes, täyttää is täyt-tä-. --Tropylium (talk) 11:50, 13 November 2015 (UTC)

The "suppletive" stems in Finnish

There are a few verb types in Finnish that seem to have a stem formation that is suppletive in some way, using different stem formants in different principal parts. I wonder if you know anything more about the origin of these.

• The tupakoida type has -oi- but some principal parts allow an optional extension to -oitse-. This formation seems relatively new, so I'm thinking of an analogical source here?
• The valita has a stem in -tse-, but in environments where syncope of the stem vowel occurred, this has been changed to -t-. What I know about Finnic and Finnish behaviour of -cc- would lead me to think that contraction would lead to -ct- giving Finnish -st-, just like in the veitsi-type nominals. But apparently something else happened here?
• vanheta is another mystery to me, I can't explain the -n-/-t- alternation. It does however appear to have a cognate in Northern Sami -nit though.

—CodeCat 19:12, 16 November 2015 (UTC)

Proto-Finnic -odak vs. -oidak

I noticed in your page on Finnish inherited vocabulary that you reconstructed a few verbs with an -oidak ending where Finnish has simply -oa. I'm guessing that the change is the same as the one that turned into -ton.

1. Is there a functional distinction between these suffixes?
2. If they are distinct, what is the origin of each one?
3. How do you know to reconstruct one or the other? I noticed Veps indeed has some -oida verbs corresponding to Finnish -oa verbs, but what if no Veps descendant is known?
4. How did the modern Finnish -oida suffix come to be, if it didn't originate in this suffix? Veps also has this one, but the two fall together. Note also my question above about the -tse- suffix in some of these, which I found to occur in Veps too.

—CodeCat 20:19, 17 November 2015 (UTC)

They're functionally distinct, yes:

• *-o- is even in Proto-Finnic only a marginally productive suffix (probably best treated as fossilized in most modern Finnic languages), probably from something like earlier *-w-, that appears mainly in intransitive verbs such as punoa. (Some *o-stem verbs are also loanwords rather than derived.)
• *-oi- is relatively productive in deriving transitive mostly instrumental verbs like vihtoa (← vihta), from earlier *-j-, and just as happens with the oblique plural and past tense markers -(o)i-, it is originally an allomorph of *-i- (as in luuta → luutia).
• -oida- though is IIRC in origin also simply an allomorph of *-oi- (so these never "fell together" in Veps): in numerous Finnish dialects *oi > *o only happened in unstressed syllables, while forms like satuloida would have retained it under secondary stress (as well as the -d-). If you look at the cases, you can notice that -oida never appears in bisyllabic roots, while -oa appears only in them.

Unstressed *oi and *o can be directly distinguished also e.g. in Ludian, in Southern Ostrobothnian Finnish (as oo vs. o), and in some cases in Livvi; indirectly also in Northern Karelian when followed by /n l s/ (as oń oľ os vs. on ol oš). For the most part though I've been following simply morphology: all transitive verbs transparently derived from nouns with the vocalism a-a, e-a, i-a are *-oi-.

As long as I'm on this topic, worth noting is that there is also a distinction between verbs ending in PF *-AdA- : *-At'Ak (e.g. pelätä) and PF *-AidA- : *-Ait'Ak (e.g. avata). --Tropylium (talk) 21:39, 17 November 2015 (UTC)

That does confirm my suspicions, that -oi- alternates with -ei- in Proto-Finnic even in these verbs. I had noted the lack of any -öä verbs in Finnish (well, I found one). But I am guessing that the frequentative -ia has another origin?

Sadly, I don't have access to sources for those other Finnic languages, so they aren't much help to me. Veps is all I have so far, and probably lots of rarer verbs aren't included.

I have avoided dealing with the -t'ak verbs mostly for now, because they seem to be a bit of a mess, with several formations coinciding, especially in Estonian. That said, I wonder why hasn't changed -ai- to -oi-, like the nouns did? Also, I had presumed that auki and avata derived from the same stem (-auga- > -aua- > -ava-) but I guess they're distinct. —CodeCat 22:42, 17 November 2015 (UTC)

I just noticed , which doesn't seem to derive from an a-stem noun (at least I can't find one). Is it derived from , rather than the other way around? —CodeCat 23:01, 17 November 2015 (UTC)

auki does come from the same root as ava- (ultimately PU *aŋa), but the *k has normally been considered part of a fossilized suffix of some sort.

The direction of derivation for tahto/tahtoa is per the sources I have around apparently so far unknown, but ultimately both would seem to be derived from a root *tahta- as indicated by Estonian (and, per SSA, apparently a single 17th-century Finnish source). The Samic cognate *tuostōtēk (“to catch”) however shows that the verb is originally an *o-stem, not an *oi-stem; making the Veps form looks like a secondary derivative *tahto-i-. I might need to look up Finnish dialect variants to be sure though. --Tropylium (talk) 23:45, 17 November 2015 (UTC)

Does Finnish distinguish final -iV from -ijV in pronunciation?

Does kulkija have the same ending, when pronounced, as sallia? Or is the -j- actually pronounced as a separate consonant? If they are the same, can I assume that j is silent between i and a vowel? —CodeCat 21:16, 18 November 2015 (UTC)

No difference. The orthography is not fully inconsistent on which to use in which forms, though. --Tropylium (talk) 09:17, 19 November 2015 (UTC)

In colloquial language sallia is pronounced as sallii, and kulkija is pronounced as kulkia. --Muhaha (talk) 21:27, 8 July 2016 (UTC)

Which is probably analogical after the contraction of -eA and -OA (sylkeä > sylkee, potea > potee, rikkoa > rikkoo, säilöä > säilöö), given how uninflected examples of -ia, -iä and similarly -ua, -yä are unaffected (no one calls Laihia "Laihii" or Lapua "Lapuu"). --Tropylium (talk) 23:36, 8 July 2016 (UTC)

-us and -os (-kse-)

Could you elaborate more on the (pre-) Proto-Finnic origins of these suffixes? Did the -us and -os variants already exist then? What happened to the plain -s version, was that still productive in Proto-Finnic? And Finnish? We currently have no entry for this sense of -s, even though the etymology of -us refers to it. —CodeCat 21:58, 3 January 2016 (UTC)

-us/-os were definitely in existence in Proto-Finnic, as they have exact cognates at least as far away as Mordvinic (the classic example is punos ~ Erzya понавкс (ponavks) < *punaw(e)kse, from *puna-). My current working list seems to have two examples: , . -s < *-kse is more widespread in Uralic yet, and it has been productive at least up to older Finnish (though I can't think of any particularly modern coinages with it).

I suspect we may have to continue datamining and sorting derivatives in the modern languages to have a better view of various suffixes' status in Proto-Finnic. There has not been a lot of explicit research on what the derivation system in those times exactly looked like. E.g. {{R:fi:SKRK}} normally just notes if a given suffix is e.g. common Finnic, but does not give any cognates, or list which uses or which words in particular can be considered to be a part of the oldest layer. --Tropylium (talk)

What was the original meaning of then? Did it always mean "red", or are there actually two different roots here? I noticed that your Samic list has both *ponētēk and *pońëtēk, which seem like they might be related at some level. —CodeCat 16:00, 14 January 2016 (UTC)

As it already says in there, PU *puna the noun is reconstructed as meaning 'hair; color', with the development to 'red' particular to Finnic.

'Braid' however derives from the verb root *puna- (“to braid”) (hyphen now added above), which appears to have been distinct already in PU (cf. Hung. fon). Samic *pońë- is from an also distinct *puńe- (“to twist”). These three roots (stems?) certainly appear to related to each other (and also to PIE *(s)penH-, as in pinti, *spinnaną!), but the derivation processes involved remain poorly known. --Tropylium (talk) 18:25, 14 January 2016 (UTC)

Meaning of "inherited" in linguistics

What, according to you, does "inherited" mean in linguistics? I think you'll have a hard time coming up with a definition that doesn't apply to the Hebrew component of Yiddish. Be prepared for my critical analysis of anything you say. --WikiTiki89 19:15, 9 February 2016 (UTC)

"Inherited" means descended through direct transmission, parent-to-child, from an ancestral language. There are two issues here: (1) substrate influence from speakers who switched from Hebrew to Yiddish; (2) later borrowings from Hebrew spoken as a second language by Yiddish speakers. #2 is straightforwardly a case of borrowing, not inheritance, by the normal definition of borrowing in linguistics. For #1 it could be argued that substrate influence is "inheritance" because the substrate language is an "ancestral language" but that's not normally how either of these terms are defined; rather, it would be considered a case of language transfer. Furthermore, it's probable that the mother tongue of Yiddish speakers prior to their speaking Yiddish wasn't Hebrew at all, but was probably Old French, and the mother tongue of those speakers might have been Aramaic, again not Hebrew. So it's doubtful there's very much Hebrew influence in Yiddish that's attributable to substrate influence, if any. Benwing2 (talk) 22:31, 9 February 2016 (UTC)

Note BTW that English has a much greater claim to substrate influence from Old French than Yiddish has from Hebrew (although in this case it's termed a "superstrate" rather than "substrate"), yet no one claims that French terms in English are inherited terms or that Old French is ancestral to English. Benwing2 (talk) 22:33, 9 February 2016 (UTC)

No one switched from Hebrew to Yiddish. Hebrew had basically stopped being a vernacular language during the Babylonian Exile; after that, Jews in Palestine spoke Aramaic, and in the diaspora they spoke the local language of wherever they were living. Jews arrived in Germany from Italy in the 4th century at the latest; they were presumably speakers of Vulgar Latin before they switched to various German dialects. Of course they still used Hebrew as a liturgical and literary language, but Hebrew hadn't been anyone's mame-loshn in over 1000 years by the time Yiddish became a separate language. —Aɴɢʀ (talk) 22:41, 9 February 2016 (UTC)

(e/c) So the thing about Yiddish, is that much of the Hebrew component does in fact come from #1. Much of it also comes from #2, and much of it is also indeterminable whether it comes from #1 or #2. Secondly, I am aware that the previous mother tongue of Yiddish speakers was in fact a Judeo-Romance language (not necessarily only Old French), and prior to that, likely Judeo-Greek, and prior to that Judeo-Aramaic, and prior to that finally Hebrew. But the thing is that the Hebrew component traveled through all these intermediate languages. Anyway, when you say "but that's not normally how either of these terms are defined", there is nothing in the definition that you gave that excludes substrate/superstrate influence. --WikiTiki89 22:52, 9 February 2016 (UTC)

I doubt that very much Hebrew in Yiddish is due to pure substrate influence through many languages due to the way languages continually throw out old content and accept new stuff; but the point is that substrate influence isn't considered inheritance in linguistics since substrate languages aren't considered ancestral languages. Benwing2 (talk) 22:55, 9 February 2016 (UTC)

Then define "ancestral language". --WikiTiki89 22:56, 9 February 2016 (UTC)

Not interested in an argument over semantics like this. Perhaps someone else will respond. Benwing2 (talk) 22:58, 9 February 2016 (UTC)

Like I said: "Be prepared for my critical analysis of anything you say." Without a definition of "ancestral language", you cannot argue that Hebrew is or isn't one. --WikiTiki89 23:04, 9 February 2016 (UTC)

You sure like to get the last word. The point is here that linguistics doesn't normally define ancestral language to include substrate languages. You seem to want to refuse to acknowledge this. Benwing2 (talk) 23:09, 9 February 2016 (UTC)

BTW I'm sure if you wanted to, you could come up with a zillion reasons why defining ancestral languages to include substrate languages leads to massive problems in quantifying what's inherited from what. Benwing2 (talk) 23:10, 9 February 2016 (UTC)

I'm not trying to come up with reasons to change the definition, I'm just trying to get a definition. The cause for most arguments is that the parties' definitions don't match up. Once we have laid out the definitions, then it becomes a matter of finding facts to back up whether something fits the definition. I'd much rather argue about the facts, but I cannot do that until I have a definition. --WikiTiki89 23:15, 9 February 2016 (UTC)

How about this: Language A is ancestral to language B if there's a direct chain of transmission from parents to children, where in each generation the parent's language is substantially the same as the child's language. This obviously excludes language replacement such as happens with substrate languages. Benwing2 (talk) 23:40, 9 February 2016 (UTC)

This is pretty much it. "Inheritance" is material being derived from an ancestor; an ancestor is a language which has gradually transformed into the language in question. (We don't have to specify "parent-to-child" in particular — in principle a language can persist indefinitely while having only second-language speakers, as might be said to be the case e.g. for Ecclesiastic Latin.)

For Hebrew to be an ancestor of Yiddish, we'd have to show the existence of a variety of Hebrew that was evolving "towards" Yiddish at the same time as its High German ancestor was. This is also theoretically possible — though all anywhere near securely known cases involve not the mixing of effectively unrelated languages, but rather fairly small shifts among related dialects. --Tropylium (talk) 02:31, 10 February 2016 (UTC)

It seems to me that the Hebrew component in Yiddish contains both inherited and borrowed elements, and while those inherited elements may not necessarily make Hebrew as a language an ancestor of Yiddish as a language, some subset of the Hebrew lexicon is the ancestor of a subset of the Yiddish lexicon by way of inheritance, not borrowing. Some such words are probably not only those with religious/cultural meaning, but I believe also words like פּנים and קול that were adapted so long ago that they only have Germanic plurals. I prefer not to use {{inh}} nor {{bor}} for such words, however, as the complexities lie deeper than those templates and their categories can handle. —Μετάknowledge^{discuss/deeds} 04:55, 10 February 2016 (UTC)

In the linguistic sense these are still borrowings. The fact that they have Germanic plurals doesn't change this. Note that English is full of Latin words, many of which were borrowed hundreds of years ago and have English plurals rather than Latin plurals, but are still borrowed from Latin (sometimes by way of French, i.e. borrowed from French which borrowed them in turn from Latin). Benwing2 (talk) 05:11, 10 February 2016 (UTC)

The difference is that the Anglo-Saxons hadn't been going through different languages and carrying some vocabulary with them as they did so. First they didn't use Latin words, then they began to; the case of Yiddish is that the speakers' ancestors used these words even before they adopted Germanic. —Μετάknowledge^{discuss/deeds} 05:36, 10 February 2016 (UTC)

Maybe. But what you're essentially asserting is that words that derive from a substrate or superstrate are "inherited"; this isn't the normal sense of the word in linguistics. For example, many English words were carried over from French speakers as they switched from French to English; but we still say they're borrowed from French, not inherited. Benwing2 (talk) 05:57, 10 February 2016 (UTC)

But those words were then disseminated into the general English-speaking population that had never spoken French; that is what, in my opinion, makes those words borrowings. --WikiTiki89 06:02, 10 February 2016 (UTC)

@Tropylium and Benwing2: Although it is a somewhat radical view, Paul Wexler proposes exactly that: that Judeo-Slavic (specifically Sorbian) was gradually relexified through contact with some variety of High German until it became Yiddish. Now, like I said that's a radical view, but it does make me wonder one thing: If Wexler's theory were true, would you then be prepared to say that that variety of High German is not an ancestor language of Yiddish?

Anyway, the situation gets more complicated when you consider the actual role of Hebrew throughout the whole process. Unlike what a few people have stated above, the pre-Yiddish Jews did in fact speak Hebrew. Even if it wasn't the everyday-matters language, it was still a literary and liturgical language (nevertheless used every day), and unlike most literary or liturgical languages you may be familiar with, every single Yiddish-speaking and pre-Yiddish Jew was familiar with Hebrew (i.e. it was not just the educated elite). Now I could suggest two extreme views of this situation: (1) that Ashkenazi Hebrew and the Germanic component of Yiddish were entirely separate languages and any instance of a Hebrew-origin word in Yiddish is actually an instance of code-switching, or (2) that Yiddish contained the entirety of the Ashkenazi Hebrew language within itself and so Ashkenazi Hebrew was simply a register of Yiddish. In truth, I actually quite favor the latter of those options, although its wording is a bit extreme, but that's beside point. The point is that words in Ashkenazi Hebrew were most certainly inherited from previous stages of Hebrew, words in the Germanic component of Yiddish were certainly inherited from some variety of High German, and that speakers of Yiddish were bilingual and passed down both the Germanic component and Ashkenazi Hebrew from generation to generation. These two languages are impossible to disentangle. I find it absurd that when a Yiddish speaker says a sentence with Hebrew grammar, the words in it are inherited, but when the speaker says the same sentence with Germanic grammar using one of the same words, somehow this same word is now no longer inherited but borrowed, even though this word had evolved from the very beginning of the history of Yiddish together with both Ashkenazi Hebrew and the Yiddish language, and had always been used in Yiddish and had never been a separate word from the one used in Ashkenazi Hebrew. --WikiTiki89 06:02, 10 February 2016 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ I don't believe what Wexler thinks, and I doubt very many linguists do. If Sorbian was relexified as High German you'd expect Yiddish to have a very strange and highly Slavic-influenced grammar and I seriously doubt that. I don't know that much of Yiddish but from what I gather it's hardly an unusual Germanic language. Anyway, Wikitiki, IMO the problem with your views as you've expressed them above is that the situation of having a population that speaks language A but is bilingual in language B and hence borrows words and phonology from that other language is far from unusual. You're making it out as if Yiddish is this super-exceptional, special case, when in fact there are lots and lots of similar contact situations. Benwing2 (talk) 06:13, 10 February 2016 (UTC)

Didn't I say that Wexler's view is radical? I don't agree with either, that was beside the point. It was an example of a possibility, and then a hypothetical question, which you didn't answer. Yiddish is not an unusual Germanic language only if you ignore the non-Germanic components, in which case, you no longer have Yiddish, nor even a fully functional language. As for similar contact situations, I'm not saying that Yiddish is super-exceptional (it is a unique situation in many ways, but then so is every language contact situation), but it is quite different from the one example you gave of English and Norman. And I'm not saying that only Yiddish should be allowed to have more than one ancestor, I am only making the case for Yiddish because I happen to know quite a bit about Yiddish, but if my arguments also apply to other contact languages, then all the better. --WikiTiki89 06:40, 10 February 2016 (UTC)

Sorry, missed your hypothetical, but if it was truly relexified I might say it was a mixed language with two parents, or something. But I still think you're falling prey to the common tendency to assume that there must be something universally special about what you're familiar with. I know this tendency -- my mother's first language is Hebrew and so my natural instinct is to think Hebrew is a unique case. (In some ways it is, as the only resurrected dead language; but as an actual language it's not exceptional, and I don't at all subscribe to the view [is that Wexler's again?] that Hebrew is relexified Yiddish or something like that. IMO the influence of Yiddish on Hebrew is in fact not significantly different from other cases of language shift, and there are parallels for the large-scale adoption of a literary language as a spoken one; modern standard Italian and German are good examples.) Benwing2 (talk) 06:50, 10 February 2016 (UTC)

"I find it absurd that when a Yiddish speaker says a sentence with Hebrew grammar, the words in it are inherited, but when the speaker says the same sentence with Germanic grammar using one of the same words, somehow this same word is now no longer inherited but borrowed, even though this word had evolved from the very beginning of the history of Yiddish together with both Ashkenazi Hebrew and the Yiddish language, and had always been used in Yiddish and had never been a separate word from the one used in Ashkenazi Hebrew."

The words of Hebrew origin weren't "always" in Yiddish. There would have been a time early in the history of Yiddish when had only just started to gradually evolve towards its later state, through (among other things) the acquisition of Hebrew loanwords. We label Yiddish as a natural, non-creole language precisely because we think it emerged gradually, not through abrupt mixing of German and Hebrew.

I'm also skeptical about the claim of Hebrew words in Yiddish "having never been separate". Some have perhaps remained in unison, but as far as I know, Yiddish is phonologically a fairly average High German variety, and thus sounds more typical of Semitic such as ḥ would have been nativized — and, after that, propagated thru inheritance within Yiddish, and not thru continuous maintenance from Hebrew.

"every single Yiddish-speaking and pre-Yiddish Jew was familiar with Hebrew"

Including also two-year-old toddlers only just getting a hang of Yiddish in the first place?

And finally, yes: if Wexler's proposal were right, Yiddish would have to be a Slavic language and it would have no inherited Germanic vocabulary at all. The proposal, however, is wrong, and is based on a conceptual confusion very similar to yours. --Tropylium (talk) 15:20, 10 February 2016 (UTC)

For convenience, I'm going to respond in reverse order:

@Tropylium said: The proposal, however, is wrong, and is based on a conceptual confusion very similar to yours.

I hate to defend things I disagree with, but it's a little harsh to call a professional linguist's theory simply "wrong". And the blog post you link to makes absolutely no substantial arguments. I have not actually read Wexler's own explanations of his theory, but I would presume it is a theoretically possible theory, however unlikely it is to be true.

@Tropylium said: Including also two-year-old toddlers only just getting a hang of Yiddish in the first place?

Well, yes. Jewish tradition (at least in the time and place in question) was to teach children a basic set of Hebrew prayers, including the first line of the Shema, as soon as they are able to speak. Even before that, they would be exposed to their parents' daily prayers in Hebrew. Boys were sent to schools called cheders at around the age of five (according to Wikipedia, but I have heard even three), where they learn to read and write in Hebrew and then immediately begin studying the Torah with Rashi's commentary. Girls in most cases lacked this formal education, but still learned informally.

@Tropylium said: There would have been a time early in the history of Yiddish when had only just started to gradually evolve towards its later state, through (among other things) the acquisition of Hebrew loanwords. We label Yiddish as a natural, non-creole language precisely because we think it emerged gradually, not through abrupt mixing of German and Hebrew.

Here's an overview of Yiddish linguists' mainstream view of how Yiddish formed (this is essentially based on Max Weinreich's view; variations on the theory generally have to do with the location in which this took place and the nature of the pre-Yiddish languages the Jews spoke, but the main idea is the same): Sometime probably around 1000 CE, two groups of Jews, the larger of which spoke Judeo-Old French and the smaller of which spoke Judeo-Old Italian (and both of which already had a Hebrew component), met in a High German-speaking territory (usually called "Loter", i.e. Lotharingia; others say it was Bavaria, and still others that this happened independently in both Loter and Bavaria). These Jews intermixed among themselves, creating some sort of mixture of Judeo-Old French and Judeo-Old Italian. They also learned German to be able to interract with the surrounding people (but keep in mind that Jews were somewhat insular and so this interraction was rather limited). The mixture started then and there, but the children of these immigrants still grew up speaking something mostly Romance-based. Gradually over one or two or three or more generations, this vernacular became more and more German-based through the influence of interactions with the surrounding German speakers, until the Romance part of the mixture became rather insignificant. At this point, you could call the language Yiddish. Now at no point here could you say that the Jews spoke the same language (or dialect, or variety, whatever the terminology) as the surrounding German speakers. Keep in mind that the role of Hebrew in all this, was that Ashkenazi Hebrew was in regular use as explained above throughout this whole process, and there was already a Hebrew component in the pre-Yiddish Judeo-Romance varieties that was carried over whole into Yiddish.

Now what does that mean? There would not have been "a time early in the history of Yiddish when had only just started to gradually evolve towards its later state", because Yiddish did not evolve from a variety of High German "toward its later state". Yiddish is not a "creole", because it did not evolve from a "pidgin", but that doesn't necessarily mean that there was no abrupt mixing going on (and honestly, I don't know where you would draw the line between "abrupt" and "gradual", but you can draw your own conclusion from my explanation above).

@Benwing2 said: But I still think you're falling prey to the common tendency to assume that there must be something universally special about what you're familiar with.

You are projecting your own experiences onto me. I am not saying that Yiddish is universally special, only that I know enough about it to hold an argument. I would make similar arguments about other examples of language contact, if I knew enough about them. But like I said before, all instances of language contact are unique in their own ways, and I don't think there is another case where I could make all of the exact same arguments, although I could sure make some of the same arguments and replace the others with other arguments specific to those scenarios. Anyway, it seems to me that experts on contact linguistics tend not to make claims about which languages count as ancestors and which do not. And I don't think we should compare the formation of Modern Hebrew to the formation of Yiddish, the circumstances were drastically different between the two.

@Benwing2 said: if it was truly relexified I might say it was a mixed language with two parents, or something.

Then you would have to change the definition you gave of an ancestor language.

--WikiTiki89 23:21, 10 February 2016 (UTC)

Regarding toddlers: Interesting. I could ask though if knowing how to recite a line implies yet any kind of understanding of meaning, though. (I recall being taught how to sing "Happy Birthday" in Spanish at the age of five or so, but only as a string of nonsense syllables sung to the tune, without even being told that they mean anything in an actual language.)

But OK, even if we suppose that all Jews would have been at least bilingual (in Hebrew and in a local Jewish language) — this still does not yet imply that the languages in question are "mixed" or have multiple ancestors that they would have inherited material from. Borrowings within a bilingual environment are still borrowings. Their status as borrowings is established already by the fact that Yiddish and Hebrew are distinct languages in the first place.

We might want to rather analyze some tokens as code-switching, of course. It's possible that this also resolves the issue of some Yiddish words allegedly "inherited from and never separate from Hebrew".

Regarding the formation of Yiddish: This is not how language shift works, regardless. There is always a stage of bilingualism involved. The supposition of a vernacular that "[g]radually (…) became more and more German-based" is untenable (at least as an ex cathedra supposition provided without explicit detailed evidence). The reconstructed pre-Yiddish Jewish community, as you admit, would need to have speaken both High German and their earlier Judeo-Romance varieties. Yiddish, judging by the fundamental Germanic-ness that people apparently see in it (I don't claim to know enough to be able to tell myself), must have evolved from the former, not the latter. (If it didn't, we would expect Yiddish to still be a visibly Romance language and not a Germanic one at all.) That being the case, if Yiddish is to have any ancestors at all, it needs to be analyzed as having gradually evolved from High German.

Otherwise the only plausible explanation will be, yes — a creole, evolved from a pidgin created by Jews who did not speak High German (older speakers still sticking with Judeo-Romance?) clashing with Jews who did not speak Hebrew (younger girls lacking the formal education?). In such a case we'd of course expect the usual features of creole grammar to emerge, instead of Germanic grammar being preserved. --Tropylium (talk) 09:47, 11 February 2016 (UTC)

Regarding: I could ask though if knowing how to recite a line implies yet any kind of understanding of meaning, though.

You could also ask whether an English-speaking toddler, early on in his acquisition of English, really understands much of what he hears and repeats in English. But as for the case in question, if the child were only taught one line of Hebrew, then of course it would be difficult to pick it apart, but once the child learns a few different prayers and blessings, they will begin to notice patterns between them, such as the names of G-d, or how the bit at the end of the blessings relates to what they are doing that calls for the recitation of the blessing (for example, the child recites the following in Hebrew before washing his hands in the morning and before meals: Blessed are you L-rd our G-d, King of the Universe, who has sanctified us with his commandments and commanded us regarding handwashing. The beginning part is the same for many different blessings, so it's not difficult to then associate the different part "regarding handwashing" with the act of washing one's hands.). This is really no different from how any child learns their native language.

Regarding: We might want to rather analyze some tokens as code-switching, of course.

I guess this is where the definition of the word "Yiddish" itself becomes important. Code-switching was essentially an integral and necessary part of how these Jews spoke. What you seem to be doing is defining "Yiddish" to be "the part of the Ashkenazi Jews' language that is consistent with being a descendant of High German" and labeling anything else as code-switching with Hebrew. But my definition of Yiddish is simply "the native language of the Ashkenazi Jews", which includes within it the things that you would call code-switching.

Regarding: It's possible that this also resolves the issue of some Yiddish words allegedly "inherited from and never separate from Hebrew".

The words I was referring to are actually the ones that are the farthest from code-switching, the words most thoroughly integrated into Yiddish, like גנבֿ (ganef, “thief”), which were likely already present in the pre-Yiddish Judeo-Romance. Yet, when these words were used in a purely Hebrew context, they were still thought of as the same words. In other words, it's not as though there were a Yiddish word ganef and an Ashkenazi Hebrew word ganef; there was just one word ganef.

Regarding: This is not how language shift works, regardless.

Look, I did not come up with this theory. As I mentioned, the version I gave is a summary based on the theory of Max Weinreich, probably the most famous Yiddish linguist, and whose son Uriel Weinreich happened to be one of the forefathers of general contact linguistics. I'll quote an excerpt of a footnote from Neil G. Jacobs' Yiddish: A Linguistic Introduction (note, "divergence" refers to the set of theories that view Yiddish as having started off as German and then having diverged to become Yiddish, while "convergence" refers to the set of theories that view Yiddish as having started off as not German and then having converged with German to become Yiddish):

Interestingly, there is most often a clear-cut divide between scholars trained in Yiddish linguistics, and those who use only the tools of German linguistics to deal with Yiddish. Yiddish linguists clearly work from a convergence approach (though aware of the German linguistic/dialectological issues). Scholars approaching Yiddish solely with the tools of German linguistics generally take a divergence approach. The latter thus frequently recognize EY [Eastern Yiddish] as a language, but see WY [Western Yiddish] as a type or types of German. They also frequently make errors which betray a fundamental lack of basic knowledge of Yiddish, and which betray a bias which sees Yiddish as essentially German. […] Significantly, scholars trained in general contact linguistics view Yiddish in terms of convergence (e.g., Louden 2000; Thomason and Kaufman 1988).

Regarding: Otherwise the only plausible explanation will be, yes — a creole, evolved from a pidgin created by Jews who did not speak High German (older speakers still sticking with Judeo-Romance?) clashing with Jews who did not speak Hebrew (younger girls lacking the formal education?). In such a case we'd of course expect the usual features of creole grammar to emerge, instead of Germanic grammar being preserved.

Two things: Firstly, there is no reason to assume there must have been a pidgin. Pidgins arise to aid communication between people who speak different languages but who do not learn each other's languages. There is no reason to assume that the Jews did not learn High German at least somewhat properly for use in communication with the surrounding people. No pidgin need have emerged and the languages (High German and Judeo-Romance) could have simply mixed in the home environment (although attested cases are rare, mixed languages do exist). Secondly, even if Yiddish did start out as a typical pidgin and then became creolized into a typical creole, there is no reason to assume that the typical creole features should have survived very long. The ongoing contact with the surrounding High German-speaking population could have influenced the creole until its grammar was for the most part "corrected". I'll re-emphasize that regardless of how Yiddish looked in its infancy, the ongoing contact with the surrounding High German-speaking population played an important role in its development.

--WikiTiki89 04:03, 12 February 2016 (UTC)

> But my definition of Yiddish is…

We're getting into "well that's just like, your opinion, man" territory here, aren't we? If your theory about Yiddish descending from Hebrew requires defining Hebrew literally as a part of Yiddish (and not as the separate Semitic language it is universally recognized as), you might want to step back a little and consider that you're not trying to improve our current work, you're trying to uproot entirely what we're even doing. There is nothing inconsistent with the idea that Ashkenazi Jews (or for that matter, most other diasporic Jews) maintained two native languages; and indeed since you're now elaborating how a Jewish child needs to learn separately, from separate contexts, the Yiddish and Hebrew vocabulary for say "washing hands", you seem to be already on board with the idea that the distinction exists.

For that matter, it is completely typical for minority languages around the world to exist in a state of universal diglossia, with their speakers natively speaking both their ancestral language and (possibly their own dialect or variety of) the majority's language. Many of these intimately incorporate loanwords and are often used with extensive codeswitching. Does this mean that the minority languages have ceased to exist already and should be analyzed merely as varieties of American English, Siberian Russian, Mexican Spanish, etc?

If you want to assert that equivalent meaning and shape is enough to claim that a word does not actually exist separately in two languages, I also await with interest your proposal for refactoring entries like internet or xenon into Translingual entries with different languages indicated merely as inflection variants.

(To be clear, this all is nonsense since a Jewish child will not merely learn e.g. ganef as a single word. They are required to separately learn that the sequence ganef with the associated meaning 'thief' is both a valid Yiddish word and a valid Hebrew word. That they sound the same, or are etymologically the same, is irrelevant.)

> Look, I did not come up with this theory.

I'm not saying you would be "coming up with" a theory, but you might be failing to understand what the linguists you're quoting are saying. Your new quote about a division between Western and Eastern Yiddish, for example, appears to be a squabble over dialectology that has nothing to do with the question of Yiddish's overall family affiliation (unless we jump right back to Wexler territory). Being "German", i.e. "a dialect of Modern Standard German" and being "High German", i.e. "descended from Old High German" are two different things.

If you now want to argue that Yiddish might be a mixed language, sure, go right ahead — I will read your or your source's full argumentation with interest. But casual hypotheses about "mixed languages" as intermediaries in language contact situations are lazy argumentation, and possibly overloading the meaning of "mixed" (which in casual descriptions and earlier linguistic terminology is often used to basically mean "has a fair bit of loanwords".) To me these hypotheses generally seem to be derived from an incredulity at the notion that e.g. Jews could have had their "own" dialect of High German, which eventually developed from a second language to their native one. (Perhaps the problem sometimes is against the concept of language change in general - that ethnic continuity must imply linguistic continuity as well?) --Tropylium (talk) 11:27, 12 February 2016 (UTC)

> We're getting into "well that's just like, your opinion, man" territory here, aren't we?

Well yes, thank you for understanding what I was trying to say. Without an agreed-upon definition of what is included in the definition of Yiddish, we cannot attempt to classify it.

> If your theory about Yiddish descending from Hebrew requires defining Hebrew literally as a part of Yiddish (and not as the separate Semitic language it is universally recognized as), you might want to step back a little and consider that you're not trying to improve our current work, you're trying to uproot entirely what we're even doing.

Well I'm not saying that all of Hebrew is part of Yiddish. At the very least, we have to narrow it down to Ashkenazi Hebrew (as perhaps you had assumed anyway), but even then, I'm not going to claim that all of Ashkenazi Hebrew is part of Yiddish, just some of it. But the problem is drawing the line. If your theory about Yiddish being a purely Germanic language requires cutting out essential parts of everyday Yiddish conversation, you might want to step back a little and consider that you're not trying to improve our current work, you're trying to uproot entirely what we're even doing (hmm, sounds familiar, doesn't it?). Anyway, my position is not that Yiddish necessarily descended from Hebrew, but that certain words from Hebrew, Judeo-Aramaic, Judeo-Greek, and Judeo-Romance should be considered as having been inherited by Yiddish.

> you're now elaborating how a Jewish child needs to learn separately, from separate contexts, the Yiddish and Hebrew vocabulary for say "washing hands"

Now where exactly did I say there was any non-Hebrew vocabulary for this? There was the Germanic verb expression זיך וואַשן עסן (zikh vashn esn), but the noun was only the Hebraic נטילת־ידים (netiles-yodayem), at least for the ritualistic sense.

> Does this mean that the minority languages have ceased to exist already and should be analyzed merely as varieties of American English, Siberian Russian, Mexican Spanish, etc?

If you are referring to minority languages that are still spoken as the everyday language of communication within the community, then this situation is entirely different and I won't comment on it (unless you really want me to). If you are referring to minority languages that have become mostly confined to specific areas of life, while a variety of the surrounding majority language is spoken as the language of everyday communication and incorporates elements of the minority language, then I would certainly say that those elements of the minority language found in their own variety of the majority language are inherited (but if these words happen to spread beyond their variety, then if viewed from the perspective of the majority language as a whole, I would say they are borrowed).

> If you want to assert that equivalent meaning and shape is enough to claim that a word does not actually exist separately in two languages, I also await with interest your proposal for refactoring entries like internet or xenon into Translingual entries with different languages indicated merely as inflection variants.

If there were a community of multilingual people who spoke every language on earth and regardless of which language they were speaking, internet and xenon were pronounced the same way, then yes, I would put internet and xenon into translingual sections. The point here is that this is from the perspective of the speakers of Yiddish, not the lexicographers.

> (To be clear, this all is nonsense since a Jewish child will not merely learn e.g. ganef as a single word. They are required to separately learn that the sequence ganef with the associated meaning 'thief' is both a valid Yiddish word and a valid Hebrew word. That they sound the same, or are etymologically the same, is irrelevant.)

Why do you say that? A child might notice that his father doesn't one particular word in front of his grandmother, and then also notice that he does still use another particular word in front of his grandmother; that doesn't mean that the latter word is actually a different word whether the grandmother is present or not, nor does it mean that the child has to learn its meaning twice. Anyway, more examples of why they are the same word: If the pluralization changes to a different suffix (like שבתות (shaboses) > שבתים (shabosem)), it does so in both languages simultaneously; and if a word acquires a new meaning (like משקה (maskhe): "drink" > "alcoholic drink" > "hard liquor"), it does so in both languages simultaneously.

> you might be failing to understand what the linguists you're quoting are saying.

Referencing, not quoting (with the exception of the one quote I gave). And I could say the same about you, except that you haven't even referenced or quoted any linguists.

> Your new quote about a division between Western and Eastern Yiddish, for example, appears to be a squabble over dialectology

I think you're missing the point of the quote. It has nothing to do with dialectology and only mentions the Western and Eastern Yiddish thing as an example. Perhaps I should have omitted everything from "The latter thus" to "as essentially German." as part of the ellipsis. The purpose of the quote was to show how a linguist's perspective going in influences whether this linguist treats the development of Yiddish from a divergence or convergence approach. Maybe you should re-read the quote with that in mind.

> If you now want to argue that Yiddish might be a mixed language, sure, go right ahead

I'm not arguing for it, I'm arguing against excluding it as a possibility, which is what you are doing.

> To me these hypotheses generally seem to be derived from an incredulity at the notion that e.g. Jews could have had their "own" dialect of High German, which eventually developed from a second language to their native one.

Coincidentally, this is discussed in part of the ellipsis that I omitted from my quote above (after a brief discussion of the Nazi ideologies of some of the German linguists who studied Yiddish): "It should be noted that some of these German-cum-Nazi scholars expressed a convergence-origins view on racial grounds – they saw Jews as racially incapable of making German fully their own true mother tongue (see Hutton 1999)."

--WikiTiki89 23:41, 12 February 2016 (UTC)

> If your theory about Yiddish being a purely Germanic language requires cutting out essential parts of everyday Yiddish conversation,

Aside from the fact that there is no such thing as "pureness" in language (in the current day, a language either is or isn't Germanic; there is no middle position), this alleged problem has a problem of circularity: you cannot complain that "everyday Yiddish" is being cut up without first already holding the point of view that a particular conversation in fact is held in Yiddish, and not in a mixture of Hebrew code-switched with Yiddish.

I make no claims about knowing a priori which particular instances fall under which class. This is an empirical question, not something that it would be even possible to have a pre-existing agreement on. But the distinction exists and is probably relevant for attempts to identify words of Hebrew origin in Yiddish. Failing to recognize code-switching will overgenerate "loanwords" that were really nothing but Hebrew all along. Failing to recognize loanwords, on the other hand, will overgenerate "code-switching" that supposedly operates only over a single technical term, and could perhaps seem to challenge the notion of Yiddish existing separately from Hebrew at all.

Do you accept that code-switching is a meaningful concept at all? Are you adding an exception clause for denying the possibility of code-switching between Hebrew and Yiddish? Or, perhaps, an exception clause for denying Hebrew-Yiddish loanwords?

> more examples of why they are the same word:

You seem to be merely attempting to equate different senses of "sameness". The claim that they are "the same" is either trivial, or trivially wrong, depending on what facts you're looking at. But I suppose you will be incapable of believing that they are distinct unless you first accept also that Yiddish and Hebrew are distinct.

> Anyway, my position is (…) that certain words from Hebrew, Judeo-Aramaic, Judeo-Greek, and Judeo-Romance should be considered as having been inherited by Yiddish.

Which would necessarily have to imply also a claim about Yiddish descending from Hebrew. The agreed-upon definition of "inheritance" is material directly derived from an ancestor. There is no other sense to the word in historical linguistics. Either we have a mere terminological confusion here, with you thinking that "inherited" means something similar to "cultural heritage" (in etymology this will not imply anything more than "derived"), or you actually want to assert that Yiddish descends from Hebrew.

Note furthermore that "inheritance" in the etymological sense is not even a property of a word in isolation. There is no analogue in linguistics to e.g. the biological process of DNA duplication that establishes a chain of descent. Learning a new lexical item is the same process regardless of what its source language is, and will likely indeed require reinforcement from multiple soures. The division of words as "inherited" versus "borrowed" is an abstraction that only makes sense in the context of languages' relationships to each others in their entirety, both their historical relationships (what has descendend from what?), and their synchronic relationships (what is a distinct language from what?). In the absense of clear answers to these, no distinction can be drawn in the first place.

>The purpose of the quote was to show how a linguist's perspective going in influences whether this linguist treats the development of Yiddish from a divergence or convergence approach.

I'm not sure what point you're trying to make about this then; it seems awfully trivial. If someone doesn't understand linguistic descent, and is mainly interested in exploring cultural heritage, of course they're unlikely to accurately identify the ancestry of a particular language variety. If someone doesn't understand language contact phenomena, and is mainly interested in linguistic genealogy, of course they're unlikely to accurately analyze language contact situations. Etc.

> I'm arguing against excluding [Yiddish as a mixed language] as a possibility, which is what you are doing.

I do not exclude the possibility (I know too little about Yiddish to conclusively reject it). But in the absense of detailed arguments I think we should follow the usual scientific consensus position that Yiddish is a regular Germanic variety that just contains several loanwords from e.g. Hebrew and Slavic. --Tropylium (talk) 18:01, 13 February 2016 (UTC)

> Aside from the fact that there is no such thing as "pureness" in language

Pure languages do not exist, because contact cannot be avoided by definition, since without contact there is not such thing as language. Also, purity is probably impossible to accurately measure. But that does not mean that the concept of purity does not exist, since no one would dispute that English-based creoles, for example, are less purely English than Southern US English. While, for another example, Southern US English cannot easily (or at all) be judged to be more or less pure than the Queen's English. Anyway, this is only tangential to our discussion about Yiddish.

> in the current day, a language either is or isn't Germanic; there is no middle position

In terms of classification and convenience, you're right, but in a purely linguistic sense, it is absurd to speak of such black-and-white contrasts and I would say that there are only middle positions.

> this alleged problem has a problem of circularity: you cannot complain that "everyday Yiddish" is being cut up without first already holding the point of view that a particular conversation in fact is held in Yiddish, and not in a mixture of Hebrew code-switched with Yiddish.

This is exactly why I'm saying that we have to first determine what counts as Yiddish without any consideration of genealogy or classification. After we have made this determination, only then can we think about genealogy and classification. Since we seem to be disagreeing about the first part, our whole argument somewhat meaningless.

> Do you accept that code-switching is a meaningful concept at all? Are you adding an exception clause for denying the possibility of code-switching between Hebrew and Yiddish? Or, perhaps, an exception clause for denying Hebrew-Yiddish loanwords?

Code-switching is one possible way to mix languages (or registers, or dialects, or whatever else). The existance of code-switching has no bearing on whether the resulting mixture should be considered its own language or not. The classic example of code-switching is when a single speaker learns one language in one context and another language in another context, and then in certain contexts code-switches. When a child acquires his first language from code-switching parents, then this can become the child's native language. I'm sure this happens all the time, but languages like this are usually unique to each family in which this takes place. Thus, even though I would call this a new language, it wouldn't be a significant enough language to give a name to and thoroughly study. However, when this phenomenon becomes the standard idiom of an entire society, then it does become significant enough to give a name to and thoroughly study. Now let me expand upon this a little further. When each of the two components in the mixture has an existing unmixed language corresponding to it, then you have a reference by which to separate out each component and you can sort-of get away with treating the components as separate languages (although you would end up with some problematic leftovers that don't fit into either component). This is unfortunately how most mixed languages end up being treated. But when one of these components doesn not have a corresponding language (or the component and corresponding language have diverged too far), then you can no longer pretend to treat the mixed language as two separate languages, because you have no way of unbiasedly separating out the components.

> You seem to be merely attempting to equate different senses of "sameness". The claim that they are "the same" is either trivial, or trivially wrong, depending on what facts you're looking at. But I suppose you will be incapable of believing that they are distinct unless you first accept also that Yiddish and Hebrew are distinct.

To clarify what I mean by same, I am trying to say that they are phsychologically handled as the same word. If I were to use an oversimplified and by-no-means-biologically-or-psychologically-accurate analogy, I would say that there is a checkbox for each word in your brain that determines which contexts it can be used in, and for some of the words I am referring to, the "every-day conversation" (i.e. "Yiddish") checkbox and the "Rabbinic literature" (i.e. Ashkenazi Hebrew) checkbox would both be checked (but the "bathroom conversation" checkbox would be unchecked).

> Which would necessarily have to imply also a claim about Yiddish descending from Hebrew. The agreed-upon definition of "inheritance" is material directly derived from an ancestor. There is no other sense to the word in historical linguistics.

What agreed-upon definition, and who agreed to it? To be honest, I've read quite a few articles on historical linguistics and I've never come across the word "inherited" or "ancestor" at all.

> Either we have a mere terminological confusion here, with you thinking that "inherited" means something similar to "cultural heritage" (in etymology this will not imply anything more than "derived"), or you actually want to assert that Yiddish descends from Hebrew.

I wouldn't mind saying that part of Yiddish descends from Hebrew and that most of Yiddish descends from German, and that as a result we classify Yiddish as Germanic rather than Semitic, but that classification in itself does not exclude the possibility of classifying words from the Hebrew component as inherited. And I'm not sure what "cultural heritage" has to do with anything, since language is not transmitted through whatever you mean by "cultural heritage".

> Note furthermore that "inheritance" in the etymological sense is not even a property of a word in isolation. There is no analogue in linguistics to e.g. the biological process of DNA duplication that establishes a chain of descent. Learning a new lexical item is the same process regardless of what its source language is, and will likely indeed require reinforcement from multiple soures. The division of words as "inherited" versus "borrowed" is an abstraction that only makes sense in the context of languages' relationships to each others in their entirety, both their historical relationships (what has descendend from what?), and their synchronic relationships (what is a distinct language from what?). In the absense of clear answers to these, no distinction can be drawn in the first place.

Of course "inheritance" is not a property of a word in isolation. "Inheritance" is a property of the history of a word (or other language feature) and how it came to be part of the language.

> The division of words as "inherited" versus "borrowed" is an abstraction that only makes sense in the context of languages' relationships to each others in their entirety, both their historical relationships (what has descendend from what?), and their synchronic relationships (what is a distinct language from what?). In the absense of clear answers to these, no distinction can be drawn in the first place.

I'm not sure what you mean by an "abstraction" here, but I agree that the division between "inherited" and "borrowed" makes sense only when you are looking at the history the language, particularly how the word was transferred from one speaker to another in relation to the speakers' languages, or from one context to another within a single speaker's speech in relation to these contexts' languages. The important thing is that "inheritance" and "borrowing" are processes independent of the classification of the languages. As I've been saying, we might not know which of these processes applied for every word in the Yiddish language, but we do have a good idea for at least some of the words.

> I'm not sure what point you're trying to make about this then; it seems awfully trivial. If someone doesn't understand linguistic descent, and is mainly interested in exploring cultural heritage, of course they're unlikely to accurately identify the ancestry of a particular language variety. If someone doesn't understand language contact phenomena, and is mainly interested in linguistic genealogy, of course they're unlikely to accurately analyze language contact situations. Etc.

You must have misread the quote then, because it clearly says that "scholars trained in general contact linguistics view Yiddish in terms of convergence", in other words they are on my side, as are "scholars trained in Yiddish linguistics". So you have on your side only "those who use only the tools of German linguistics to deal with Yiddish", who happen to also "frequently make errors which betray a fundamental lack of basic knowledge of Yiddish".

> But in the absense of detailed arguments I think we should follow the usual scientific consensus position that Yiddish is a regular Germanic variety that just contains several loanwords from e.g. Hebrew and Slavic.

Once again, you seem to have a strong misconception of what the "usual scientific consensus position" is. See my preceding paragraph.

--WikiTiki89 19:21, 16 February 2016 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘

> In terms of classification and convenience, you're right, but in a purely linguistic sense, it is absurd to speak of such black-and-white contrasts and I would say that there are only middle positions.

"Germanic" is in the end a concept of classification and convenience, yes. So are "inheritance" and "borrowing".

> When a child acquires his first language from code-switching parents, then this can become the child's native language. I'm sure this happens all the time, but languages like this are usually unique to each family in which this takes place.

I'm not sure I buy this idea in the case of mutually unintelligible languages. It's theoretically possible I guess, but I would expect the environment that led to the parents' code-switching in the first place to similarly lead to the child learning to break their nascent "native language" into two "registers" that are in fact two different languages. I'd similarly expect them to figure out in any case, already from their parents' speech, that there are two registers, not simply a large amount of synonyms and alternate constructions.

> When each of the two components in the mixture has an existing unmixed language corresponding to it, then you have a reference by which to separate out each component (…) But when one of these components doesn not have a corresponding language (or the component and corresponding language have diverged too far), then you can no longer pretend to treat the mixed language as two separate languages, because you have no way of unbiasedly separating out the components.

This does not seem not necessary. In a code-switching situation between different languages (as opposed to a mixed-language or register-variation situation), the distinction is already inferrable from what components get usually used together with which other ones. Codeswitching, as I understand it, works along the lines of "YYYYYHHHHHHHHHYYYYY" (alternation between the language of larger syntactic units) or "YYYYHYYYYYHYHYHYYYYYYHYYYY" (the insertion of what we could call on-the-spur loanwords), not "YHYHHHYHYHYYHYHHHHYY" (grammar and languages mixed willy-nilly).

> The important thing is that "inheritance" and "borrowing" are processes independent of the classification of the languages.

Disagree. At the grassroots level, there is no real distinction between the two, only a single process of learning vocabulary (or shifting registers). They can only be distinguished with reference to language classification. (For example, to ward off any age-based arguments, it's impossible to define some kind of a clear cutoff point for where you "finish learning" your native language. Inherited words don't stop being inherited even if you learn one at 15, or 85 even; loanwords don't stop being loanwords even if you adopt one at 3.)

> it clearly says that "scholars trained in general contact linguistics view Yiddish in terms of convergence"

Ah, that line (as opposed to the "scholars trained in Yiddish linguistics"). The same point applies, though. I'm sure contact linguists have argued for viewing Yiddish in terms of convergence, but this does not necessarily imply anything special about Yiddish — I see contact linguists often argue for viewing this or that or the other language variety in terms of convergence just as well. Also, some claims about convergence require stronger evidence than others, in particular claims about the convergence of unrelated languages to produce a third language yet. Most such cases I've seen have held little water, and nothing I've heard here suggests that the case of Yiddish would be any different. --Tropylium (talk) 21:21, 16 February 2016 (UTC)

> I'm not sure I buy this idea in the case of mutually unintelligible languages. It's theoretically possible I guess, but I would expect the environment that led to the parents' code-switching in the first place to similarly lead to the child learning to break their nascent "native language" into two "registers" that are in fact two different languages. I'd similarly expect them to figure out in any case, already from their parents' speech, that there are two registers, not simply a large amount of synonyms and alternate constructions.

Since I guess we're both speculating, do you know of any research on this type of situation? And I'm asking out of interest, not argumentation.

> This does not seem not necessary. In a code-switching situation between different languages (as opposed to a mixed-language or register-variation situation), the distinction is already inferrable from what components get usually used together with which other ones. Codeswitching, as I understand it, works along the lines of "YYYYYHHHHHHHHHYYYYY" (alternation between the language of larger syntactic units) or "YYYYHYYYYYHYHYHYYYYYYHYYYY" (the insertion of what we could call on-the-spur loanwords), not "YHYHHHYHYHYYHYHHHHYY" (grammar and languages mixed willy-nilly).

It's not completely willy-nilly, but that doesn't mean you're able to pick the two languages apart. You can have alternating syntactic units with on-the-spur loanwords withing them and you could quite possibly end up with something looking much like "YHYHHHYHYHYYHYHHHHYY". But most imporantly, once this code-switched language begins to evolve on its own (which could easily happen within the first generation raised with it), you're going to have a lot of trouble classifying the newly formed elements.

> Disagree. At the grassroots level, there is no real distinction between the two, only a single process of learning vocabulary (or shifting registers). They can only be distinguished with reference to language classification. (For example, to ward off any age-based arguments, it's impossible to define some kind of a clear cutoff point for where you "finish learning" your native language. Inherited words don't stop being inherited even if you learn one at 15, or 85 even; loanwords don't stop being loanwords even if you adopt one at 3.)

In my mind it's like this: If you learn a word as part of acquiring the language of a particular context, then you have inherited this word into your idiolect of that context. If you learn a word in one context and use it in another context where you have not also heard the same word, then you are borrowing it into your idiolect of that other context. When a borrowing occurs between two contexts that we see as the same language, we tend to ignore it.

> Ah, that line (as opposed to the "scholars trained in Yiddish linguistics"). The same point applies, though.

So you want to tell me that "scholars trained in general contact linguistics" do not "understand language contact phenomena"? I'm thoroughly confused. And keep in mind that the "scholars trained in Yiddish linguistics" does not refer to kind of some amateur linguists. They would certainly understand perfectly well "linguistic descent" (whatever you even mean by that) and "language contact phenomena". Perhaps you should read their work before insulting their skills.

> I'm sure contact linguists have argued for viewing Yiddish in terms of convergence, but this does not necessarily imply anything special about Yiddish — I see contact linguists often argue for viewing this or that or the other language variety in terms of convergence just as well. Also, some claims about convergence require stronger evidence than others, in particular claims about the convergence of unrelated languages to produce a third language yet. Most such cases I've seen have held little water, and nothing I've heard here suggests that the case of Yiddish would be any different.

So first you accuse me of ignoring the field of contact linguitics, and now you're accusing the field of contact linguistics of producing nonsense? The only conclusion I can draw from here is that you consider yourself to be the only reliable authority on contact linguistics. If that is the case, please go study Yiddish and publish your dissenting research and enlighten the world.

--WikiTiki89 22:35, 16 February 2016 (UTC)

lähi- and lähteä

Would you know if these two are related at all? —CodeCat 19:18, 16 February 2016 (UTC)

They aren't. The first is from PU *läse-, the latter from *läkte-. --Tropylium (talk) 20:16, 16 February 2016 (UTC)

Ok, thanks. —CodeCat 23:59, 16 February 2016 (UTC)

The verbs with stems in -ne- and -cce- in Proto-Finnic

Is there a (post-)Proto-Finnic process that assimilates *-ndak and *-ctak into *-t'ak before a third or higher syllable? In Finnish and Veps, verbs in these classes (Finnish vanheta and valita types) have infinitives in -ta, which suggests that such a change took place. Veps apparently has a similar change with -ntadak > -ta, judging by semeta vs. siementää. In Votic, the -tse- class seems to have been mostly reformed into a -tsa- class, with infinitives in -tsaa (valittsaa, tarittsaa). However, I found one relic, iloita, present iloitsõn, which perfectly matches its Finnish cognate.

A curiosity is the pair of kaitsea vs kaita, where the change seems to have happened before the second syllable: *kaiccetak > contraction *kaictak > assimilation *kait'ak. The change apparently failed in seistä. Could you shed some light on this? —CodeCat 00:20, 17 February 2016 (UTC)

I feel we've gone over this before, but {{R:fi:SKRK}} explains the Finnish development being mostly due to analogy with verbs of the salata type, in the first group especially due to almost complete synonymy (as in e.g. aueta, indeed analyzed by us as one verb with two possible conjugations). Note also the imperative forms, e.g. vanhetkoon, valitkoon where there's no *t-material in the suffixes to begin with. He mentions however that the development could have been regular in forms going back to *-C-tt- (*vanhenttu > vanhettu, *valicttu > valittu).

The Veps change could be analogy as well, I'm not sure if we should expect syncope in *seementädä(k) in the first place (cf. varastaa ~ vargastada, not **vargastta; OTOH still puhaltaa ~ puhalta, not **puhaldada). At minimum there is no general shift *ntt > t, cf. > antta.

Your Votic example looks like an Ingrian loan? It is reported in Vadja keele sõnaraamat from only one village, while ilottsaa is recorded from three. --Tropylium (talk) 01:09, 17 February 2016 (UTC)

Analogy in Veps may be possible but it has been applied to an entire class in that case, see Category:Veps words suffixed with -eta (ndab) which is the equivalent of the -entaa class of Finnish. What is more curious is that, due to this change, the transitive (-nta-) and intransitive (-ne-) classes have partially fallen together. There's a fair number of Veps verbs in -(e)ta which are intransitive or transitive depending on their conjugation type. It would be strange for analogy to cause this kind of class-wide homonymy.

In any case, I wonder how these might be reconstructed for Proto-Finnic. Would we simply assume and ? That seems strange somehow when the Northern Finnic languages agree so widely on -ta. Even in other Finnic languages there is some agreement: Votic also has vanata (present vananõn), which is a rather basic verb which speaks against borrowing. Võro has vanadaq alongside vananõdaq, and validaq alongside valitsaq. The Võro forms are not a result of Estonian influence either, as Estonian has reformed these with simple -da suffixes: vananeda, valitseda. Perhaps Livonian can tell us more, but all in all it looks like if analogy occurred, it occurred early enough to affect all these languages, so infinitives in -ta must probably be reconstructed for Proto-Finnic. —CodeCat 01:31, 17 February 2016 (UTC)

At least *-ne- : *-t'ak looks Proto-Finnic to me, yes. --Tropylium (talk) 02:44, 17 February 2016 (UTC)

North Karelian š

Does North Karelian even have a phonemic s distinct from š? If not, then I think we should only use the spelling with s, and indicate the difference in the pronunciation section. —CodeCat 23:06, 3 March 2016 (UTC)

It does, though the contrast is fairly marginal in native vocab. (Closest thing to a minimal pair I can find: kuši 'urine' vs. šusi 'wolf'.) (It's also the actual spelling, not simply transcription.) --Tropylium (talk) 23:25, 4 March 2016 (UTC)

Something interesting I came across

I was compiling descendants of and noticed that Võro has a -v-: püvi. This strikes me as a definite archaism, reflecting a form that must have existed before the contraction -üve- > -üü-. Is this indeed the case? And what does it mean for the shape of this reconstruction? Võro does have the contraction in all the other long-u/ü stems that we have entries for, but those all seem to have different origins as well. —CodeCat 20:20, 5 March 2016 (UTC)

Probably we'll just have to reconstruct both variants. There are other cases where *Vwe > *VV fails too, such as (but ), or Livonian õvā (but vuo < < *uwa), or within a single lexeme in a single language even, Fi. tyven ~ tyyni.

One hypothesis could be that the contraction originally applied only in consonant-stem forms (*püw-tä etc.) and was later in most words generalized across the paradigm (but in others perhaps levelled away), but this seems to be not really directly demonstrable. --Tropylium (talk) 22:09, 5 March 2016 (UTC)

Reconstruction:Proto-Finnic/papu

This is clearly a late borrowing, it doesn't seem to exist across Finnic so it's areal. I wonder if you could provide more descendants, to get an idea of the general area where this term is used? Also, I'm curious if the borrowing could be dated based on that. It must have occurred before the rounding of Slavic o occurred, and before the final vowel was lost, so that places it in early or Pre-Common Slavic territory (5th century-ish?). Would you happen to know at what point Slavic o starts to be borrowed as o into Finnic? —CodeCat 18:12, 11 March 2016 (UTC)

Finnish inessive -hna

Doesn't the Finnish dialectal inessive -hna require the reconstruction of proto-Finnic inessive *-sna rather than *-ssa? Also, standard Finnish has siinä as the inessive of se. --Muhaha (talk) 07:36, 14 May 2016 (UTC)

The usual approach to this (and the similar South Estonian inessive) that I've seen has been to assume *-hnA and *-ssA as parallel allomorphs. (Plausibly the former could have been used after secondarily stressed syllables, the latter otherwise, though to my knowledge such distribution has not been actually attested.)

siinä is just an instance of the archaic locative sense of the modern essive ending *-nA (with the long vowel presumably by analogy to the plural stem nii-). Similarly siitä retains an archaic separative sense of the modern partitive ending *-tA. Calling these "inessive" and "elative" might be a bit misleading.--Tropylium (talk) 16:51, 14 May 2016 (UTC)

I'm curious as to why s became h in the ending, myself. —CodeCat 16:57, 14 May 2016 (UTC)

*s > *h before nasals/liquids is regular, as in e.g. *pićlä > (→ > pihlaja), *gīslaz → kihla. --Tropylium (talk) 17:25, 14 May 2016 (UTC)

I'd infer, then, that sn > ss in the ending preceded this change. It begs the question what caused it to fail in some dialects. Past participles also have sn > ss, are there any examples of hn here? —CodeCat 18:09, 14 May 2016 (UTC)

I have not heard of any such cases, but it's not as if I have a full list of all past participles in all Finnic varieties to consult. --Tropylium (talk) 18:40, 14 May 2016 (UTC)

What's the etymology of the Sami inessive singular -s and plural -in? Is it a regular development from *-sna and *-j-sna?--Muhaha (talk) 18:24, 7 July 2016 (UTC)

Northern Sami -s is indeed from -stē < *-snē < *-sna (Southern Sami still has -sne; most of Eastern Samic has -st), the plural though is from plain *-j-na (so cognate to the Finnic plural essive). --Tropylium (talk) 19:14, 7 July 2016 (UTC)

The cases in Samic are, overall, much less symmetric between singular and plural, and resemble Indo-European in that respect. The Wikipedia article on Proto-Samic has more. —CodeCat 19:36, 7 July 2016 (UTC)

Was sn>ss a sound change or an analogical change in suffixes, like rn>rr (purrut but saarna)? The ln>ll change seems to be an actual sound change that ignores grammar (villa, tullut). --Muhaha (talk) 12:33, 12 September 2016 (UTC)

What's the cause of the shortening of the case suffixes in some dialects (e.g. inessive -s, Stadis, Turus). It doesn't seem to be a regular sound change. --Muhaha (talk) 12:44, 12 September 2016 (UTC)

We might be able to see about the former if you manage to find an inherited word or a reasonably old loanword that has had stem-medial *-sn-. Thus far none are known (to me, anyway). Sound changes that have applied solely in loanwords are common enough in the history of Finnic (what with Uralic phonotactics being much more restricted than Indo-European), but so are analogical developments.

The latter: indeed no. There might be newer, more detailed accounts, but the traditional explanation connects this to sandhi on one hand (loss of any -A before a vowel-initial word, as in tarkk’ampuja), on the other hand to dialectal syncope processes in some western dialects (Estonian-style loss of all 3rd+ mora final vowels in the southwest; sandhi loss of final vowels between two voiceless consonants in lower Satakunta), followed by generalization and spreading of the short allomorphs into numerous other dialects. --Tropylium (talk) 13:22, 12 September 2016 (UTC)

User:Tropylium/Proto-Indo-European/Verbs

I've changed itr. into intr. (and fixed the entry).

Thanks, that certainly doesn't hurt.

I have typed a list too from a Spanish source of 1996: User:Sobreira/Roberts-Pastor, only a few translations left. They give also the LA and GRC reflexes, which will take me much more time. I will have some questions if you don't mind, about how to conflate or coordinate or relate these two lists, or at least better understand mine. Sobreira (talk) 08:41, 14 June 2016 (UTC)

For bridging Spanish and PIE, an obvious starting point would indeed be Latin. I know too little about Romance to help you too much with the data otherwise. (You can probably find a PDF version of {{R:De Vaan 2008}} online if you google around a little bit, by the way.)

For transforming traditional PIE transcription into laryngeal theory transcription, the following basic substitutions usually help:

• word initial *e, *a, *o ⇒ *h₁e, *h₂e, *h₃e (respectively)
• long *ē, *ā, *ō in roots ⇒ *eh₁, *eh₂, *eh₃ (respectively)
• root-final *ə ⇒ *H (any laryngeal)

So e.g. "*pā-" 'to protect' translates into *peh₂- (“to protect”); or, "*arə-" 'to plough' translates into *h₂erH-, which we find as *h₂erh₃- (“to plough”) on the LIV list.

If you have other questions, go ahead. --Tropylium (talk) 15:28, 14 June 2016 (UTC)

Thanks for the idea of PDFing the deVaan, I never tried because I wouldn't have never thought I could find it, but I already did. I realised the correlations between laryngeals and changed/long vowels already, but some of them resist. I try also going up from the reflexes, and through the meaning, but some are resistant however. I will need more practice. I have a goal: I like the entry *ten-, it's clear, quite explicative, referenced and systematic. Until *tén-onts, gives the tense/aspect.. of the verb so I guess they can be derived from the root. Pity for the rest. I don't know whether there are rules for those Extended form[ation]s. The topic is much like the one of that User:Sobreira/Roberts-Pastor, they state + suffixes every now and then (see first three), and I don't believe it too much until I start seeing coincidences (and so far not much) and then I begin introducing into the Wiktionary. Thanks for the patience. Sobreira (talk) 09:44, 15 June 2016 (UTC)

Not all of the "roots" listed on that page are actually roots in the usual PIE sense. "agro-" for example is most certainly a noun stem. Roots begin and end with a consonant, only ever have at most one vowel, and there is a strict ordering of the initial and final consonants meaning that a root can never end in -gr- for example. See w:PIE root. —CodeCat 13:17, 16 June 2016 (UTC)

New try

I have started with a more simple word/root, with much less reflexes/derivated: Reconstruction:Proto-Indo-European/h₃nobʰilos < Reconstruction:Proto-Indo-European/h₃nebʰ-, "navel". The compilation is at User:Sobreira/*h₃nobʰ-. I guess that the two first and the last, with omissions, are wrong just for typos, which leaves the questions for:

• of the E-grade or O-grade (and zero-grade), which do we use as main lemma? (I start from the supposition that all post-PIE evolutions can begin from any of them unless added by a suffix/declension/conjugation which could change the vowel somehow).
• when do I know whether a metathesis is legitimate? Well, I guess they happen sporadically and random, and it's legit as far as it can explain an evolution. So the real question is: when are the claims of metathesis supported? My own answer: when it's referenced, so we should ask the source to the editor that included it. The real question comes from the fact that the metathesis is only defended by one etym among two others in one word for one language: ὀμφαλός. Someone knowing about Old Greek could know whether it is (or they are) possible claims. As far as I know, the most logical would be a chain:
  • *h₃n̥bʰ- (zero) > Ablaut to e > *h₃nebʰ- > metathesis > *h₃enbʰ- > Ablaut to o > *h₃onbʰ-; or
  • *h₃n̥bʰ- (zero) > Ablaut to e > *h₃nebʰ- > Ablaut to o > *h₃nobʰ- > metathesis > *h₃onbʰ-; or even better for shorter and simpler
  • *h₃n̥bʰ- (zero) > Ablaut to o > *h₃nobʰ- > metathesis > *h₃onbʰ-

but as h₃ gives already /o/, I don't know the need for arriving into -onbʰ- as I think there are more ways to get /o/ (or /ó/ or /ō/) in Greek.

• can the nasal/sonorant n be transmutated into m̥ in zero grade or we should notate n̥? Or is a generic/archiphonemic nasal?
• could really *h₃nobʰilos be a diminutive of *h₃nobʰ-/*h₃nebʰ- as said in navel#English? (then we have to create the entry for the suffix

I have to go now. No hurry, thanks. Sobreira (talk) 13:08, 16 June 2016 (UTC)

On transcription:

• E-grade is the default for roots, any other formations should signify whatever vowel the descendants indicate (and if they indicate disparate vowels, the form would need to be broken in multiple entries).
• PIE had, as far as I can tell, no *-nbʰ-, so the zero grade would be expected to be realized as *h₃m̥bʰ- at least phonetically. However, we seem to lean towards more abstract rather than more concrete root forms in general (e.g. in writing *h₂e- and not *h₂a-), so perhaps that suggests that we should notate the zero-grade root as *h₃n̥bʰ-?

On reconstruction:

• Did laryngeal vocalization predate the vocalization of *N̥ in Greek? If it did, then we would seem to just have *h₃m̥bʰ- > *om̥bʰ- > *ombʰ- > *ompʰ- without any need to assume metathesis.
• If it didn't, then we seem to have here a case of the phenomenon called Schwebeablaut: since *R̥ is the zero grade of both *eR and *Re, sometimes new root variants like *h₃embʰ- / *h₃ombʰ-, with the full-grade vowel in the "wrong" place, were analogically generated from zero grade forms like *h₃m̥bʰ-. --Tropylium (talk) 20:42, 21 June 2016 (UTC)

bidjat

Is this from *piejëtēk? The vowel doesn't match, but it's not a stretch to assume that Northern Sami might have had iej > ij > idj as a regular change. —CodeCat 20:14, 21 June 2016 (UTC)

This seems to be the etymology, yes. The change looks irregular though: Northern Sami still has numerous words with -iei- (nieida (“girl”), hieibma (“gentle wind”), gieibmi (“rut”) etc.) or -iedj- (biedju (“lair”), niedjat (“to descend”), etc.) Perhaps it is analogical from some contracted-vowel inflected form (since *ie-VV > *ii-VV is regular). Loaning from Skolt Sami piijjâd would also seem to work… though probably only for Finnmark dialects, not the Torne Valley ones. --Tropylium (talk) 20:26, 21 June 2016 (UTC)

jakŋat vs *jënëtēk

Your Samic list has *n, but the Northern Sami term reflects *ŋ. Is this an error in your list? —CodeCat 17:44, 25 June 2016 (UTC)

Yes, typo on my behalf. Fixed. --Tropylium (talk) 17:53, 25 June 2016 (UTC)

Uralic words for ‘copper’ and ‘iron’

Hi. How does modern research explain the two groups of Uralic words mentioned in երկաթ (erkatʻ)? --Vahag (talk) 18:10, 1 July 2016 (UTC)

1. #Vrg- group: the Hungarian is a literary coinage, while the others cannot be shown related to eachother either (though likely the Mansi is an old Permic loan). Both might be from Alanian (cf. Ossetian ӕрхуы (ærx°y)).
2. *kärt group: loans from Iranian *kart- (evidently mostly reflected as 'knife' in modern languages, e.g. Ossetian кард (kard), but still 'steel' in e.g. some Pamir languages).

A recent article by Viitso is a good review of these and other metal words if you want the full scoop. --Tropylium (talk) 18:56, 1 July 2016 (UTC)

Thanks, that was a useful article. --Vahag (talk) 17:38, 2 July 2016 (UTC)

Inflection of Northern Sami participles

I don't know if you know much about Northern Sami, but I don't really know anyone else to ask. Northern Sami has two participles, present and past. Adjectives normally have a separate attributive form when modifying a noun, and more noun-like case forms when used otherwise. I can't figure out if this includes participles. That is, whether the present and past participle have case forms or if they exist only as attributive forms. Would you know this? You may be able to find it in Finnish-language sources that I'm not able to understand. —CodeCat 19:03, 2 July 2016 (UTC)

njárbat and *ńārpē

Are these the same root? Your list of Samic words glosses it with "think", which is rather strange, is it perhaps a typo for "thick (of a liquid)"? —CodeCat 15:09, 3 July 2016 (UTC)

That's another typo, yes: should be 'thin'.

Also, for future reference — Lehtiranta's lexicon {{R:YSS}} is a bit inexact with the reconstruction of adjectives. For most of them we'd really need two separate Proto-Samic stems, i.e. the attributive and predicative forms, while he reconstructs only a somewhat abstract "root" underlying the two (so e.g. njárbat would be from a PS attributive form *ńārpëtē). --Tropylium (talk) 17:15, 3 July 2016 (UTC)

Did the attributive-predicative distinction already exist in Proto-Samic, then? —CodeCat 17:26, 3 July 2016 (UTC)

Definitely. A recent talk I was at distinguished about 5 morphological adjective classes in PS, each with a specific attributive/predicative ending alternation (though I gather this particular research is so far unpublished). --Tropylium (talk) 20:53, 3 July 2016 (UTC)

I have just been creating entries for attributive adjectives and noticed a few patterns too. It seems that -s is generally associated with the attributive, but sometimes the entire paradigm has it. The pattern -ad- with attributive in -es is common. —CodeCat 21:28, 3 July 2016 (UTC)

Finnish week days' sound development

Were the Finnish week day names maanantai, tiistai, torstai, perjantai, lauantai and sunnuntai borrowed from a Germanic language with a g to j sound change, or is it a Finnic sound change, like consonant gradation after unstressed vowel (*tiistaki -> *tiistagi -> tiistai, analogized into the other days like maanantai where the k would be after secondary stressed vowel)? Is the n in lauantai (instead of laugardagr -> **laugartaki/**laugartagi -> **lauartai) analogy after the other days? --Muhaha (talk) 21:13, 8 July 2016 (UTC)

The sources I have at hand don't seem to comment on this, but I'd suppose that the development of -tai was a bit of both — Old Swedish had -ɣ- in these, and the lenition -ɣ- > -j- is native Finnish (as in e.g. petäjä < *petäɣä < *petägä). As you may know, Eastern dialects have maanantaki etc. but this probably involves the fact that loss of -ɣ- was much earlier in these, so they had to substitute something else.

Lauantai looks to me the most like as if the word was reanalyzed as a genitive compound lauan tai 'laua's day' (and the same applies to perjantai as well, for which we'd expect ˣperjartai). --Tropylium (talk) 00:17, 9 July 2016 (UTC)

Huh. I thought the reason for -tai, väri, -pori etc. was simply the Modern Swedish pronunciation of word-final -g as [j]. --Florian Blaschke (talk) 21:53, 27 July 2016 (UTC)

So, was it laughantaghi in Old Finnish? (Is it even attested?) --Florian Blaschke (talk) 18:29, 28 July 2016 (UTC)

Doesn't seem to have been attested with -gh-, but *ɣ was lost earlier from between unstressed syllables than when post-tonic anyway.

Some cases like väri can well be directly from later Swedish, but as far as I know /j/ for -g only applies in -lg -rg: tog could be /tuːg/ ~ /tuː/, but not ˣ/tuːj/. --Tropylium (talk) 19:18, 28 July 2016 (UTC)

I didn't know this resource. Cool! So it was basically *laughanta'i (or *-taji?), but [ɣ] became [w] early on after labial vowels, I guess.

Oh, of course, I forgot: jag is [jɑː(g)], after all. --Florian Blaschke (talk) 15:07, 30 July 2016 (UTC)

Sorry for meddling with your user subpage

I just realised that User:Tropylium/Proto-Indo-European/Verbs is a personal user subpage and not a page in the main namespace of Wiktionary. D'oh! I got there from your contributions page and missed the "User:Tropylium" part in the title; it looked so deceptively like an "official" Wiktionary page to inattentive, delirious, tunnel-visioned me! I'm not sure if it was appropriate to meddle with the introduction there. I was only feeling defensive of the controversial "single-brancher" practice used by the authors, as I happen to think their rationale makes a lot of sense. I believe other authors such as Schumacher or Kroonen have essentially taken the same stance. You are of course welcome to disagree and revert my commentary. --Florian Blaschke (talk) 21:52, 27 July 2016 (UTC)

No problem, looks like a good addition to me. I wouldn't mind moving the list to an appendix at some point, if anyone feels it would be useful. --Tropylium (talk) 00:17, 28 July 2016 (UTC)

Phew. As far as I could see, you don't class roots as single-branchers if nominal formations are known from other branches, right? --Florian Blaschke (talk) 18:25, 28 July 2016 (UTC)

Nope, after all we don't maintain a distinction between verb and noun roots for PIE. (Though I've considered still noting this in some fashion.) --Tropylium (talk) 18:59, 28 July 2016 (UTC)

Hm, I guess it makes sense wiki-internally, though I recall that Karin Stüber was quite insistent that there must have been a distinction between verbal and nominal roots, and roots which occur in both functions, i. e. as both a verb and a root noun (such as *ped- or *h₃ekʷ-), are actually quite rare, IIRC. --Florian Blaschke (talk) 15:14, 30 July 2016 (UTC)

palaa-polttaa

Is the vowel change in palaa/polttaa regular (*pala- / *poole-)? Would *pooltta- be shortened into *poltta- because of non-allowed cluster? --Muhaha (talk) 09:26, 30 July 2016 (UTC)

This remains unresolved, but it seems to go back already to a pre-Finnic *a/*o alternation, reflected also in e.g. Erzya паломс (paloms) : пултамс (pultams); it seems IMO unrelated to *a > *oo in Finnic. No weirder than a number of other apparent but unexplained vowel alternations in Uralic really (e.g. jalka : jälki). ---Tropylium (talk) 13:58, 30 July 2016 (UTC)

Samoyed socks

Hey, I was padding up this Mongolian entry with an etymology, and I came across a note relating it to Proto-Samoyedic *päyma in "The Mongolic Languages" by Juha Janhunen et al. I'd like to use concrete attested vocabulary instead of a reconstruction for this, but I don't know how to find Samoyedic words, so I thought I'd ask you, being the leading internet-approachable expert (I love your blog by the way) on Uralic languages, if you could dig them up? Crom daba (talk) 16:59, 17 August 2016 (UTC)

Janhunen in his etymological lexicon of Samoyedic gives Nganasan хуайму; Enets [script needed] (fē); Tundra Nenets пива (pīwa); Forest Nenest [script needed] (pĭemmŏə); Selkup ; Kamass [script needed] (peiima); Mator [script needed] (хи́ма). These all mean 'boot'.

I probably should be creating Proto-Samoyedic entries here and there at some point, but alas, there's only 24 hours in a day (and also I'd like to have some entry guidelines for extinct languages like Mator and Kamass clarified a bit first, but that's another can of worms). Maybe I'll do a test run with this one though. --Tropylium (talk) 17:52, 17 August 2016 (UTC)

Thanks a lot! This word is weird, Mongolic *h- usually corresponds to Turkic *h-, and I haven't seen *yi corresponding to *ĺ/ş before, I'll mark it down as 'relation uncertain' Crom daba (talk) 21:54, 17 August 2016 (UTC)

It looks to me as if it would have gone Turkic → Samoyedic → Mongolic rather than directly between T and M (the vocalization *ĺ > *j is a regular change in Samoyedic), but I wouldn't know how well this squares with the usually accepted chronology of *p > *h in the two families. (Also, Altaicists apparently reconstruct two separate consonants: *pʰ for *h in both, *p for cases like these with Turkic *b ~ Mongolic *h.) --Tropylium (talk) 22:51, 17 August 2016 (UTC)

Regarding Erzya

You may have already noticed this already, but I suspect that situation with Erzya entries could be pretty bad, aka, essentially conlanging. I remember I noticed this with pičemar' ("apple from a pine") which in Moksha is a pinecone or a kidney (if I'm not mistaken, according to published dictionaries the same in Erzya) but it somehow came to mean a pineapple in Erzya, when variations of "ananas" rule from Netherlands to Kazakhstan and beyond (aka, no cultural context for calquing the English word, especially as it's already pretty overloaded semantically, as happens often in the Volgaic languages.)

Easterner / post-soviet "work ethic" doesn't help either, for example, this Ščankina lady actually received an EU grant for compiling a Volgaic-Russian dictionary (I presume way before EU-Russia relations deteriorated), what she (they?) did was "repackage" an old already existing dictionary as it were, an exact carbon copy, so there's that.

So on one hand you have Russian academics that don't give a single f*ck and on the other -- enthusiasts that go crazy with conlanging / ex nihilo word coining. Neitrāls vārds (talk) 22:35, 28 August 2016 (UTC) Neitrāls vārds (talk) 22:35, 28 August 2016 (UTC)

Honestly I haven't kept much track of our situation with the various Volga region languages, beyond occasional formatting for inherited material. (I might like to do some cleanup projects on word derivation at some point, but currently I don't even have basic dictionaries available for consultation at home — aside from Mari.) And they're living languages that can well contain neologisms, of course, but it also wouldn't surprize me if there were language activists out there trying to promote their own coinages in places like this.

If you run into anything especially dubious though, feel free to drop them at RFV and ping me, I can try looking up if they're citeable in printed materials. --Tropylium (talk) 22:54, 28 August 2016 (UTC)

Not that I work with these languages very much, but this information scares me. Anyway, this is why we have RFV and why we can't be lenient in attestation requirements even for small languages. --WikiTiki89 11:02, 29 August 2016 (UTC)

Selkup caps

Hey, it's me again, I found another Mongol-Uralic comparison. Ramstedt mentions, in connection with Mongolian малгай (malgaj, “hat”), Taz Selkup, which Alexander Castrén links to Khanty. Do you know anything further about the Khanty word? (I've already checked Janhunen (1977), it doesn't have anything that seems related under m.)Crom daba (talk) 16:06, 7 September 2016 (UTC)

It certainly exists, at least: müḷ ~ miḷ ~ mil, with regular variation per dialect, meaning 'hat', 'winter hat' ('tuque'?), 'men's hat'. Some modern dictionaries have as written forms Eastern , Northern . Steinitz' dialect/etymological dictionary of Khanty does not seem to list any known origin for this. (Neither does Alatalo's dialect dictionary of Selkup for that one.) The Mongolian and Selkup look like they could well be related, but linking in the Khanty word to this seems more difficult, given the front vowel and no indication of a 2nd syllable. --Tropylium (talk) 12:33, 12 September 2016 (UTC)

A proposal on splitting Monguor into Mangghuer and Mongghul

Hey, a proposal I've made at Wiktionary:Requests_for_moves,_mergers_and_splits#Splitting_Monguor_into_Mangghuer_and_Mongghul seems to be stuck for a long time now, could you perhaps take a look at it, share your thoughts and vote? Crom daba (talk) 00:32, 21 December 2016 (UTC)

//NOTE: This message was crossposted to multiple talk pages. Crom daba (talk) 00:32, 21 December 2016 (UTC)

kalma

Is kalma derived from the root of kuolla (vowel change *a -> *ō due to original open/closed syllable difference)? --Muhaha (talk) 10:38, 28 January 2017 (UTC)

It's either that, or from Germanic; either seems to work perfectly well. We can in any case reconstruct both *kalma already for Proto-Uralic, given e.g. Erzya калмо (kalmo, “grave”) and Nenets хальмер (xalʹmyer°, “ghost”); and the seemingly unrelated *kwalmaz (“death”) for Germanic.

(It also looks a bit like as if either the later derivative kuolema or a hypothetical Samic cognate **kuolmā- could have been loaned into Germanic, but I don't think the chronology works, especially since *kwalmaz has no Scandinavian descendants.) --Tropylium (talk) 15:58, 28 January 2017 (UTC)

Proto-Uralic verb endings

Wikipedia doesn't have anything at all on Proto-Uralic verbs, so I wonder what you know about the endings. Specifically the 3rd-person singular ending. If I'm not mistaken, this form was endingless?

I ask because I noticed that we have a verb at *kuńśe, without the following hyphen, rather than at *kuńśe-. If the form is indeed endingless, then we could lemmatise all verbs this way and get rid of the hyphen. —CodeCat 14:28, 4 February 2017 (UTC)

Probably endingless, yes. I'm not sure if getting rid of hyphens accomplishes much though; they currently seem to work for disambiguating verb and noun roots. (The difference is that the nom. sg. remains endingless everywhere in Uralic, while the 3P sg. in about half of the languages develops a suffix of some kind, e.g. *-ja in Samic and Mansi.) With cases like "urine, urinate", one option would be to use something like *kuńśe(-). --Tropylium (talk) 16:33, 4 February 2017 (UTC)

There is already disambiguation by membership in Category:Proto-Uralic nouns and Category:Proto-Uralic verbs. I don't think it's particularly necessary that the lemma form reveals the part of speech. —CodeCat 17:21, 4 February 2017 (UTC)

Why not make two entries for *kuńśe and *kuńśe-? Crom daba (talk) 02:12, 5 February 2017 (UTC)

I agree with you, for example Proto Turkic has *doŋ meaning "frost" and *doŋ- "to freeze" and many examples like *gēč meaning "late" and *gēč- "to be late". I would create two entries and mention they are cognates. --Anylai (talk) 10:13, 5 February 2017 (UTC)

But in the case of Proto-Uralic, the form *kuńśe is both a noun form and a verb form. So the hyphen is actually misleading, it indicates that something extra must be attached, which is not true for the 3rd person singular nor for the nominative singular. —CodeCat 13:46, 5 February 2017 (UTC)

Proto-Uralic *päĺkä

Hey, do you have anything on this root? Its listed at Russian палка (palka) (which is interestingly well rooted in Slavic, there's {Serbo-Croatian opaliti (“to strike”) and palica (“bat”)). It looks a lot like Proto-Tungusic *palụka (“hammer”) (correspondences seem more or less regular except for Manchu) and Mongolian алх (alx) (it appears already in Persian 14th century glossaries so I'm presuming it's not a Russian loan).

I keep coming across these Slavic-Altaic parallels like *tolkъno, талх (talx) or *bykъ, *buka. Two of these preserve l so I suppose that these aren't merely Iranic (inherited) cognates.

Crom daba (talk) 11:08, 12 February 2017 (UTC)

What would that one be supposed to mean? There's a widespread root *pälkä (“thumb”), and UEW has a much less certain "*päĺkä" - which I suspect to be an accidental similarity, as it is reconstructed from just two branches which don't quite agree in form (*e versus *ä) or meaning ('foot' versus 'hoof'). In any case neither seems to be very close to "stick", "hammer", etc. A slightly better Uralic counterpart might be the verb *polke- (“to tread, stamp”).

Then there is the traditional comparison with other IE words, too; see परशु for some of that stuff.

— талх (talx) on first glance looks like some kind of an early loan from Slavic to Turkic (maybe through Bulghar), but don't trust me on that topic. --Tropylium (talk) 21:44, 12 February 2017 (UTC)

No idea, Simboyd added the comparison, I'll guess I'll have to wait until he responds. Doesn't look very promising if it's the foot/hoof thing.

*pälkä looks a bit like palac (and cognates), but you're surely familiar with this.

I know about the Sanskrit word, I wrote the Tungusic entry. There's also Turkish balta. If you want to drive yourself crackpot there's also Tagalog palakol and Malagasy faikona, not to mention Esperanto hakilo (with debuccalization and metathesis regular for Esperanto-Tungus cognates).

talqan is attested in 8th century Uyghur, a century before the Moravian mission so loaning from Slavic is unlikely. Clauson makes a native Turkish derivation of the word but it doesn't feel very convincing.

Crom daba (talk) 22:42, 12 February 2017 (UTC)

Reconstruction:Proto-Germanic/hakkōną

The Estonian and Livonian words have wildly different meanings from all the other Finnic languages. ("to start to do something" in Estonian, "to catch" in Livonian.) Here's the ETY entry page. It says: "Läänemeresoome tüve või ainult soome, karjala ja vadja vastet on peetud ka germaani laenuks, ← alggermaani *χakkōjan-". This means: "The Finnic root, or only the Finnish, Karelian(actually Livvi, I made this mistake myself for a long time), and Votic words have also been considered a Germanic loan ← Proto-Germanic *χakkōjan-" (which is *hakkōną in the style that wiktionary uses). The alternative etymology for Estonian is given as Baltic; compare Lithuanian "šokti", and Lativan "sākt", which means "to start".Strombones (talk) 12:45, 18 April 2017 (UTC)

Yes, fair enough. Some similar semantics are attested also from Karelian ('do intensely'), but I suppose it's possible this group and the 'chop' group are indeed two unrelated verbs. --Tropylium (talk) 17:11, 18 April 2017 (UTC)

kolmas

Are you sure the Ingrian form can't have arisen by analogy? The palatalisation in Veps is analogical too after all. —CodeCat 23:58, 4 May 2017 (UTC)

That would not be impossible, but it's what is actually argued in some sources I'm looking into. There turns out to be other evidence as well, e.g. kolmasi, kolmans, kolomansi, kolomanś in several small isolated groups of Eastern Finnish dialects. Note also the other Ingrian forms like neľľääz: these get the long vowel from a regular sound change *CV(C)CVCV > *CV(C)CVVCV, thus also from earlier *neljäsi. --Tropylium (talk) 00:07, 5 May 2017 (UTC)

But it rules out *neljänsi. —CodeCat 00:12, 5 May 2017 (UTC)

No, it means that *nc > *ns > *s (normal in Ingrian after unstressed vowels) was earlier than apocope. --Tropylium (talk) 00:16, 5 May 2017 (UTC)

But then why was n retained in "third"? —CodeCat 00:18, 5 May 2017 (UTC)

It might be just particular to one dialect; the Ingrian dialects are quite divergent from each other. I'm looking the big picture currently on how many ns forms are actually attested (if it's just '3rd' or more of them). --Tropylium (talk) 00:23, 5 May 2017 (UTC)

Seems that kolmans is indeed the only ordinal given from the Hevaha dialect in my dialect dictionary, which at least matches with the article I'm going off of. --Tropylium (talk) 01:25, 5 May 2017 (UTC)

Tuhat (1000)

Where does Finnish tuhat (*tušante-?) come from? The Baltic languages (e.g. tūkstantis) do not have an š in the word. Germanic has *þūsundī. The Germanic s could be borrowed into early Finnic š, but the second vowel is an u instead of a. --Muhaha (talk) 16:19, 12 May 2017 (UTC)

Normally it's considered to be from some early Baltic form that did have RUKI as expected (so *tu(k)štanti-, close to Old Prussian tūsimtons), but that is indeed a bit of a circular argument.

In principle a second approach is consonant gradation; it's sometimes assumed that there used to be also a gradation pattern *s : *z > *s : *h (to explain inflection patterns such as  : ). But there are no examples of this in a root syllable; e.g. the gen. of is not **pehän.

Another example where Finnic "shows RUKI" while Baltic itself doesn't is laiha. --Tropylium (talk) 17:49, 12 May 2017 (UTC)

The word mies has h in inflected forms where it's between vowels. Is the declension on the reconstruction page *mees, *meesen correct? --Muhaha (talk) 20:42, 12 May 2017 (UTC)

It's not what we actually see in any of the descendants AFAIK, but it's kind of what would be predicted. The *s : *h alternation is after all a case of suffixal gradation, which is conditioned by stress rather than syllable closure (= weak grades occur after unstressed syllables, strong grades after stressed syllables, i.e. basically identical to Verner's Law). On the other hand, if were to come from an earlier disyllable such as (this is suggested by the Germanic loan etymology) and only contracted to a monosyllable after suffixal gradation set in, then things will work fine. I should add sometime a reference to Viitso's article arguing basically this. --Tropylium (talk) 09:46, 13 May 2017 (UTC)

Hungarian definite/indefinite conjugation

What's the origin of the indefinite conjugation (-k, -sz/-l, etc.) of Hungarian? Is it from Proto-Uralic? Do other languages have similar separation of indefinite and definite conjugations? PIE has some kind of separation (*-mi, *-h₂e) but it's not exactly definite and indefinite. Is this more common in active-stative languages than in nominative-accusative? Also, the suffixes are very fusional type in a language that's usually called agglutinative (e.g. -sz is indefinite present indicative second person singular) . --Muhaha (talk) 21:31, 3 July 2017 (UTC)

No one knows for sure. There are definite / indefinite conjugation systems in a variety of Uralic languages (e.g. in the 3rd person in the Samoyedic languages; cross-indexed for almost all subject/object combinations in the Mordvinic languages), but they have fairly poor matches with the Hungarian system. --Tropylium (talk) 17:25, 4 July 2017 (UTC)

Basic lexical form of Mansi verbs

As soon as possible I am going to start adding also Mansi verbs to the Wiktionary. But I do not really know which grammatical form should be used as basic lexem. Russian, Finnish, Estonian, German verbs are given according their infinitive form, however, Hungarian verbs are given in their finite form (present singular/third person), since Hungarian finite (megy, eszik) and infinite forms (menni, enni) generally differ. The dictonary of Munkács and Kálmán [1] follows the Hungarian tradition, while Russian dictionaries follow their own one [2]. Which one should be chosen? --Martinus Poeta Juvenis (talk) 18:15, 26 July 2017 (UTC)

I would check a couple Russian-Mansi dictionaries and see what they do. If there is a de facto standard, we should probably follow that. The infinitive is in principle usable, but it's a bit lengthy for me to support unconditionally, and something like 3rd person singular would make a shorter lemma. --Tropylium (talk) 18:47, 26 July 2017 (UTC)

After a lookaround: the infinitive does seem to be the standard in all Northern Mansi reference works, so I suggest going with that. --Tropylium (talk) 14:25, 3 August 2017 (UTC)

Until this time, I've created both forms for the verbs [3] and [4]. Then, the current lemmas will be regarded as inflected forms of the future lemmas.

By the way, I think the application of the infinitive form is based only on the tradition of German/Russian lexicography. --Martinus Poeta Juvenis (talk) 09:40, 4 August 2017 (UTC)

Well, it's not as if there's any other tradition of Mansi lexicography out there… --Tropylium (talk) 16:40, 4 August 2017 (UTC)

Done. The infinitive form is and will be the lemma of the Mansi verb forms. --Martinus Poeta Juvenis (talk) 18:35, 4 August 2017 (UTC)

Sammallahti on Northern Sami

I've been looking for more resources on Northern Sami, phonology especially. Do you know an easy way to get this one? —Rua (mew) 15:05, 22 September 2017 (UTC)

Not really. I've fewer resources myself than I'd like, on dialect variation at least. To my knowledge there is no comprehensive comparative description of dialect variation so far at all. On individual varieties, I presume you have the "Topics in North Saami Phonology" paper on the Kautokeino dialect. In print there is Sammallahti's licenciate thesis monograph on the eastern Eanodat dialect. Anything else has to be dug up from journals.

The Saami Linguistics Online blog has some very old public-domain grammars etc. to peruse, but the only one useful for modern purposes might be Wiklund's Entwurf…, which covers several NS dialects (even then in phonetic rather than phonological transcription). --Tropylium (talk) 16:37, 22 September 2017 (UTC)

What'd I'd especially like to know is how on earth you're supposed to tell whether vowels are long or short. The orthography doesn't tell you in most cases. Since Proto-Samic vowel length was tied to quality, there must be some kind of regular/predictable process that changed the length in the development to Northern Sami. Do Sammallahti or any of your other sources say anything about this? —Rua (mew) 17:05, 22 September 2017 (UTC)

The details seem to vary a whole lot depending on the dialect. As far as stressed position goes though, a simple etymological approach predicts most of the cases:

• á (from *ā) is usually long;
• i e o u from the smoothing of *ie *ea *oa *ua are usually long;
• i u o from *i *u *o are usually short; short /e/ is very rare;
• diphthongs have no phonemic lenght contrast (short in closed syllables ~ long in open syllables).

That is, most variation in length involves either a (from *ë), or unstressed vowels. --Tropylium (talk) 18:43, 22 September 2017 (UTC)

"Usually" is kinda vague though. What are cases in which these general assumptions don't apply? Are they predictable? I'm hoping to learn enough about the phonology that I can start adding pronunciations. So far I've gone by the Kautokeino description that you mentioned, but it says very little about the conditioning and assumes that the length of vowels is known. It also distinguishes between short and long diphthongs, which you say are nonphonemic. —Rua (mew) 19:22, 22 September 2017 (UTC)

As far as I can tell they at least morphologically predictable, but the main thing is that the conditions vary by dialect. In most cases it's probably impossible to add pronunciations for Northern Sami without specifying also a locality, since there is no standard "newscaster" accent. E.g. čáhci can vary from Eastern Finnmark [tʃæːhtsiː] to western Finnmark [tʃaːhtsiː] to northern Torne Sami [tʃaːhtʃiː] to southern Torne Sami [tʃaːhtsie]. The second would be the form used by most speakers, but it's regardless still just the pronunciation of one dialect area. --Tropylium (talk) 19:41, 24 September 2017 (UTC)

Do the dialects vary even phonemically? All I am interested in adding is phonemic pronunciations for now. Also, I see that in Torne Sami you replaced hc with hč, but Wikipedia (sourced) actually says it's the other way around. —Rua (mew) 19:53, 24 September 2017 (UTC)

All I can say with certainty is that Sammallahti's phonemic analysis of Eastern Eanodat is substantially different from the Kautokeino description paper. At least some of it can be reduced to issues of analysis; e.g. are tn dn to be analyzed as clusters /tn dn/, as clusters /nʔn ʔn/, or as glottalized consonants /nˀː nˀ/? But it is also fairly obvious that e.g. the merger of overlong and long consonants in Eastern Finnmark is phonemic, as is the associated lengthening of short vowels. Similarly e.g. the qualitative consonant mergers:

• ŋ > nj (Western Finnmark, most Torne Sami)
• ŧ > ht (southern Torne Sami ) or > /s/ (Western Finnmark subdialects)
• đ > /r/ (Finnish Wedge)
• and, from the data I have (Sammallahti in The Saami Languages), Torne Sami dialects have affricate mergers in both directions: *č > c in Karesuando, *c > č (transcribed ć actually) in Finnish Wedge.

So there is no such thing as a neutral phonemic transcription of Northern Sami, any more than e.g. for English. If we want to indicate features that could be used to predict pronunciation (morphophonemic level?), taking a step back towards the Nielsen orthography and distinguishing ē ō ī ū from ĕ ŏ ĭ ŭ, similar to CˈC versus CC, would do most of the work.

But maybe adding a few variant pronunciations would be a better solution. Including at least one of Kautokeino and Karasjok should be a shoo-in; maybe even both. Eventually also some particular Torne Valley pronunciation might be a good addition for at least some of the more divergent cases (but I have little idea where to get those from consistently). --Tropylium (talk) 08:21, 25 September 2017 (UTC)

My intention was to indicate the pronunciation in a maximally phonemic way. This would mean preserving ŋ, ŧ, đ, dj, č and c as all being distinct, and the pronunciations of the individual dialects could be derived from these via regular rules. If there is no dialect that preserves all of these, then maybe it's somewhat hopeless. Does it make sense to indicate some kind of abstract pan-Northern Sami pronunciation, the way the orthography does, or would such a thing be useless in practice?

Aside from all this, I'd love it if more of these details were added to Wikipedia. I don't have access to the sources you have, so Wikipedia would be a way to share some of it. —Rua (mew) 12:15, 25 September 2017 (UTC)

Template:se-IPA

I created this to automatically generate IPA pronunciations based on the spelling. Long vowels (if they are short by default) and overlong consonants need to be indicated with a '. It works well, but it has some limitations/missing features for now: it doesn't show secondary stress, and it doesn't handle consonant clusters. It only shows the form of the Kautokeino dialect. I can also provide no guarantees for words of more than two syllables, since I don't know how consonant clusters after unstressed syllables behave or how it works with respect to vowel length. It does shorten vowels and diphthongs in the appropriate environments, however, based on the rules I added to the Wikipedia article.

Any information you can give on other dialects and long vowels to improve it would be welcome. —Rua (mew) 19:37, 30 September 2017 (UTC)

Monosyllables in Proto-Samic

I'm a bit puzzled by the behaviour of some monosyllabic words in Proto-Samic. The stem of the negative verb ought to be *ea-, and Northern Sami indeed reflects that, but only in the dual and plural. In the singular, for some reason, the stem is i-, and it appears that way in other Sami languages as well. Southern Sami also has *i- in the 2du, 1pl and 2pl. Skolt Sami has jiõm, jiõk and ij as the three singular forms. Are we dealing with Proto-Samic *jiem, *jiek for the 1sg and 2sg, with some kind of regular shortening rule in monosyllables? What about the 3sg? It has a weird ending *-j that I don't recognise.

The interrogatives mii and gii are similarly puzzling. Their stems ought to be *më- and *kea- respectively, but the nominative singular forms are odd. The plural personal pronouns seem to have -i- throughout, but that's certainly not the pre-Samic situation. Is there a sound law I'm missing here? —Rua (mew) 20:42, 11 October 2017 (UTC)

Notation on Álgu

The link to Álgu you have given me is quite useful, but I'm having some difficulty figuring out how it works, still. Apparently, things are presented in a variety of orthographies, not all of which are straightforward to convert to the standard one in a given language. In any case, I was wondering what the dots mean that are placed underneath vowels. According to the FAQ, it's to indicate that it's short, but that seems rather redundant as the vowels they're placed under are already short. E.g. in sáhkọhit, is there any reason to assume the o would be long if not for the dot? And how does the unstressed o in gōrodit differ from this special ọ?

There's also a ḥ mentioned in the FAQ, which apparently indicates dialectal pronunciation differences. Do you know what those differences actually are, so that they may be incorporated into the Wikipedia article and {{se-IPA}} sometime? —Rua (mew) 15:31, 27 October 2017 (UTC)

Odd-syllable (stressed) o is usually short, but even-syllable (unstressed) o and e are usually long; their typical sources are overlong contracted *û and *î. ḥ apparently indicates earlier *ht; I do not know what the dialect variation for this is. --Tropylium (talk) 19:56, 27 October 2017 (UTC)

According to Sammallahti in "The Saamic Languages", the vowels that originally triggered diphthong simplification were short i and u, which were in turn created from ē and ō by a following j and by a following ë respectively. Where contraction occurs, he indicates the late Proto-Samic vowel in the resulting syllable as short, identical with any other short i and u. In modern Northern Sami these short vowels are lowered to e and o except when j follows, while Lule Sami preserves the original height. In neither language is there anything to hint that these vowels are long. Moreover, Sammallahti says that the only long vowels that occur in unstressed syllables are i, á and u, along with the full complement of short vowels. —Rua (mew) 20:17, 27 October 2017 (UTC)

Hmm, sorry, I'm confusing a few things with the older orthography (where unstressed /i e o u/ are spelled as e i u o respectively). On re-reading, this appears to match with the old Nielsen orthography's contrast between ĕ ŏ (= Álgu's ẹ ọ?) and i u (= current e o), though they are (in most dialects?) pronounced the same. The difference is primarily etymological: i u come from *i *u (including *î *û), while ĕ ŏ occur in allegro forms (for *ē *ō) and recent loanwords (in this case sáhkọhit ← Fi sakottaa).

More relevant for the modern language, however, is that the latter fail to trigger diphthong smoothing: e.g. guoktẹ (Nielsen guoktĕ) and not ˣgukte, versus gorrot (Nielsen gorrut) and not ˣgoarrọt.

(As an aside, I keep thinking that w:Northern Sami orthography needs a comparison table between the most widely used schemes; but I keep wishing to first find a source that already does a part of this, since it'd be a fair bit of work to do completely from scratch.) --Tropylium (talk) 21:42, 27 October 2017 (UTC)

Thank you for clarifying. The reason I asked about the pronunciation is that Álgu indicates them, but I can't really imagine that they're doing it just to indicate etymology. Some Álgu entries have ạ, which isn't even relevant to diphthong simplification (or smoothing as you call it), so there must be some other difference going on. —Rua (mew) 21:48, 27 October 2017 (UTC)

It's not strictly etymological as much as backwards compatibility (to a system that included some etymological considerations). As for ạ, this is a similar story. Here's a sample from Knut Bergland's chapter on Sami in On Simplifying of the FU transcription (Helsinki 1973):

As noted above (…), between Nielsen's i, u, â and ĕ, ŏ, ă there is no phonetical difference directly. Thie distinction was introduced primarily in order to make it easier for Norwegians to learn Lapp grammar: according to Nielsen's rule, ie, uo, æ, oa are "changed" into ī/i, ū/u, ē/e, ǭ/ǫ (before x, xx, xy / xˈx, xˈy) before an i or u of the following syllable but not before ĕ, e or ŏ, o, or â etc. The actual distributional rule, of course, is that ie, uo, æ, oa occur before all vowel phonemes of the following syllable, whereas ī, ū, ē, ǭ occur only before the short (and half-short) vowel phonemes, so that only in the latter position there is an opposition between ie and ī etc. (this, too, is a simplified account). This opposition is sufficiently taken care of by itself, so so far there is no need to distinguish between i, u and ĕ, ŏ. But Nielsen's distinction also indirectly marks a quantitative difference which, according to Nielsen, obtains under certain conditions in the first syllable, the "bright a", for example, being (before consonants other than xˈx, xˈy) shorter in those cases where Nielsen writes ă, e.g. KN akkălâgges [= modern áhkalạkkis] etc. 'Greenland shark' vs. bākkâdâdˈdât [= modern báhkadattat] etc. 'heat several times etc.' In the case of the "bright a" and of æ the difference is marked also in the first syllable, but in the case of â, ie etc. the notation is indirect only. This is not to critizize Nielsen's orthography, which may be good for the purpose it was designed for. But for a transcription designed for purely scientific purposes it seems preferable mark the differences where they are actually found. Nielsen's description of the quantity system of the Finnmark Lapp dialects probably needs checking with modern tools (if it is still possible, for the dialects may also have changed since the days of Nielsen's field work). The vocalism of the dialect of Karasjok, for example, may turn out to be less complicated, and less "historically regular", than Nielsen conceived it. But it seems fairly certain that both the "bright a" and the "dark a" have in a first syllable two phonemically different quantitative varieties, whereas the difference between the two sets may be basically qualitative. Thus, perhaps, all the vowels of the first syllable need a quantitative mark, either the comparatively short ones or the comparatively long ones, for example ŭo vs. uo, or uo vs. u˙o, and â̆ vs. â, or â vs. â˙, or the like.

I have no idea where ạ is to be marked — if this is predictable at all. --Tropylium (talk) 13:04, 28 October 2017 (UTC)

I've added support for the underdots to {{se-IPA}} and it correctly modifies any preceding diphthongs, see ađaheapme. I'm not sure how it should be notated for any other vowels, or whether it affects the consonants as well. —Rua (mew) 13:26, 28 October 2017 (UTC)

Another Álgu question. Since the lists of word families often have terms written in linguistic orthography, that isn't particularly useful to me if I want to add these terms as descendants. Is there any way to get a term in the official orthography of the language (where one exists)? —Rua (mew) 13:26, 28 October 2017 (UTC)

Grouping descendants by language family

Hey, I think I might be the progenitor of this (harmful?) habit so I think I should try defending it. I feel that this increases readability for large lists while the danger of confusion is relatively small, for example.

• → Proto-Foo: *bar
  • Qux: baz

As opposed to

• Foo:
  • → Qux: baz

Of course for small lists this would just be clutter, but one should apply editorial discretion to these things. Crom daba (talk) 16:29, 5 November 2017 (UTC)

I can see some benefits to this, but not for extending this to "macro-families" if a term is not even reconstructible for its subfamilies (e.g. it would not make sense to group two recent loanwords in English and French as "being into Indo-European"). Secondly, "Dagestanian" is not a family at all; the term "Nakh-Dagestanian" is built in a similar way as "Indo-European" or "Hamito-Semitic". --Tropylium (talk) 19:23, 5 November 2017 (UTC)

I agree in principle but in this case Northeast-Caucasian seems like a handy delineation of geography and typology.

Breaking it down into subfamilies would still leave readers wondering if it is NEC or Pamir languages or something third, which is what we're trying to avoid.

(also I thought that Nakh is basal to NEC, but checking Wikipedia this is apparently controversial.) Crom daba (talk) 21:05, 5 November 2017 (UTC)

"which is what we're trying to avoid" — We are? We're not really a source on language taxonomy. If a user doesn't know what the affiliation of a language is, they will be able to click thru to Wikipedia on the main category page on here, or better yet, thru {{cog}} and similar, to read more.

What I would like to avoid instead is the implication that all NEC languages have borrowed their term for "sand" from Turkic; whether this was through PNEC or later on. If neither of these is the case, I don't think we should be using language family labels for lists. (I'm okay with such labels if a given word in fact did get introduced everywhere applicable; we even already sometimes reconstruct such words for proto-languages.) --Tropylium (talk) 21:17, 5 November 2017 (UTC)

I guess we disagree then, maybe I'll start a thread to see if anyone else supports this. Crom daba (talk) 21:40, 5 November 2017 (UTC)

The two kinds of s-suffixes on Samic attributive adjectives

I've been doing a bit more work on adjectives in the last few days and noticed something curious. Northern Sami has many attributive adjectives that end with -s, but these seem to be actually two different suffixes. The first, exemplified by čeahppi, triggers a weak grade and has a shortened ẹ in the second syllable that does not trigger diphthong simplification. The second, exemplified by suohkat and ruoksat, triggers an extra-strong (Q3) grade and has e from older i, which triggers diphthong simplification.

To explain the weak grade of the first variant the syllable must have been closed, while the Q3 grade and *i of the second must result from contraction. However, it appears that both suffixes originally had a final *-ē, since Kildin Sami has final palatalisation. So would the first suffix have been something like *-ssē, and the second perhaps *-jësē? Do you know anything more about these? —Rua (mew) 21:56, 15 November 2017 (UTC)

I don't think the palatalization in Kildin indicates lost original *ē; this e.g. ought to be otherwise be retained in Southern. As far as I know everyone has reconstructed this ending as just *-s.

For the latter type, Korhonen reconstructs *-ëtës > *-ëës > *-ês, but the assumed loss of medial *-t- (yet not in e.g. the infinitive *-tēk) is unclear to me. Samallahti seems to agree but doesn't clarify either.

The whole attributive system is a bit of underresearched anyway, from what I gather. I have a colleague who continues postdoctoral work on this, though; would you like me to put you in touch? --Tropylium (talk) 22:13, 15 November 2017 (UTC)

I suppose, if they can elaborate further on this subject. —Rua (mew) 22:17, 15 November 2017 (UTC)

Do you have any news on this? I'm interested to learn more about the contracted type of attributive. —Rua (mew) 21:03, 28 December 2017 (UTC)

Also, I have now created Proto-Samic entries for clear-cut cases like *ručkëtē. However, the old entry *ručkë is now pretty much useless, as it has no direct descendants, only a derived term. Should we get rid of it? —Rua (mew) 21:08, 28 December 2017 (UTC)

Reconstruction:Proto-Finnic/keerä

Reconstruction:Proto-Finnic/keerä: "Etymoloygy Unclear. Has been compared with words in other Uralic languages, e.g. Hungarian kerül, but the long vowel cannot be inferred from the reconstructed Proto-Uralic form *kerä." En tiedä, miten kielitieteitä pitäisi ajatella, mutta meillähän on ainakin sellaiset sanat kuin keruu ja kerjuu, joissa on pitkä vokaali. Ja voisiko johonkin väliin mahtua sellainen kuin kehrtädäk? Saataisiin kehruu ja kehrä, eli auringonkehrä, auringonkehä, päivänkehrä, päivänkehä mukaan tähän rinkiin... ja mukaan voitaisiin ottaa vielä kerjätä < PFU *kerä- (vrt. kerjuu, kierjuu? ja kiertolainen, kierre). Ja kerta (kertoa, kerrata, kerrasto): "Euroopassa alettiin käyttää 1400–1500-luvulla langan kehräämiseen ja kertomiseen rukkia (Viires, 2006, s. 180). Aluksi rukinpyörää kierrettiin käsin, myöhemmin yleistyi poljettava rukki." Villakäsityöt Itämeren saarilla --Liedes (talk) 14:55, 15 December 2017 (UTC)

Hei!

Odottamaton "pitkävokaalisuus" viittaa luonnollisesti 1. tavun vokaaliin *keerä (suomessa diftongi kierä), ei 2. tavun. Keruu ja kerjuu eivät muutenkaan ole suoraan liitettävissä "kiertää"-sanueeseen; ensinkään merkityksenkään nojalla; toisekseen muoto-oppinsa (nämähän ovat nominijohdoksia verbeistä kerätä ja kerjätä, eivät kantasanoja). Merkitykseltään läheisemmät kehrä(tä), kehä, kerta ovat kaikki nekin eri sanueita äänneasunsa nojalla. Epäilisin että vaikutus on ollut pikemmin päinvastaista: samantapainen äänneasu on johtanut merkitystenkin lähestymiseen.

Osa näistä sanueista saattaa kyllä olla jossain varhaisissa yhteydessä toisiinsa, mutta suhdetta *keerä / kerül ne eivät mitenkään selitä, pikemmin vain sekoittavat pakkaa. Kuulostaa tosiaan oikeastaan siltä, että et juurikaan tunne äännehistorian perusteita (joiden perusteellinen tässä selvittäminen olisi ikävä kyllä vähän turhan pitkä projekti). **kehträ- ei ensinnäkään edes ole mahdollinen kantasuomen rekonstruktio; toisekseen sillä pitäisi olla vain yksi vastine, ei kolmea-neljää; ja kolmanneksi muutoksesta *eh(t) > *ee ei ole mitään muita todisteita (esim. meillä on lehmä mutta ei **leemä > **liemä, tai ehto mutta ei **eeto > **ieto, tai lehti mutta ei **leeti > **lieti.)

En ole varma vastasiko tämä kysymykseesi, mutta ehkä tästä jotain apua oli. --Tropylium (talk) 18:16, 15 December 2017 (UTC)

*hepoi

The issue is we should rather have *hopoi, but that would be even more drastic change from me. Liedes (talk) 11:23, 12 January 2018 (UTC)

No, the basic word for 'horse' has quite clearly e in the Finnish through Veps range, and also as expected, õ in (Courland) Livonian and partly Votic. You are probably going by the Estonian words, but it is considered a late separate development, similar to kollane from *kõllane < , or kord from *kõrd < . Phonetically a likely explanation in this case is that o in the second syllable has colored earlier õ in the first syllable to also o (so a similar case as Finnish pisty becoming pysty in most dialects).

We are not opposed entirely at Wiktionary to editors making their own etymological commentary, or to including rare poetic/dialectal terms, but you will need to understand what you're doing. I would recommend e.g. starting by writing the actual entry for a rare word first and worrying about its etymology only after that; and if it's not found in common etymology sources, it's perhaps better to leave it off of reconstruction entries, unless you know what sound changes are expected in the language. --Tropylium (talk) 11:48, 12 January 2018 (UTC)

Thanks alot. --Liedes (talk) 21:52, 26 January 2018 (UTC)

*kiwe

Is Uralonet considered a secondary source not to be duplicated as such? I mean, the entries you created differed in many places from what I produced on the other page earlier. Is that just a different ortography in Uralonet, or is it inferior or outdated information? --Liedes (talk) 21:52, 26 January 2018 (UTC)

On the first: I suppose we could use Uralonet also as a primary source, but it is so far only an online duplicate of UEW with no updated information. I prefer giving UEW as the primary source, to show that its etymological etc. information is ~30 years old.

As for orthography, there are lots of transcription variants on the Uralic languages spoken in Russia. We should be mostly using the official Cyrillic spelling on Wiktionary, not any of the Latin transcription schemes. (If you do not have any sources on these, you can leave a transcription request by using the tr= parameter; so e.g. {{m|udm|tr=ke̮}} if you have an Udmurt term given just as ke̮.) But yes, the UEW / Uralonet primary language data is still valid and can be in many cases mechanically retranscribed (e.g. in the case of Udmurt vowels: e̮ → ӧ, i̮ → ы). --Tropylium (talk) 23:11, 26 January 2018 (UTC)

Admin?

You have a pretty good understanding of our policies, are responsible and sensible, and you already do a lot of excellent work improving and protecting our etymologies. I think you would be an excellent admin.

I know that your plate is full offwiki, so I'm not suggesting this to get you more involved here. I just think that having admin tools would make what you're doing anyway much more effective. Aside from patrolling and rollbacking tools (which can be granted without adminship), it gives you the ability to block, delete, move without leaving redirects, change/bypass protections, and to view/restore deleted entries/edits.

As a bureaucrat, I try to avoid getting involved directly in the selection of admins, but this seems so obvious that I would happily make my first-ever nomination for adminship on your behalf.

If you're not interested in that, we could at least make you a rollbacker and/or patroller: that way when you deal with edits in Recent Changes you can mark them as patrolled so others don't need to check them.

Let me know if you're interested. If you are, I'll set up the vote, and all you have to do is indicate that you accept the nomination on the vote page. Chuck Entz (talk) 20:08, 13 April 2018 (UTC)

siveä, sivistys

The word siveä comes Proto-Uralic *śiwa according to the page. Is it a borrowing from PIE ḱey-wo-? --Muhaha (talk) 10:53, 1 May 2018 (UTC)

I think this is likely. At least {{R:Katz 2003}} proposes this etymology, for Finnic only though. I don't recall what, if anything, he says about the proposed cognates in Khanty. --Tropylium (talk) 16:11, 2 May 2018 (UTC)

a > ea / _ś[σ

I didn't know of this law, and it's not mentioned on w:Proto-Samic#From Proto-Uralic. Could you add it? —Rua (mew) 16:51, 10 March 2020 (UTC)

päivä, päivyt

I appreciate your help, but on the other hand the edits raise questions. What is this suffix -yt? What does it mean? The Finnish suffix doesn't seem to satisfy the semantical change "day" > "sun". Could you explain this? Thadh (talk) 21:04, 18 August 2020 (UTC)

Honestly I don't know if -ut, -yt has any active specialized uses in Ingrian, just that it is really very productive for deminutives across Ingrian / Karelian / Ludian / Veps. It's not even clear to me if there are any up-to-modern-standards descriptive sources on word derivation in the smaller Finnic languages. E.g. E. V. Ahtia's posthumously published booklet on derivation in Karelian only lists it as one of 23 (!) undifferentiated deminutive suffixes. Semantics-wise though this shift makes sense to me: "the little shiney" i.e. just the actual light source rather than the entire state of lightness. Estonian does the same too, FWIW (päike).

There's also a homophonous pan-Finnic adjective formant (in e.g. kevyt, lyhyt). Aside from the identical shape, I don't actually know of any reasons to think that this would have to be the same suffix etymologically though, so this may not matter. --Tropylium (talk) 21:57, 18 August 2020 (UTC)

Now you say it, "the little shiny" does sound logical. Thanks! I've gone on and changed the suffix to -ut instead, as to avoid two separate categories for suffix endings in Ingrian (and thus avoid inconsistency). And as of my observations following the ittseopastaja, Ingrian suffixes are mostly akin to Finnish (except perhaps -kkoi, although I am considering cognancy with -kki). Thadh (talk) 07:18, 19 August 2020 (UTC)

I'd indeed recommend following a similar front variant of… formatting for vowel-harmonic suffixes across vowel-harmonic languages in general.

-kkoi looks like it might be from agglutination of akka + -oi; I wonder if there are examples where the first -a- still survives (i.e. along the lines of Eesti > eestakkoi)? But yeah, "Finnish" covers such a large palette of Finnic varieties that I wouldn't expect there to be many suffixes at all in the closest relatives Ingrian / Karelian that wouldn't have cognates at least in the eastern dialects. In Karelian there's the reflective conjugation, but IIRC Ingrian does not have that… --Tropylium (talk) 15:37, 19 August 2020 (UTC)

Question about the usage of "√" for roots

Hello, I noticed your recent use of "√" in a Etymology Scriptorium post and have seen in a few other places, though haven't ever been quite sure what it meant. Notably, at the page for √ the related meaning doesn't seem present. Would be willing to tell me how you use "√" for linguistics and whether or not you know if the usage is translingual? Best. —The Editor's Apprentice (talk) 17:41, 15 December 2020 (UTC)

Proto-Permic

Hi, I was wondering if you could help me figuring out how to handle Proto-Permic best on wiktionary. So, my first compulsion had been to just follow Csúcs's book and stick to it, but considering what Ante Aikio wrote about the book in his 2021 review of Permic etymologies, I'm not so sure that was a great decision.

We could probably follow Lytkin's model, which is kind of ideal, since I'm making Proto-Permic entries along with Komi-Zyrian ones (this way I could use the reconstructions in his etymological dictionary), but of course we'd probably need some character substitutions (using carons below doesn't seem practical). But I'm wondering what you think! Thadh (talk) 22:18, 11 November 2021 (UTC)

Yeah, as Aikio says there's basically no consensus on Proto-Permic vocalism; ask ten researchers and you'll get twelve reconstructions. But I would not recommend Csúcs as the starting point, his reconstruction is phonologically clearly the worst in the last 30 years. The most recent take is Metsäranta's thesis from 2020 [5] which uses (unfortunately does not explicitly present) a relatively conservative system close to Lytkin. For the sake of being able to accommodate future updates, we probably should prefer a system that has more distinctions rather than fewer, so Lytkin might be indeed the best option. His is also clearly the system that has been argued in the greatest detail.

On the open/close-mid vowels, I do like Csúcs' notation of *ɔ *o for earlier *o̭ *o (followed by Zhivlov 2010 and Metsäranta; I used this myself in a short paper on Permic recently too), but this would need to be thought over if we wanted to generalize it, e.g. on how we deal with Proto-Komi / Old Komi, or the *ḙ *e *ɛ triplet (could be turned into *e *ɛ *ä perhaps). And of course whatever we might decide probably should be presented in some detail at the expected PP considerations page. --Tropylium (talk) 14:37, 12 November 2021 (UTC)

I have a started page at User:Thadh/Proto-Permic which summarises Csúcs's notations (of course we'll need to at least add ȯ̭ and remove u̇, but it's a start), feel free to edit it! See there for some proposals on further simplification of characters on my side. I think that using ɔ and ɛ for the open-mid vowels is a good idea.

Unrelated to the vocalism, in how far is Csúcs a good source for morphological reconstructions? I don't think I know of any other attempts at reconstructing PP case morphology for instance.

And finally, we probably want to make the codes for Komi-Yazva and Old Komi. Being as complete as possible from the start seems to be a good idea when talking about wiki work. Thadh (talk) 17:37, 12 November 2021 (UTC)

Huh, we didn't even have a code Jazva yet? And yes Old Komi seems like a good idea (already for the native alphabet).

On morphology I would probably call Csúcs the most reliable single source, though there's plenty of literature on individual details. --Tropylium (talk) 22:15, 13 November 2021 (UTC)

I've updated the overview with Lytkin's vowels. I have two questions left: 1) Should we use *å or better stick with *a? 2) Do you agree with all the modifications as given?

I'll create an RFM for Komi-Yazva and Old Komi. Do you think we should lemmatise at Anbur or Cyrillic? And where do you think we should set the upper border for the language: 1700, 1917? Thadh (talk) 12:42, 15 November 2021 (UTC)

About Proto-Samoyedic *tättǝ

Hi! I have some questions about Proto-Samoyedic reconstruction of *tättǝ: in Juhanen's "Samojedischer Wortschatz" it's *tettǝ. So I wonder, is everyone (Wikipedia, Wiktionary) wrong or am I just miss something? Maybe it's some PSa reconstruction features? Kaarkemhveel (talk) 15:55, 17 March 2023 (UTC)

This is a recalibration of Proto-Samoyedic, proposed by Helimski in 2005 and since then adopted basically universally in Samoyedology:

• Janhunen's *i is to be split in *i and *e, continuing Proto-Uralic *i/*ü and *e, respectively (e.g. *pir (“tall”) from *pide, but *wet (“water”) from *wete);
• Janhunen's *e (in initial syllables) is to be reassigned as *ä, which is also its Proto-Uralic source (e.g. *käəj (“tongue”) from *käle);
• Janhunen's *ä (in initial syllables) is to be reassigned as *a, which is also its most typical Proto-Uralic source (e.g. *kajwə- (“to dig”) from *kajwa-).

Our current WT entries *wit, *kitä, *keəj and *näkər (“three”) probably should be accordingly moved — the last actually to *nakur with also unstressed *u, another feature that has been reconstructed for PS since 1977.

For reading on the various updates on PS reconstruction, the best single starting point might be Salminen 2012; though quite a bit has happened really and there is a clear need in the field for some kind of an updated source (as it happens I am myself working on publishing one, "soon™").

--Tropylium (talk) 17:13, 17 March 2023 (UTC)

Thank you, it helped a lot. I'll move these pages and now I can finally add some of the "questionable" entries (like *tättə).

P.S. I'll be waiting for updated source! :) Kaarkemhveel (talk) 19:28, 19 March 2023 (UTC)

Also, about "*pir (“tall”) from *pide": isn't it *pirë (Jah)/*pirә (Sal)? Kaarkemhveel (talk) 15:26, 20 March 2023 (UTC)

Right, I'm misrecalling slightly; as Janhunen reconstructs, *pir looks to have been essentially a bound root from which *pir-ə 'height' and *pir-kä 'tall' are derived (neither can be derived directly from the other). We might not need a separate entry for that and could just refer the derivatives to the PU root.

(Curiously, as also noted by Janhunen, a similar situation appears in Finnic too: e.g. Finnish pit-uus (“length”), pit-kä (“long”), also pite-mpi (“longer”), but a bare **pite- does not occur. There might have been some paradigm like this already in Proto-Uralic, but full details await working out, maybe above all what the function of *-kä would have been here.)

-Tropylium (talk) 21:06, 21 March 2023 (UTC)