Why can't linguist just write it in a simple way instead of trying to be Elon Musk naming his children?
>>18718737Because they dont know.
>>18718737/ħwə're:ks/
>>18718737*h3rḗǵs is fairly abstract. Let's start with laryngeals:In early PIE *h2 was plausibly */χː/ and *h3 */χ/ *[χ ~ ʁ]and perhaps in later PIE *h2 = */χ/ and *h3 = */ʁ/ (unclear).For Pre-Proto -Greek, -Armenian, and -Tocharian, probably*h2 = */ħ/ and *h3 = */ʕ/.The vowel *e was probably */ɛ/.*ǵ was plausibly something like */gʲ/.However, the identification of the laryngeal in this word is not secure since the languages involved do not show any vowel reflex for a laryngeal in word initial *HC-. Some linguists such as Kim McCone have proposed *h2. I prefer *h2rḗǵs.In IPA this would be */χːrɛːgʲs/ *[χːrɛːkʲs].
>>18718780>*ǵ was plausibly something like */gʲ/.Although some linguists fancy *ǵ = */g/ and *g = /G/. I need to read up on this again as I'm not sure how well the uvular plosive theory is holding up right now.
>>18718780>*/χːrɛːgʲs/I almost forgot. *ǵ could be an implosive phonemically (although a particular allophone may be chosen in *-ǵs#).Later PIE might not have implosives though. It looks like there might have been a move toward glottal constriction in the "voiced" series of stops.
>>18718780>/χ/What,?
>>18718899That's an uvular fricative.
>>18718780In English, doc.
>>18718977NTA but:There were three sounds in protoindoeuropean that we know were there but we don't know what exactly they were. Usually for sound reconstruction we can look at words like pater, πατηρ, pitar, and father and see that three languages kept the p and one "lenited" the p for an f sound. Therefore the original word began with p, not f. However there are some vowel reconstructions that really stumped everybody for a few decades until de Saussure came up with the idea of laryngeals: guttural sounds that "colored" vowel sounds around them. For example in the pater/πατηρ/pitar/father set, three languages have the sound "a" in the first syllable while sanskrit has an i. However the vast majority of words that have a's in greek and latin and the rest correspond to words with a's in sanskrit, so there must be something special about the first vowel in pitar. Now it is commonly accepted that the vowel in question comes from a syllablic laryngeal that evolved into "a" in most branches of indoeuropean but "i" in Sanskrit. The laryngeal theory has been the dominant one since Hittite documents were deciphered that showed guttural sounds exactly where de Saussure had expected them.
>>18718999>three languages kept the p and one "lenited" the p for an f sound. Therefore the original word began with p, not f.There doesn’t seem to be any reason to believe this. I don’t understand. How can someone not just say it actually began as f?
You read the H 3 times
>>18719005You can say whatever you want, but you have to provide convincing evidence and the simpler explanation is generally better. You could say the original sound was a voiceless 'w' sound, but then you would have to explain why it became 'f' in one branch of the languages and 'p' in the rest of them.p > pp > pp > pp > fis simpler than f > pf > pf > pf > fIf you look at Grimm's Law, the letters that in non Germanic languages tend to be unvoiced plosives (p, t, k) in Germanic languages are unvoiced fricatives (f, θ, h). That the same phenomenon happens to multiple sounds suggests a regular sound change in a certain direction. It is all hypothetical, of course, until we find written evidence of PIE, but it is the best explanation thus far.
>>18719030>It is all hypotheticalSo? Its all fake? Literally trust me bro?
>Why isn't the 5300 year old reconstructed language easy to learn?Better question is why has there been a sudden surge of bots asking this retard faggot question
>>18719146>>18719143Yes. Cope. Linguistics are basically like>my personal theory
>>18719143>Literally trust me bro>t. Every historian ever.As >>18718999 said, hypotheses are formulated, and then newly discovered languages/texts can demonstrate evidence for the hypothesis, as with laryngeals and Hittite. A similar thing happened with Greek and the digamma. Words in Homer were believed to originally contain 'w' (eg 'wanax' for 'anax,' 'lord'), and Mycenaean texts in Linear B show the presence of the 'w' ('wa-na-ka'), according to the prevailing interpretation of Linear B texts. Of course, someone could present a new hypothesis for Linear B or laryngeals or Proto Indo European, but if you formulate a coherent hypothesis, and then new evidence seems to confirm it, that is pretty convincing.
>>18719143No, none of it is fake. For 400 years now, people have closely studied thousands of years of evidence on the evolution of indo european languages, and they've discovered consistent patterns that are never violated.All the evidence from Germanic sources - Gothic, Norse, German, etc. tells us that p evolved into f, k evoled into h, et cetera. These patterns are universal among consonants. f almost never evolves into p, i've never heard of an example of it, ever. Maybe you can find some obscure edge case but it's dominated by p evolving into f.In the age of AI you could look up a neat summary of all of this at a moment's notice, the only reason to post like this is to troll.
>>18719148Gravity is fake, your personal theory is just cope. Sure you can find some examples of the apple falling from the tree, but I can also just as easily find examples of apples rising up from the ground when an earthquake happened or the tectonic plates converged
How did any of you even pass high school math if you can't handle seeing variables
>>18719153>f almost never evolves into p, i've never heard of an example of it, ever. Maybe you can find some obscure edge case but it's dominated by p evolving into f.https://chridd.nfshost.com/diachronica/search?q=p A few instances, but yes, overwhelmingly p > f.
>>18719160according to your database there, there are quite literally zero verified examples of an f to p sound change. Did you even check it?
>>18719162Did you scroll down to the 'to /p/' section? Literally the first entry is f > p.Here are the examples, since you are too lazy to check:>Proto Afro Asiatic to Proto Osmotic>Proto Boreafrasian to Egypto Berber>Classical Arabic to Cypriot Arabic>Proto Turkic to Proto Kypchak>Proto Erromango to Sye>Old Norse to Shetland Norn>Proto Mande to Southwest MandeThere are also quite a few examples of φ to p and v to p.
>>18718737>*h3rḗǵs"“HEH-rays”
>>18719170Yeah, that's what I'm talking about, did you actually read those examples and read what's written about them? In order, you have1. an unverified reconstruction2. another unverified reconstruction from the same guy3. f to p only as parts of other fricatives in arabic, "f theta" to "p t", so not at all the f to p in the Indo-European consonant sense we're talking about4. Another unverified entry from someone's personal wiki reconstruction, not an actual historical example, and it's not even clear if it's f to p or something that is phonologically like f, and apparently the p is sometimes an intervocalic w?5. Another random reconstruction (Proto-Erromango??? this is not real)6. Not f to p, it's ft to pt7. Someone's reconstruction of an arabic dialectThis is the problem with wiki's and newfag internet in general. People are more interesting in trying to win an argument so they just stumble blindly through resources they don't understand. No, a random guy's reconstruction of "Erromango" or reconstructed Turkic "Kypchak" are not examples of an actual language changing f to p.
Like I thought you were going to give me an example of an obscure, barely known Indo-European word that changes f to p. Instead you're citing a random dude's personal reconstruction of "Proto-Boreafrasian" (what even is this? what is the boreafrasian language family, the language of the fucking hottentots?) without any context
People think linguistics isn't a serious field because their idea of linguistics is a random dude deciding at his own discretion that "Proto-Boreafrasian" had a certain way of pronouncing its fricatives because he felt like it.
>>18718780Let's take a look at few words:ἄρχων "ruler, commander"ἀρχός "leader, chief, ruler"Armenian arkʿay "king"Latin rēx "king"Is there any way to connect these words? If you assume rēx < *h2rḗǵs, then perhaps we have a root *h2reǵ- ~ *h2erǵ- (by schwebeablaut). If there was an s-stem *h2érǵos ~ *h2érǵes-, then by adding a suffix *-kó- ~ *-kéh2- we would arrive at*h2r̩ǵ-s-kó-*h2r̩ǵ-s-kéh2-*ǵ would be lost regularly in this cluster, then so would *s in *-rsC, and the *k would become aspirated due to this cluster.*h2r̩-kʰó- > ἀρχός*h2r̩-kʰéh2-Then *h2r̩-kʰéh2- is the stem behind *h2r̩kʰéh2-ti- > arkʿayStefan Schaffner explains everything necessary for this except for connecting the Armenian and Greek words to the root *h2erǵ-, which is an attractive option considering the proposed clusters would be reduced by regular sound change.
>>18719219
>>18719205>>18719204Don't forget that this is the same guy who creates personal charts and tables. He, for example, believes in Indo-Uralic and that Sumerian has Proto-Indo-European words.
>>18719235different poster
>>18718999>Hittite documents were deciphered that showed guttural soundshow did we figure out the sound values?>Mycenaean texts in Linear B show the presence of the 'w' ('wa-na-ka'),again, how do we know the sound values? feels like a huge arena of circular reasoning.>according to the prevailing interpretation of Linear B textsah, opinion sciences, majority opiniton, voting as a way to find out truth.
>>18718977Proto-Indo-European had consonants called "laryngeals" which I would liken to the class of fricatives in Semitic called "gutturals". Their exact sound vowels are disputed, so they are referred to abstractly with the symbols*h1*h2*h3These laryngeals are identified by how they modified the pronunciation of adjacent vowels*h1e = *[h1e]*h2e = *[h2a]*h3e = *[h3o]Since *h1 has no effect on the vowel, it is commonly thought to be */h/ or */ʔ/.*h2 and *h3 are thought to be uvular (*/χ/ ~ */ʁ/) or pharyngeal (*/ħ/ ~ */ʕ/) fricatives, because these are the guttural fricatives which are cross-linguistically likely to have effects on adjacent vowels.Because *h3e = *[h3o], many people hypothesized *h3 is labialized, but this is contradicted by Anatolian evidence which shows *h3 has no inherent labialization. Proponents have to assume *h3 lost its labialization in Pre-Anatolian even though Anatolian has secondary labialized laryngeals.Having thought about this recently, I think the best way to explain the different vowel outcomes of *h2 and *h3 is probably duration or voicing onset time. Following Alwin Kloekhorst, I think the best reconstruction of the early PIE stop system includes fortis and lenis stops. Analogous to the stops, I think it makes sense that *h2 and *h3 may have been fortis and lenis fricatives respectively.So we could have *h2 = */χː/ and *h3 = */ʁ/ (correcting what I said earlier). */χː/ causes backing and lowering of the adjacent vowel. */ʁ/ only causes backing because the duration of the fricative is shorter than */χː/. I suspect in later PIE *h2 became ungeminated to */χ/, but how do we explain the continued *h2e -> *[h2a]? This is a weakness of the theory. Voicing onset time may be enough despite the lack of gemination. Maybe it's */χː/ > */χˑ/? (half-gemination). Alternatively, what we mean by a "fortis" */χ/ may require something other than gemination.
>>18719678Too much theory and too little objective truth.
>>18719779The Hittite reflex of word-medial *h2 is ⟨-ḫḫ-⟩ unless secondarily lenited. *h3 is usually lost word-medially unless secondarily labialized. Alwin Kloekhorst interprets consonants written doubly ⟨-CC-⟩ as fortis and those written as singletons ⟨-C-⟩ as lenis. The easiest interpretation is that *h2 was fortis and *h3 was lost frequently because of its lenis character.Anatolian evidence and a broad overview of the PIE obstruent system supports the theory that the different vowel coloring effects of *h2 and *h3 are due to a fortis : lenis opposition rather than labialization of *h3.The reason uvular fricative phonemes should be reconstructed rather than pharyngeal fricatives is because both Anatolian and Corded Ware branches support this pronunciation whereas pharyngeal phonemes are really only supported by the Balkan languages, Armenian, and Tocharian. Positional pharyngeal allophones are still open to debate though.
bump for more /ˌlJŋtɑːˈdeJʃən/beats having another ankle-deep christianity thread
>>18719837>Anatolian evidence and a broad overview of the PIE obstruent system supports the theory that the different vowel coloring effects of *h2 and *h3 are due to a fortis : lenis opposition rather than labialization of *h3.Now we're talking about evidence! I understand. Thank you for the explanation and for your time. Very interesting. It's a shame that Tocharian is relatively fragmented.
>>18719855
>>18720144
>>18720147
>>18720150
>>18720144>>18720147>>18720150>>18720151Thank you
>>18720177We probably can't assume *h2 was */χː/ at all stages of PIE. Thomas Steer has identified a process where *h2 became fortitioned to *g in the environment *-nh2V́-. This voicing is probably impossible for */χː/ but possible for */χ/.For example,*moneh2 (n.) "uprightness"=> *monh2ós -> *mongós (adj.) "characterized by uprightness"=> *mongi̯ós > Proto-Balto-Slavic *mangjás "man, husband"Also,*monh2ós -> *mongós (adj.) "characterized by uprightness"=> *mónh2us (n.) "man" (u-substantivization) > Vedic ⟨mánu⟩ "man"https://en.wiktionary.org/wiki/Reconstruction:Proto-Balto-Slavic/mangj%C3%A1s(Wiktionary gives *mon-gʷ-yó-s, but this is because the editor didn't read Derksen right. Derksen says *mon-g(ʷ)io- because the velar is ambiguous from the perspective of Balto-Slavic.)*-nh2V́- > *-ngV́- is also interesting in that it seems to exclude the possibility that this *g is an implosive.
>>18720226Steer assumes a different semantic development from *moneh2 and *monh2ós though. However, there is a methodological problem in that there is no good evidence for a nominal root simply meaning "man" by itself without additional suffixes. Instead, a verbal root *men- "to protrude, to be erect" exists which is behind Latin ⟨mōns⟩ "mountain". I'd rather see a derivational process where the word for man was created to describe men as upright beings.
>>18720237>>18720226That's cool. I'm wondering about the word for woman in PIE.
>>18720226In proto-indo-european, doc.
>>18718737I think that two things don't make sense here. It doesn't make sense to include s in the reconstruction, as that is obviously just the nominative ending. Second, if we reconstruct ǵ as dʲ rather than gʲ, a lot of things become more straightforward. Centum (and very much later Czech) depalatalized it to ɟ. This not only simplifies everything in the table, but it allows to add slavic languages as well, as a far more straightforward origin of "řídit" and its equivalents.
>>18719030Good explanation of parsimony anon
>>18721046>a far more straightforward origin of "řídit" and its equivalents.I don't think this word continues a palatovelar.As far as I can tell, the easiest way to account for the reflexes of the palatovelars in the daughter languages is by*ḱ = */kʲ/*ǵ = */gʲ/Unless we want to use the uvular stop hypothesis that *k = */q/ and *ḱ = */k/, but the evidence for *k = */q/ appears to be lacking.For example,*ḱ > Ancient Greek /k/*ḱ > Armenian /s/*ḱ > Albanian /θ/We can't have something further forward than */kʲ/ because it's otherwise difficult to explain how something like */c/ or */t͜s/ became /k/ in centum languages. This kind of backing would be typologically unexpected.*/kʲ/ > Ancient Greek /k/*/kʲ/ > /c/ > */t͜ɕ/ > */ɕ/ > Armenian /s/*/kʲ/ > /c/ > */t͜ɕ/ > */t͜s/ > */t͜θ/ > Albanian /θ/
>>18722293Proto-Kartvelian loanwords from PIE do not show the difference between velars and palatovelars.PK *uɣ- (yoke) < PIE *yugóm (id)PK *usx- (sacrificial bull) < PIE *uksḗn (bull)PK *otxo- (four) < PIE *(H)oḱto- (four fingers)Palatovelars might have never existed in Centum languages. The palatals could've been an innovation among some dialects.
>>18722327>Palatovelars might have never existed in Centum languages. The palatals could've been an innovation among some dialects.I believe this way of thinking falls apart once Indo-European is divided up into proper subclades. E.g., the Balkan languages (including Greek) are one subclade, and Armenian added to that is a parent subclade.Then the only way to explain the reflexes of the palatovelars in these languages is by the starting point *Kʲ, and we can't suppose Albanian and Armenian did this independently while Pre-Proto-Greek never had *Kʲ because we have to explain why Anatolian, Balto-Slavic, and Indo-Iranian coincidentally created this same series of palatovelars in the same environment. The simplest explanation is that palatovelars go back to PIE.I was looking at Proto-Kartvelian phonology just now and it does appear to have palatal stops, just some postalveolar affricates. It also doesn't differentiate between velar and uvular fricatives. I don't think the Kartvelian examples you gave are very diagnostic about the nature of PIE stops.
>>18722341>it does appear to have palatal stopsdoesn't
>>18722341Palatization is a trivial process, it can happen many times, independently, and with a similar result. Meanwhile the depalatization is a much rarer process, and under the traditional model it would have to happen several time independantly: Tocharian, Greek, Western European Centums (let's even say it was Proto-Italo-Celtico-Germanic dialect of PIE), and maybe Anatolian (but the last is uncertain).
>>18722349Depalatalization */kʲ/ > */k/ is trivial, however.Palatalization in general may be typologically common, but there is a significant obstacle to suggesting that the palatalization seen in Indo-European is the result of independent developments:1. The environment for palatalization must be first identified.2. It must be explained why various distantly related branches are coincidentally palatalizing the same velars in the same environment instead of a mixture of environments.If it was as simple as all these languages are palatalizing something like *ki, I might agree with you that the palatalization could be explained by independent developments.
>>18722362>Depalatalization */kʲ/ > */k/ is trivial, however.Are there attested examples? Is thw depalatization unconditioned?>The environment for palatalization must be first identified.PIE had a lot of *e and *i. Palatovelars are more common than plain velars, perhaps one should search for the environments where the palatization didn't occur.>It must be explained why various distantly relatedBalto-Slavic dialects of PIE could have formed a continuum with Indo-Iranian, and maybe with Albanian (or, rather, the Paleobalkan languages). They aren't that distant. The only oddball is Armenian, but it's not the only odd characteristics it has.
>>18719207>People think linguistics isn't a serious fieldyeah, I'm one of those people. renowned academicians disagree over basic facts of my native language (the random amateurs, of which we have many, are just plain unhinged, they are easy to spot).
>>18722293>We can't have something further forward than */kʲ/ because it's otherwise difficult to explain how something like */c/ or */t͜s/ became /k/ in centum languages. This kind of backing would be typologically unexpected.It has happened rather recently in Czech, where /tʲ/ /dʲ/ split into /c/ /ɟ/ and /t͜s/ /z/, depending on what followed.
>>18722373>Are there attested examples? Is thw depalatization unconditioned?Yes. Yes, inasmuch as the same vowel that caused the palatalization is still present when the velar loses palatalization. /kʲ/ is a velar stop with a secondary palatal gesture. /kʲ/ > /k/ is retention of the velar stop with loss of the palatal gesture.https://link.springer.com/article/10.1007/s11049-018-9430-3[kʲεfʲir] > [kefʲir]See: (66) Pronunciation reversals>PIE had a lot of *e and *i.This isn't sufficient to explain palatalization in PIE. *e doesn't palatalize plain velars. See for example *gʰei̯dʰ- (which would have a zero-grade *gʰidʰ-). On the other hand see *ḱei̯- "to be lying down". You could make the argument that *Ki is rare enough that maybe it was an environment at some point. Suppose in Pre-PIE stressed vowels were long (like Anatolian) *ḱei̯- might be the result of */kiː-/ > */kʲei̯-/. Essentially, */iː/ broke into a palatalizing diphthong. This idea would push these developments into Pre-PIE though, so you can't have palatalization as a development of individual languages.https://en.wiktionary.org/wiki/Reconstruction:Proto-Indo-European/g%CA%B0eyd%CA%B0->Balto-Slavic dialects of PIE could have formed a continuum with Indo-Iranian, and maybe with Albanian (or, rather, the Paleobalkan languages). They aren't that distant.Balto-Slavic and Indo-Iranian as a subclade is a valid hypothesis. Adding a Balkan language to that is very unlikely. Balkan languages and Armenian come from Yamnaya. Balto-Slavic and Indo-Iranian come from Corded Ware. Intuition says they should be distant. I know that isn't a linguistic argument, but if you wanted to make one, the primary way to do that, in my opinion, is to use a table of sound correspondences.
>>18718780PIE reconstruction is where historical linguistics starts looking suspiciously like eldritch phonology
>>18722401I can't tell you it is universally impossible but something like */c/ > */k/ is probably unlikely, or at least less likely than */kʲ/ > */c/. Typological arguments have limitations, but it is useful to know what is likely and unlikely.
>>18722655>inasmuch as the same vowel that caused the palatalization is still presentSo, the depalatization is unvonditioned? But Centums would have to have the unconditional change of palatovelars into plain velars.>Adding a Balkan language to that is very unlikely. There is a theory Dacian language was closely related to Baltic languages. And other paleobalkan languages (including Albanian) related to Dacian. Admittedly, the theory is weak, since we don't know much about the languages.
>>18722655>>18723241>So, the depalatization is unvonditioned?Sorry, I meant there's a condition for the depalatization in your example. But for Centums it happened unconditionally?
>>18723241For the Polish example, I will quote the article:>Since the derivation g gj is fully predictable from Surface Velar Palatalization, the underlying representation is underspecified for Palatalization, //lεgεnd+a//. In the 20th c. Surface Velar Palatalization changes status and becomes restricted to derived environments. Consequently, it can no longer affect morpheme-internal //gε// and hence legenda starts surfacing as [lεgεnda], with hard [gε]. Some words such as giermek [gjεrmεk] ‘page’ and kiedy [kjεdɨ] ‘when’ restructured their underlying representations from //gε// and /kε// to //gjε// and //kjε//, respectively, because there were no alternations between [kj gj] and [k g]. The restructuring process was lexically diffused and proceeded in an item-by-item manner, so, for example, the restructuring occurred in giermek //gjεrmεk// ‘page’ but not in legenda //lεgεnd+a// ‘legend’. When, in the 20th c., Surface Velar Palatalization lost jurisdiction over morpheme-internal structures, underlying representations emerged as surface representations, yielding [gjεrmεk] and [lεgεnda], respectively.>These pronunciation reversals highlight a new theoretical point. Namely, they show that reversals are possible in the absence of alternations.It's also probably worth discussing the case of Ancient Greek */kʲ/ > */k/. */kʲ/ is an economical starting point for Greek, surrounding Balkan languages, and Armenian, but why did Greek lose palatalization? It's possible that it has to do with the rise of secondary palatovelars, and the new rule for palatalization competed with inherited palatovelars which did not have this new environment for palatalization. Because they did not have the new environment, their palatalization was removed in accordance with a new rule.There may be different circumstances in the various centum languages, but I believe Polish and the case of South Jutlandic show that not much is required.
>>18723422Yes, some Slavic languages have depalatalized their consonants before the front vowels. But I think these are bad examples in the context of centumization.The Salvic languages historically have always palatazed their consonants before front vowels. The palatals were merely allophones, conditioned by the vowels. In case of centums, the depalatization had to be unconditioned.
>>18722341Proto-Kartvelian had uvular stops: plain /q/ and ejective /q'/. If these correspondences >>18722327 are true, then they are a strong argument against the uvular theory in PIE.
>>18722711There are other examples of sounds moving backward. Spanish /x/, /r/ in many European languages, /t/->/k/ and /k/->/ʔ/ in Hawaiian, as well as /t/->/ʔ/ that is increasingly common in English, or /q/-/ʔ/ in some Arabic dialects, so it isn't a one way road.
How did the Greek word end up as “basileus”, and the English as “king”, words that do not even remotely resemble each other or “rex”?
>>18724744It's debuccalization, and it's fairly common. Depalatalization is less common, and I cannot think of a single example where it would happen unconditionally (which is supposed to have happened to the centums). Various Slavic languages have depalatalized quite often, but only in certain phonetic environments, which varied from language to language. While all centum languages had to unconditionally depalatalize all their palatovelars. Possibly, this had to happen independently several times to different groups of centum languages (or dialects at that point).
>>18724289>If these correspondences >>18722327 are true, then they are a strong argument against the uvular theory in PIE.There are no Kartvelian stops in that post. See:>I was looking at Proto-Kartvelian phonology just now and it doesn't appear to have palatal stops, just some postalveolar affricates. It also doesn't differentiate between velar and uvular fricatives. I don't think the Kartvelian examples you gave are very diagnostic about the nature of PIE stops.>>18724744Here's the problem: moving a stop series in association with the PIE palatovelars backward is far from economical. *ḱ = */kʲ/ just works. Every branch can take */kʲ/ and either 1) stay at the same place of articulation or 2) move it forward into an affricate.There is zero motivation to posit an initial backward move such as */c/ > */kʲ/, and while it may be possible to argue that something like that can technically happen, developments such as /kʲ/ > /k/ and /c/ > /t͜ɕ/ have to be much more likely than /t͜ɕ/ or /c/ > /kʲ/. A backward move means we cannot explain the centum languages as independent developments, and there's no such thing as a centum branch. "centum" is just a feature of a bunch of unrelated branches. /kʲ/ > /k/ is trivial enough to happen independently many times. If that's still unbelievable for some reason, then it would point to *k = /q ~ k̠/ (/k̠/ should be underlined but it may not show on 4chan), *ḱ = /k/ and chain shifts that later created palatal series.
>>18723440>I think these are bad examples in the context of centumization.>The palatals were merely allophones, conditioned by the vowels.Polish is still a fairly good example, and I would also add South Jutlandic Danish as another example of depalatalization. However, you seem concerned about further details in that centum languages need depalatalization of phonemes quite specifically, but the level of detail needed to see the exact trigger for depalatalization in the centum languages is usually not available to us.But it is not difficult to imagine scenarios where depalatalization can occur quite naturally. I already mentioned the case of the secondary palatals in Ancient Greek. Secondary palatals also happened in Tocharian. Secondary palatals could easily interfere with inherited palatals, and we may not always be able to see if and when secondary palatals could have been created in the prehistories of the various daughter languages. It is also entirely reasonable for a phoneme */kʲ/ to develop an allophone *[k] (before back vowels as an example) and once the *[k] allophone exists, it's only a matter of time before before it takes over and the *[kʲ] allophone is dropped.
