Time to start a new standalone post, though as always most of what I publish will be in the comments section. I'll be continuing work on the Eskaleut origins of Tungusic words, and the sound laws that link them.
To briefly recap the latest mega-discoveries, I began with a new model for the birth of the Altaic languages as creoles that are mostly Samoyedic (in the Uralic family), with a crucial Yeniseian influence / phonotactic filter. This was partly about historical sound changes, but largely lexical, showing how Proto-Turkic in particular is "Sayan Samoyedic with a Yeniseian accent". But it created an Altaic foundation which the other two families used as well, Monogolic / Para-Mongolic, and Tungusic.
Then in the middle of that, I happened upon the mother of all secret passageways in Altaic studies -- Tungusic used to be a member of the Eskaleut family, which up until now has thought to be entirely confined to the New World, namely the Arctic New World, mostly the Eskimo branch, but also Aleut in the Aleutian Islands. The more I unearth the etymologies, the more it looks like Tungusic is mostly Eskaleut, and owes only a minor part of its nature to the Altaic foundation. It can still be considered a creole, neither pure this or pure that, but the influence and inheritance from its Eskaleut side is pervasive.
This represents a new contribution to the discoveries about linguistic connections between Siberia and the northern New World, such as Vajda's discovery of the sibling status between Yeniseian in Siberia and the Na-Dene family mainly in Northwest America (with a major recent spread into the American Southwest). I already added to that specific family in the past year or so, by showing that Japonic belongs to it as well. But this latest discovery of mine, about Tungusic being Eskaleut, provides a whole new link between Siberia and Northwest America.
Without rehearsing every single point made in the previous comment section, here are they key principles and sound laws:
The main problem that faced the speakers of Proto-Tungusic, was that the Altaic foundation came from Uralic, which is a half-sibling to their own native tongue, Eskaleut. If they let the two intermix willy-nilly, the outcome would be inbred and confusing. So, they made a concerted effort to distinguish these two sources of lexical items, when forming the lexicon of their new creole, Tungusic.
The Altaic base resulted from the Eastern Steppe nomad confederations of the 1st millennium BC (such as the Deer Stones culture), who were victorious in driving out the Indo-Iranian apocalyptic charioteer cult that had laid siege to the Sayan Mountain gate during the Andronovo era. As victors and protectors of the region, their new Samoyedic-Yeniseian creole enjoyed prestige among other residents of the Eastern Steppe, all the way east into Manchuria and the Amur region.
As such, Tungusic deferred to the Altaic foundation when separating its lexicon into two source. Words of Uralic origin would keep their Altaic form, while words of native local Eskaleut origin would be systematically altered, though still recognizable. Both Uralic and Eskaleut have a redundancy in their consonant series, where stops are voiceless and fricatives are voiced. Uralic-origin words preserved this redundancy. Eskaleut-origin words inverted this redundancy, but still kept a redundancy, to please their Eskaleut ears -- now the stops would be voiced, and fricatives would be voiceless. This creates a complimentary distribution between the two sources of words, allowing them to keep them separate.
The only major change to the phoneme inventory involved the uvular consonants from Eskaleut, since Uralic does not have consonants in this place. Briefly, "q" -> "g" and "ʀ" -> "r" in Tungusic. For the vowels, "ə" -> "e", although subject to moving due to vowel harmony.
At the earliest stage, vowel harmony seems to have been the imported Uralic pattern, since Eskaleut has no vowel harmony of its own, based on front vs. back. This may have changed over time, toward the typical Tungusic pattern based on high-front vs. low-back. This is a minor quibble to me at this point, so I'm just using front-back as a guideline.
Phonotactically, the major change was adopting a Uralic word-shape of CVCCV, perhaps involving deletion of a V between the medial C's in an Eskaleut root, or metathesizing the final two segments of an Eskaleut root of shape CVCVC. Also, new Uralic phonotactic rules for the nature of that medial consonant sequence, e.g. if there's a liquid, it must be the 1st of the sequence in Uralic, but not necessarily in Eskaleut, requiring metathesis or feature-alteration.
One other important sound change is deletion of voiced velar consonants between vowels, including Eskaleut "ɣ" and "ŋ", as well as voiced "w"-like "v" between vowels. Intervocalic velar deltion occurred in New World Eskaleut, and in the other Altaic languages, especially Turkic. Intervocalic deletion of "v" occurred in New World Eskaleut.
At the end of this initial process of adopting the Altaic foundation, and separating the two sources of words, these sound changes stopped. The job was done. However, words don't stop entering the lexicon just cuz the major initial stage has been completed. Late-coming words do not appear to show these sound changes as much, whether they're from Uralic or Eskaleut origin. This allows us to date when words entered the lexicon. It's similar to Ancient Greek, where earlier "s" -> "h", so if you see "s", that word is probably a late-comer to Greek, after that law ceased operating.
Specifically, the descendants of Proto-Samoyedic had already weakened "p" toward "x" / "h", possibly further deleting the weak consonant and yielding null. This was inherited into Turkic the strongest, where earlier "p" is null, while Mongolic and Tungusic held onto the "x" / "h" forms. Also, Tungusic changed "p" -> "b" in its Eskaleut layer, as part of the effort to distinguish Uralic from Eskaleut origin. Both of those laws erase "p".
And yet, there's still a decent amount of "p"-initial words in Proto-Tungusic -- these must be late-comers, regardless of their source. And indeed they are from less important word classes and semantic fields, unlike the crucial classes and semantic fields that show the sound changes of "p" -> "x" (Uralic) and "p" -> "b" (Eskaleut). They were lexical afterthoughts, late-minute filler, flourishes, secondary / learned / extended meanings, and so on.
There are also a handful of "t"-initial words of Eskaleut origin, entering the lexicon after the "t" -> "d" change in the Eskaleut layer. They join the words of Uralic origin, which never altered "t" in the first place.
One major common problem / shibboleth of Eskaleut and Tungusic, is palatalization of both alveolar "t" -> "c" and velar "k" -> "c", before "i" or even "ə" ("e" in Tungusic). Uralic / Samoyedic did not take part in this, let alone to such an obsessive extent. In Proto-Eskimo, no words begin with "ti", so they must have been palatalized to "ci" by the P-E stage. There are even gaps in the "tə" space, so some of those must have been sent to "cə". However, there are basically no words with "cə", so those also must have been raised to "ci", an overflowing container in P-E.
Thankfully, this palatalization of "t" and raising of "ə" was far less extensive in Old World Eskaleut, i.e. the non-Altaic parent in the Tungusic creole. So some cognates between Old World and New World Eskaleut clearly have "t" in Tungusic, vs. "c" in Eskimo, revealing that they used to begin with "t" rather than "c" in Eskimo, before the Proto stage.
Tungusic does not have "c", and it treated Eskaleut sources differently, seemingly based on chronology, although maybe there's some phonological conditioning that I can't see just yet. At the early stage, or certainly preceding non-palatal vowels, "c" -> "s", much like in several New World descendants of Eskimo. As mentioned, Tungusic did not palatalize "t" to "c" in some cases, so that source of "c" was avoided.
However, another source of "c" was earlier "ki" or "ke" (from "kə" in Eskaleut), which had been palatalized as well. Even among late-entering words, where initial "k" is preserved, there are no Tungusic words with "ke" or "ki" -- they must have been sent somewhere else, and there are a bunch of words beginning with "ʒ" (voiced palatal affricate), so they must have gone there. Eskimo did not palatalize "k", and there are many words beginning with "ki" and "kə". This palatalization of "k" is strictly Tungusic, and reflects the compresses phoneme inventory that happened when the uvular consonants of Eskaleut had to be squeezed into the Uralic-esque phoneme space, where the furthest-back place is velar.
Since "k" already has unaltered Uralic-origin words, and since that is an incredibly common consonant in Uralic, this container was going to be overflowing. So in some cases, "k" -> "x" in the Eskaleut layer, part of the inversion of the consonant feature redundancy, which produced voiceless fricatives and voiced stops. This seems to be the early stage.
Late-entering words, though, still tried to avoid overflowing "k", and palatalized them to "c" before "e" or "i" -- but since "c" doesn't exist in Tungusic, these were then voiced to "ʒ", in the same place. That keeps the change within the overarching inversion of feature redundancy, since an earlier "k" maps to a voiced stop / affricate.
Also worth noting that in late-coming words with initial "t", where "t" was preserved rather than voiced to "d", palatalization was still an active threat, and could have resulted in "t" -> "c", unlike "d", since voiced consonants do not palatalize as easily as voiceless ones. But even "d" palatalized would result in "ʒ", which would merge earlier "d" and "k", creating homophones and confusion. Since "c" is an empty container, earlier "t" words palatalized to "c" would therefore be mapped further into "ʒ". But that would have merged earlier "t" and "k" before palatal vowels, creating homophones and confusion. So, to prevent this homophone problem, the triggering palatal vowels "e" and "i" were rounded to "ö" and "ü", since round vowels tend to impede palatalization (too difficult to articulate both rounded lips and flexing the tongue tip toward the palate).
What I have not yet fully determined, is the origin of Tungusic words beginning with unusual nasals, "ń" and "ŋ". My brief look says they are either shifted from an earlier "n", or are nasalization of an oral consonant in the same place, like "k" -> "ŋ", since "k" was overflowing and had to be mapped elsewhere. They are not merely conditioned by the following vowel, as though "ń" only appeared before front vowels -- both of these nasals appear before all the vowels of Tungusic. If anything, "ń" is *less* likely to appear before a front vowel, so it could be a nasalized "c", for all I know.
To summarize the major sound changes in the Eskaleut layer of Tungusic. Remember, this is from Proto-Eskaleut, that is the shared ancestor between the Old World and New World branches. Since the New World branch palatalized "t" to "c" before "ə" or "i" already in its proto-stage, sometimes there's a correspondence between NW "c" and OW "t", due to NW palatalization of "t", and other times where NW "c" corresponds to OW "s", due to OW feature redundancy inversion (producing voiceless fricatives and voiced stops / affricates). Voiced fricatives from Eskaleut were mostly hardened to stops, as part of the inversion of feature redundancy.
Vowels:
"ə" -> "e" (subject to vowel harmony)
Consonants, voiceless stops:
"p" -> "b"
"t" -> "d"
"c" -> "s"
"k" -> "x" (early?, unconditional?)
"k" -> "ʒ" (late?, before "ə" or "i")
"q" -> "g"
Consonants, voiced fricatives:
"v" -> "b"
"ð" -> "d"
"ɣ" -> "g"
"ʀ" -> "r"
With all that said, today will see more etymologies, covering some of the "ʒ" words, including another numeral, the 4 seasons... and another massive hunting trophy, the Altaic sky-god's name Tengri. Stay tuned! ^_^
August 20, 2026
Eskaleut origins of Manchurian / Tungusic language family
Categories:
Archaeology,
Geography,
Language,
Mythology,
Religion
Subscribe to:
Post Comments (Atom)
P-T "abdu" = "cattle, herd; household, property; cloth, fabric"
ReplyDeleteP-E "avðiʀ" = "to hinder, protect", "avðaʀ" = "hindrance, protection"
The first 2 senses of "abdu" are "possessions", which in a nomadic society are livestock, and then the generic "property". Your possessions are things that you protect, so they stay yours and you get to benefit from them, rather than someone else rustling your livestock, stealing your jewelry, etc. You also protect your household as a whole. These sense have a passive role -- "things that are protected".
"Cloth, fabric" protects what it covers, at least from the elements, but also spilled food or liquids, weapons, and so on. This has an active role -- "things that protect".
Eskaleut verbal bases are ambitransitive, so they can refer to either active / passive or causative / anticausative, with those distinctions only spelled out by further overt postbases. Here, there are none, so it could mean both "things that are protected" and "things that protect".
"v" -> "b", "ð" -> "d", "ʀ" is the meaningless absolute case marker for the dictionary listing, yielding "abda".
There is no reason for the final "a" to raise to "u", under any theory of vowel harmony. Rather, I propose the "u" is from a fossilized case ending -- and what case could be more appropriate for "possessions" than the possessive case? In the absolute case, it's "vut" for the first-person possessor, plural possessed things. The original combination was something like "abdavut", meaning "our possessions", "our property".
"v" -> 0 between vowels, word-final C deleted since there's no way to work it into the Uralic / Tungusic word shape of CVCCV. That yields "abdau", still containing the 2nd "a" instead of just being "u", so there is a remaining irregularity to be accounted for -- but at least I got the "u" in there.
In any case, it's clear that the P-T word was inherited from Eskaleut, and with a fossilized case ending to boot. This is even stronger evidence that Tungusic is heavily Eskaleut in origin -- otherwise, it would borrow a word in its bare form, not with a fossilized bound morpheme like "our".
*1st-person plural possessor
ReplyDeleteYou could say the sequence "au" resolved into "u" by the "u" being the more important vowel, to signal possessive case, but usually that would result in compensatory lengthening of the deleted vowel. And yet it's a short "u", not long, so there's still an irregularity in that final vowel. I will leave this 99% solved for now.
ReplyDeleteFixed the final 1% on "abdu". In P-T, long vowels can only appear in the 1st syllable of a root. Suffixes can have long vowels, but not root syllables after the 1st. Since the "u" is a fossilized segment of a case ending, not a productive case ending assembled on-the-fly, "abdau" is treated as an atomic root, not root + suffix. Since "au" is not the 1st syllable, it can't be long.
ReplyDeleteSo the "u" was more important to preserve, as it was the fossilized possessive ending, and it simply replaced the "a", with no compensatory lengthening.
P-T "aŋaʒukān" = "orphan, widow", suffix "kān", root "aŋaʒu"
ReplyDeleteP-E "aŋayunʀuq" = "co-in-law", "aŋayukliʀ" = "oldest one", "aŋayuq(q)aʀ" = "parent or leader", "aŋayuɣ" = "elder sibling of same sex"
The root of the Eskimo words is "aŋayu" and it means "to be elder, senior", and the various endings specify what kind of "elder kin relation" it is.
However, the P-T word clearly means "to lack" -- or rather, "to have lost or been deprived of" -- elder or senior kin (husband being an elder / senior role, relative to wife).
There's a "deprived of" postbase in P-E, "i(ʀ)", which is the stronger version of the "lacking" postbase "it". It attaches to a noun X, and produces a verb meaning "to have lost / been deprived of X". It still requires a nominalizer ending, which in P-T is "kān". So we're just trying to get P-T "aŋaʒu" from P-E "aŋayu + [noun ending] + i(ʀ)".
In several branches of Eskimo, P-E "y" hardened to voiced sibilant "z" or affricate "ʒ". That's what happened in the Eskaleut precursor to Tungusic, "y" -> "ʒ".
Since the Eskaleut word is so multisyllabic, and Uralic / Tungusic wants word shape CVCCV, or at most CVCVCV, there was a lot of compression that was going to happen. So they must have taken the elder kin term that was the simplest, "aŋayuɣ", for the base. Then, "to have lost the elder kin member" = "aŋayuɣi". If it had final "ʀ", that would have been deleted in the compression, as there are already 3 C's.
"y" -> "ʒ", "ɣ" -> 0 between vowels, non-initial syllable cannot be long, so sequence "ui" resolved into "u" to respect vowel harmony with the other V's, which are back "a" (contra front "i"). That yields "aŋaʒu", which takes the nominalizer suffix "kān" to yield "aŋaʒukān", QED.
This is a very bizarre word in P-T, as it begins with a vowel, has only exotic segments "ŋ" and "ʒ" in its root, and is quite multisyllabic. The fact that its root so closely matches the shared multisyllabic base of those Eskimo "elder kin" terms is no coincidence -- Tungusic inherited it from Eskaleut.
However, it did not inherit the P-E word for "orphan" = "əłiɣaʀ(aʀ)". Rather, its Old World Eskaleut precursor had assembled a word of its own, using different Eskaleut morphemes, and passed this phrase onto its Tungusic creole descendant.
P-T "ʒȫr" = "two", with long vowel suggesting a deleted consonant
ReplyDeleteP-E "kəvɣaʀ" = "messenger, helper", "kəvəɣ" = "to lift", P-I "kəvɣumiaq" = "to carry in arms"
The P-E noun "helper" derives from the verbs meaning "lift" or "carry", i.e. to be the physical support which elevates the superior to his high status.
Playing such a crucial role as the supporter of the #1 guy in a social pyramid, entitles that person to be called "number 2", as in English.
Although Latin does not use figures of speech to coin the cardinal number for 2, it does for the ordinal "secundus", which is where English gets "second" from. It's derived from the verb for "to follow", referring to the role of the "second in command" as a "follower" rather than "supporter" ("upholder, etc."). Still, alluding to some key role played by #2 toward #1. If nobody followed the #1, there'd be total mutiny or no corps to begin with.
"ə" -> "e", "ʀ" -> "r", yielding "kevɣar".
Since Tungusic tries to separate Uralic vs. Eskaleut origin, Eskaleut "k" cannot remain. The main option was to soften it to "x", but the strategy with other Eskaleut voiceless stops was to voice them. However, "k" can't go to "g", since that's already taken by Eskaleut "q". Softening it to "x" threatens to overflow the very full container of "x"-initial words in P-T, some of which are of Uralic origin anyway, as weakenings of "p".
What else can it do? Here it is tempting to palatalize "k" before high-ish front vowel "e". It did this, sending "k" -> "c", but it still has to voice it so that Eskaleut roots are either voiced stops or voiceless fricatives. So it voiced "c" -> "ʒ", yielding "ʒevɣar".
"v" absorbed onto the "e" as a labial / rounding element. This is to minimize the number of C's in Uralic / Tungusic word shape. That yields "ʒöɣar".
"ɣ" -> 0 between vowels, sequence "öa" resolved by 2nd V absorbed into 1st, with compensatory lengthening (allowed, since it's 1st syllable), yielding "ʒȫr", QED.
This is not cognate with the P-E word for "two" = "malʀuɣ". But as I keep saying, numeral systems are just a kind of technology (for counting), they can be replaced entirely (mostly what happened in Japanese, which only kept native words for "4" and "7", and replaced the rest with Chinese loans).
And yet, just cuz the Old World and New World branches of Eskaleut chose different figures of speech, the NW word isn't so dissimilar -- it's based on the same semantics as Latin "secundus". In P-E, "mal" belongs to words relating to "follow" and "pursue", e.g. "maliʀqə" = "to chase".
You can debate whether "follow" is intended to mean "next in sequence", or "obey orders", but certainly in the Tungusic case, "support" can only be the social rank metaphor. They had belonged to the nomadic Steppe confederations, and were primed to use military rank metaphors when devising their numeral system. So, "2" had to be "2nd in command", or "the guy who holds up the #1 guy".
Military terms to coin the names for numbers -- who said jocks and nerds can't be friends? ^_^
P-T "ʒegde" = "to burn". This is intransitive, like "it burns". With causative suffix "gi", it becomes "ʒegdegi" = "to set on fire, make it burn".
ReplyDeleteP-E "kəɣə" = "to bite", "kəɣiʀ" = "to be in pain", "kəɣðaɣ" = "to be rough", P-Y "kəɣ(ɣ)əvə" = "to chafe"
The simplest root is "kəɣ" relating to biting, then extended to refer to any sort of pain, and given another particular form of pain, based on friction like from surfaces that are rough or that chafe the skin.
In English, we refer to "rug burn" and "rope burn", linking intense friction and abrasion with burning. We also refer to little sparks of fire, or maybe a little splash of hot oil, that land on the skin, as feeling like "something bit me". Hell, even extreme cold feels like "frost bite" or "freezer burn".
Intense pain, regardless of the source, is treated the same. But "biting" and "burning" are two of the most common forms of intense pain, so they are easy cognates, which they are in this case.
The P-T word is intransitive, as in English when we say "my hand is burning" after touching something hot, not "the hot oil burned my hand" (that is grammatically correct, but the focus is on the source of the burning there, whereas pain tends to focus on the sufferer).
It looks like P-T chose an earlier Eskaleut form phonetically like P-E "kəɣðaɣ". From there:
"ə" -> "e", triggering the "k" -> "c" -> "ʒ" shift we saw before. "ɣ" -> "g", "ð" -> "d", final C deleted since there's already 3 C's, and "a" harmonized with front "e" by fronting to "e" as well (the lower of the front vowels, since "a" is low).
That yields "ʒegde", QED.
P-T "ʒuga" = "summer". Now we tackle the 4 seasons.
ReplyDeleteP-E "kiðaɣ" = "heat or summer", "kiððiʀ" = "sweat"
The OW Eskaleut precursor to Tungusic named "summer" as "the sweaty season". As I'll emphasize again for winter, Siberia has cloudy / rainy summers (part of the Asiatic Low-pressure system) and clear-sky / sunny winters (part of the Siberian High-pressure system). So summer is not just hot, but humid and therefore sweat-inducing.
"k" -> "c" -> "ʒ" before "i", "ð" -> "d", "ɣ" -> "g", yielding "ʒidag".
To fit standard word shape of CVCCV, "ag" metathesized to "ga", and "i" pulled back to harmonize with "a", becoming "u" (preserving high height), yielding "ʒudga".
The sequence "dg" is banned, although "gd" is fine. But perhaps only 1 metathesis is allowed per derivation, otherwise you can really play musical chairs and end up somewhere unrecognizable from the source. And that was already used up for "ag" -> "ga".
There's a sonority principle for what kind of medial sequences are allowed, where the 1st member wants to be a vowel, liquid, or nasal (except m). If both are stops, labials have priority as the 1st, then velars, then alveolars / palatals. Since "dg" has a low-priority alveolar as the 1st member, it just gets dropped. No medial sequence begins with "d", regardless of the 2nd member.
That leaves "ʒuga", QED.
Before going onto the other seasons, a brief note on when "k" -> "c" -> "ʒ". These look like later layers in the lexicon -- numerals are not basic, often borrowed, and seasonal terms are also not the most basic, subjec to shifting around or coining new epithets. "Burn" is the only basic word.
ReplyDeleteI think Tungusic took this strategy for dealing with Eskaleut "k" when the other option had already been exhausted, namely softening to "x". There are tons of "x"-initial words, even after removing those of Uralic origin (from earlier "p"). But not so many "ʒ"-initial words.
It's not just the phonological environment, like being before "e" or "i", since there are "xe" and "xi" words.
After awhile, they got worried about every word beginning with "x", not just from Eskaleut origins but Uralic as well. So they opted for a different strategy, which would still adhere to the "inversion of feature redundancy" to distinguish the Uralic vs. Eskaleut layers of the lexicon. Instead of making "k" a voiceless fricative like "x", they made it a voiced stop / affricate like "ʒ". Crisis averted.
Another mini-note on "ʒuga", why did the vowel harmony move the 1st V rather than 2nd? Why didn't "a" get pulled forward to harmonize with "i"?
ReplyDeleteWell, there's another phonotactic rule against "ʒi", whether initially or medially. There's "ʒü", which shows the round vowel blocking a rule related to palatalization, as mentioned earlier. So the rule is about dissimilating palatal consonant "ʒ" from palatal vowel "i", similar to the ban on "yi" in Japonic.
To avoid "ʒi", the word for "summer" had to shift the 1st V, so the 2nd V became the target rather than the mover.
P-T "ńeŋńe" and "nelki" = "spring season", with a metathesis between the velar and palatal C's of the medial sequence.
ReplyDeleteP-E "nəŋłinaʀ" = "to be cool", "nəŋłaʀ" = "to cool down", "nəŋłi" = "to be or get cold"
These share a more basic root "nəŋ" having to do with "tight", with "nəŋə" = "to stretch or lower a rope", and "nəŋ(ŋ)aʀ" = "to be angry" (lots of tension, "wound up tight" in English, taut muscles, etc., linking "anger" with "cold" or "frozen" instead of the usual "hot").
So, the cold season is also one where things become hard and inflexible, due to freezing over, making them tight and rigid instead of pliable.
Originally the word applied to late winter / early spring, when the effect of winter freezing / hardening had been completed. Even in early spring, there hasn't been enough time to unfreeze all 3 months' worth of freezing, despite warmer temperatures. This Eskaleut word shifted a half-step forward to refer to "spring" in Tungusic.
Phonotactically, medial sequences require a liquid before a stop, and if they're the same manner, velars before alveolars and palatals. Triangulating with the P-E words, P-T "nelki" must have originally been "nekli", with the same "velar, palatal" sequence as "ńeŋńe". And the initial C in "ńeŋńe" was originally "n" but palatalized by following "e".
The original Pre-PT word, which was resolved into both "ńeŋńe" and "nelki" (via "nekli"), must have been "neŋli", a dead ringer for the Eskimo root for "getting cold". However, in P-T, the sequence "nasal, liquid" is banned for sonority reasons, and only "liquid, nasal" is allowed, as well as "nasal, nasal" in different places of articulation.
Since "lŋ" is legal, yet the form with the liquid has "lk", first the "ŋ" -> "k", then metathesized to fit "liquid, stop", yielding "nelki".
The other form left "ŋ" and altered the manner of "l" to become a nasal in the same place as "l", to fit the "nasal, nasal", yielding "ŋń". It also palatalized initial "n" before "e", and lowered final "i" to "e", perhaps on analogy with the 1st syllable.
That accounts for the Pre-PT form underlying both of the cognates in P-T. What about from P-E to Pre-PT? Only the usual, "ə" -> "e", "ł" -> "l" (no lateral fricatives in P-T).
BTW, now we get P-T "ńaŋńa" = "clear sky" for free. Recall that winters in Siberia are clear and sunny, while summers are cloudy and rainy. Since the Pre-PT "neŋli", which gave rise to P-T "ńeŋńe", originally referred to late winter / early spring, it originally referred to a period where the skies were clear rather than cloudy.
ReplyDeleteThe only phonetic change was a vowel alternation, using the other low unround vowel, "a" instead of "e", in both V slots.
P-T "bola" = "autumn", not derived from the Uralic verb for "to become", which became "bol(V)" in the Altaic foundation. To become what? Every season is a time of becoming something. We need something specific. So...
ReplyDeleteP-E "pəłu" = "leaf", "pəlu(ʀ)" = "dust, dirt", P-YS "pəłuɣ" = "to scrape skin"
The simpler root "pəl" refers to small things covering an object to be loosened and then shed and fall to the ground. Like beating the dust or dirt from a carpet, or shaking dust / dirt loose from your clothing after getting dirty, or scraping flakes or scales of skin from a hide -- or leaves being loosened and shed from trees, and falling to the ground.
Hey, very similar figure of speech as in English -- "fall", referring to leaves. In Tungusic, "fall" is more properly "the shedding season", but clearly referring to trees shedding their leaves.
"p" -> "b", "ł" -> "l", "ə" -> "e" then pulled back to harmonize with "u" at the same mid height, yielding "bolu".
Then the Eskaleut passive / resultative suffix "a" for "that which has been X-ed". Instead of the noun "leaf" or verb "to shed leaves", it's the season "which has been leaf-ed", as though leaves were the agent and the ground and environment were the patient. Especially in the forest climate where Proto-Tungusic spawned, fall results in a lot of shed leaves and pine needles, blanketing the ground and transforming it.
The ground, nay, the entire land has been leaf-ed! ^_^
Finally, P-T "töge" = "winter"
ReplyDeleteP-E "ciqinəʀ" = "sun, sunshine", "ciq(q)it" = "to pour out or dump with a splash"
In Eskaleut, the word for "sun" comes from the verb for "to pour", which has a simple root "ciq" in P-E, shared with non-liquid substances that are still very fine-grained solids, like "ciqumə" = "to break into pieces" and P-I "ciquptəq" = "to pulverize". As long as it's fine enough, solid particles can be "poured". In English, we can say that the sand in an hourglass "pours down" from the top part into the bottom part.
Also in English, we say the sunlight is "pouring" into a room, or that we have installed a massive-watt "flood light". It's easy to think of sunshine or light from any source, as a physical mass noun that flows downward by gravity from a high-up source, like the sky.
And so, the sunny season is also the "pouring" season -- not pouring rain, but pouring sunlight, a semantic link already made in the Eskaleut word for "sun" as "the thing that does the pouring". It doesn't rain or snow heavily during winter in the Siberian High, but sunlight does pour and pour and pour some more, since their winters are very high-pressure, hence no clouds form, and the sky is clear and open to the sun to pour its light down.
Winter in Siberia may be freezing cold, but the skies are cobalt blue, and there's nothing but sunlight pouring down during the daytime.
It doesn't look like Tungusic derived "winter" from "sun", which has the "n" and "r", but the simpler verb "to pour", i.e. it's "the pouring season", where "pour" is understood to refer to sunlight. Most of the cognates don't have "t" after the 2nd V, so the base seems to have been "ciqi", not "ciqit".
Recall that in P-E, there was a massive merger, where both Pre-PE "ti" -> P-E "ci" and "cə" -> "ci". There are no P-E words beginning with "ti" or "cə". However, even though "tə" is allowed, for some reason "təq" is not, so it would have been mapped to "ciq", which is right where we find "ciqi" relating to "pour" and "sun".
Well, as we saw before, Old World Eskaleut did not take part in this mega-merger relating to "t" -> "c", nor raising "tə" to "ti" / "ci". Here we see yet another example, since P-T "töge" refers to the "pouring down" sunlight season, yet it begins with "t" rather than "c" as in the P-E cognates. So in fact, the P-EA root must have been "təqə" or "təqi". The P-E words for "splash" have either "ə" and "i" as the 2nd V, "ciqəʀ" and "ciq(q)i", so that 2nd V is hard to tell. Possibly there was some ambiguity in the P-EA origin.
Taking the P-EA form to be "təqə":
"q" -> "g", "ə" -> "e", and 1st "e" rounded to avoid palatalization into "c", which is an empty container in P-T (aside from a couple loans / substrate words). There are some words beginning with "te" and "ti", but it's much more common to find their rounded counterparts, since the former are unstable in Eskaleut, in both Old World and New World branches.
That yields "töge", QED.
If the P-EA form was "təqi", the only difference is that intermediate form "tögi" shows sequence "gi", which is banned in roots (only as the causative suffix). Probably that's to avoid palatalization toward "ʒi", since that sequence is also banned. And similar to why "ki" doesn't exist, outside of suffixes, since that would threaten to wind up as "ci" or "ʒi", both banned. So, lower "i" to "e", still high-ish front and unrounded, yielding "töge" again, QED.
The P-T word for "winter" was adopted late, not early, into the lexicon. It doesn't show the typical "t" -> "d" change for words of Eskaleut origin.
ReplyDeleteHowever, there's only 1 word beginning with "de", so the "t" -> "d" strategy was not necessarily the best or only way to deal with underlying "te", under threat of palatalization. One strategy is to voice "t" since voiced consonants don't palatalize as easily. The other strategy is to round "e" since rounding impedes palatalization.
There are no words beginning with "dö", which shows that this would have been unnecessarily redundant -- it could only have come from "te" (and earlier "tə" in P-EA), so why do both strategies? Either one will prevent "te" and its threat of winding up with "c".
At any rate, "töge" was a late-entering word, since the "t" -> "d" sound change was no longer obligatory, although "q" -> "g" was still in effect, as was "ə" -> "e". Seasons are not the most primitive words, since they're a bit more abstract -- lasting months instead of the most concrete weather cycle, which is a day, or maybe a lunar cycle for a month. Thus, they can be coined later as afterthoughts, as it was in this case.
And that brings us to Tengri, the sky-god and chief object of worship in Altaic religions, especially Turkic but also Mongolic, althought not Tungusic, oddly enough. But languages can borrow words for deities, as several Uralic branches did with the Indo-Iranian word "bhagas" meaning "deity" -- Russian borrowed it too, as "bug", probably indirectly since the Indo-Iranians were long gone by the time Slavic was born in the 1st millennium AD.
ReplyDeletehttps://en.wikipedia.org/wiki/Tengri
Although Tungusic speakers did not get hardcore into Tengrism, and kept to their existing shamanic religion, they at least provided to lexical origin of Tengri's name, which is Eskaleut. Since Mongolic shares some of the Eskaleut-isms that Tungusic has, possibly this originated from the precursor to Mongolic, which had some Eskaleut character as well, just not as strongly as Tungusic did. And from Mongolic, it passed into Turkic. Or, Mongolic borrowed the word from Tungusic, and then loaned it into Turkic. Either way, it's Eskaleut.
In Proto-Turkic, his name is "Teŋri", and in Proto-Mongolic "Tenger" ("sky"). He is not just associated with the celestial body, the sun, but sunlight and the clear blue sky, which admits sunlight to pour down onto the land.
The etymology section at Wikipedia reviews the main theories, but they're all grasping at straws, which I would be doing too, if I hadn't discovered the Eskaleut nature of Tungusic.
Lo and behold, Mongolic "tenger" comes from the Eskaleut word for "sun", in the sense of "sunlight pouring down", linked to clear blue skies.
Rewinding the P-E form back to P-EA, by taking into account the Old World forms that did not palatalize "t" to "c", we get something like
P-EA "təqinəʀ" = "sun, sunlight"
"ə" -> "e", "q" -> "g", "ʀ" -> "r", but unlike Tungusic, leaving "t" alone, since Turkic especially but also Mongolic did not like initial "d", preferring "t". That yields "teginer".
Since Uralic prefers 2-syllable words, not 3, the 2nd V was deleted. And since Turkic and Mongolic share the Tungusic ban on "stop, nasal", the medial sequence metathesized to "nasal, stop", yielding "tenger", QED for the Mongolic case.
Turkic is the most Uralic of the 3, and insisted on CVCCV word shape, while still preserving the word boundary consonants. So it fused the "ng" into a single "ŋ", and metathesized the "er" into "re", yielding "teŋre". Although word-final "e" and "o" are allowed, these mid vowels are rare word-finally -- either high or low vowels are strongly preferred. And since "e" is front, where there's not much of a very low vowel to move to, it raised to "i", making the end vowel more euphonic to Turkic ears, yielding "Teŋri", QED.
And there you have it -- the chief god of the religion that powered the westward Turkic and Mongolic expansions into Eurasia, had a name that we would only expect to find among the Eskimos! But in one of the biggest unlockings of ancient mysteries ever, Tungusic has been revealed to be largely Eskaleut, especially at the early stage, when this god's name was coined. It was recorded by the Chinese in the 4th C BC, in use among the Xiongnu confederation, when the Altaic creolization was still in its infancy.
Yet already they were elevating Old World Eskaleut words to god-like status... except among the speakers who provided the original non-theonymic form of the word! But languages are happy to borrow words for gods, spirits, and such, including Tengri himself. ^_^
*Russian "bog"
ReplyDeleteTungusic word for "chipmunk" cognate with Eskimo word for "cheek". Awwww...
ReplyDeleteP-T "ulgukī" = "chipmunk", animal name suffix "kī", root "ulgu"
P-E "ułuɣaɣ" = "cheek", "uluɣ" = "to soften an animal skin by chewing or rubbing"
Chipmunks' most salient feature is their large cheek pouches that they store things in while transporting them.
The Eskimo word for "cheek" is derived from the verb about softening an animal skin to make it more flexible and pliable, by adding a nominalizer "aɣ" to the verb, meaning the cheek is "the thing that softens an animal skin by chewing".
Sometimes skin softening was done by rubbing with the hands, or crumpling it to introduce folds in the structure. But sometimes it was done by putting it between the jaws and chewing it -- not biting or tearing all the way through, not swallowing it, but to using the jaws in a repeating clamping-down motion, to cause folds and bends to form in the structure, making it easily workable instead of a solid stiff plane.
Here's a picture of the practice from 1951, "Inuit woman “Josie” chewing sealskin to soften it for making kamiits (boots), Kinngait, Nunavut":
https://commons.wikimedia.org/wiki/File:Inuit_woman_%E2%80%9CJosie%E2%80%9D_chewing_sealskin_to_soften_it_for_making_kamiits_(boots),_Kinngait,_Nunavut_(31497043966).jpg
The simplest phonetic derivation is from the verb form "uluɣ", but semantically that's not good enough, since chipmunks aren't notable for chewing things but for having big cheeks. So, starting from the noun form "ułuɣaɣ":
"ł" -> "l", metathesis of "uɣ" -> "ɣu" to produce standard Uralic / Tungusic medial CC cluster, final "aɣ" deleted since standard Uralic / Tungusic word shape ends after the V that follows medial C cluster. Or, "a" absorbed into preceding "u" (producing the best kind of harmony, identical V's), with no compensatory lengthening, since this isn't the 1st syllable, then word-final C deleted since it's not "n". Either way, that yields "ulgu", QED.
Note that Eskaleut not only provides the semantic basis for Tungusic "chipmunk", namely "cheek", it also provides the semantic basis for Eskaleut "cheek", namely the thing that chews a skin to soften it. That is a deeply rooted semantic chain-of-custody, establishing without a doubt that Tungusic is from Eskaleut origins.
This rules out Tungusic merely borrowing a word for "chipmunk" from Eskaleut, as may happen with animal names. In fact, there is no word for "chipmunk" in Eskimo languages -- they don't live way up in the Arctic, but they do in the forests of Manchuria and nearby, where Old World Eskaleut was spoken, before being creolized into Tungusic.
And of course "ɣ" -> "g"
ReplyDeleteP-T "ulgian" = "pig"
ReplyDeleteP-E "uliɣ" = "to cover, blanket"
First, I agree with Fortescue that this Tungusic word was loaned into Nivkh, rather than vice versa. Nivkh words rarely begin with a vowel, whereas "ulgian" does, and so does "ulgukī".
https://en.wiktionary.org/wiki/Reconstruction:Proto-Tungusic/ulgian
But I go further and claim it's of Eskaleut origin. It has the same beginning "ulg" as "ulgukī". And there's another animal name in P-Yupik, "ulaɣaq" = "fox", which would begin with "ulg" if it had been adapted into Tungusic (though it was not). These are native Tungusic names, ultimately from their Eskaleut origins, not the Uralic / Altaic side (and not loans from Nivkh).
To the Eskimo verb base, add "a" which produces a noun, "uliɣa", that means "that which has been X'ed". In this case, "something that has been covered or blanketed". The "n" in the P-T word is the standard / generic word-final "n" in P-T, not an Eskaleut word ending. So we just have to derive P-T "ulgia".
"iɣ" metathesized -> "ɣi" to fit standard Uralic / Tungusic word shape with medial CC cluster, "ɣ" -> "g". That yields "ulgia", QED.
Semantically, a pig is an animal noted for being habitually covered or blanketed -- probably alluding to its habit of wallowing in mud to cool itself off (no sweat glands), which when the mud dries into its fur, leaves the body covered in a muddy / furry blanket.
Less likely, but possibly, it alludes to the distinctly thick layer of back fat, acting as a natural blanket. Or perhaps the lacy blanket-like caul fat that pigs are prized for.
Somehow, though, Eskaleut provides the semantic basis for "pig" in Tungusic, relating to "blanket". Somehow, too, modern American English intuited this semantic connection, and coined the name for a food item, "pigs in a blanket". ^_^ Those are hot dog pieces wrapped in pastry dough.
So yet again, we see that the animal name in Tungusic is not just a borrowing with no semantic basis. Eskaleut provides a semantic basis for the name, not only its phonetic form as is true if it's merely a loanword.
I spoke too soon! That Yupik word for "fox" *does* show up in Tungusic! It's a late-entering word, which replaced the P-T word for "reindeer" = "oran" in the Oroch language, and which means "domestic animal" in Manchu. Both of those are the expected reflexes of P-T "ulaga", although it's not attested that far back:
ReplyDeletehttps://en.wiktionary.org/wiki/Reconstruction:Proto-Tungusic/oran
Because it's late-entering, it didn't reflect the pressure from Uralic / Altaic as much as the early-entering words. For example, the word for "winter" keeping "t" instead of voicing it to "d". In this case, it didn't obsess over making words fit the Uralic word shape of CVCCV, and it did not metathesize "ag" -> "ga" to yield "ulga". Rather, it left them in place.
The Yupik word is "ulaɣaq", so all that changed was "ɣ" -> "g" and deleting word-final C since it's not "n" and there are already 3 syllables, so metathesizing it would not create a CC cluster after the 1st V, a la Uralic.
That yields "ulaga", QED.
What's the meaning of this word? Well, "aɣ" is a frequentative / habitual postbase, and attaches to a verb base to mean, "does X habitually", in this case to describe some enduring trait of an animal.
That verb base is "ulə", where the reduced "ə" drops out when the postbase attaches to it. There are two homophones for this base, one meaning "to overflow", and the other having to do with "to mistake or confuse".
The Yupik word for "fox" certainly derives from the one about "mistake, confuse", since Eskaleut verbs are ambitransitive and ambicausative, meaning this could also refer to causing someone to make a mistake or get confused -- like the sly, cunning, deceitful, trickster fox.
I think this base could also refer to "reindeer" or other domestic animal, since they tend to wander, get confused about where to go, make mistakes, etc., as herd animals -- and it's the job of the herder to guide them where they're supposed to go, so they don't lose their way by straying or wandering.
Alternatively, it derives from the sense of "overflow", relating to the tendency of livestock species to increase in size exponentially by reproduction, until there's just too many of them for one single herder or family to properly manage. The frequentative / habitual postbase means they do this habitually, and it's a recurring problem of herd sizes getting too big.
But since the "mistake, confuse" sense already has a precedent in New World Eskaleut, I favor that sense for the Tungusic word as well. Livestock are the "easily straying" animal that require constant vigilance and guidance and direction and driving toward the intended destination.
P-T "erün" = "time"
ReplyDeleteP-E "ərə" = "to run or fade (of colors)" and in P-YS "dawn", P-Y "əʀnəq" = "day"
The Eskimo root has to do with fading colors, and colors are initially talked about in terms of brightness and darkness (objective light levels, not a subjective qualia like "red" -- pre-introspection vs. post-introspection). So fading brightness could have referred to the sunrise and sunset cycle, leading to "dawn" and the length of that cycle, "day".
There's a P-E postbase "un" which usually denotes a tool or instrument or medium, but it can also be abstract, including a time or occasion when something happens. Either "time" is conceived of as a physical medium or instrument that enables the daily fading cycle, or it's the abstract "time / occasion" when the fading cycle takes place. IDK exactly how ancient Tungusic cultures conceived of time, but it was either of those.
Base-initial "ə" -> "e", base-final "ə" drops out when postbase "un" attaches, "u" fronts to "ü" to harmonize with front "e". That yields "erün", QED.
P-T "eri" = "to breathe"
ReplyDeleteP-E "əʀina(ʀ)" = "voice", P-I "əʀiala" = "to shout"
Since "na(R)" is postbase, the verbal base is "əʀi", which means "to vocalize". Vocalizing requires exhaling, one phase of the breathing cycle.
"ə" -> "e", "ʀ" -> "r", yielding "eri", QED.
Easy-peasy, lemon squeezy! ^_^
I doubt "o"-initial Tungusic words are Eskaleut, since Eskaleut lacks "o", somewhat like the Crescentine family, where Eskaleut "ə" is like Crescentine "e", then "a, i, u". But no "o". It could be a vowel-harmonic alteration of "e", from "ə", but so far it seems like the 1st V is the target, and the 2nd V harmonizes with it. Also, "e" -> "o" is 2 changes, in frontness and in rounding.
ReplyDeleteThere aren't many "o"-initial words in P-T, and one of them, "omŋa" = "to forget", has a cognate in P-Turkic "umnï" (ditto), all but confirming that it's from Samoyedic / Uralic origin, not Eskaleut. But you could have told that just from its initial vowel being "o" -- very unlikely to be Eskaleut, since that vowel doesn't exist in their inventory.
Also, the Tungusic word for "time" shows that it was loaned into Nivkh, not vice versa. Only Tungusic can provide a semantic basis, from its Eskaleut side.
ReplyDeleteP-T "üse" = "seed, seedling; to plant, grow"
ReplyDeleteP-E "uci" = "load", "ucuɣ" = "sexual organ"
The sex organ carries a "seed" to make babies, and the seed is treated like a "load", much as in English for the seed in the male organ.
The P-T word reflects earlier "ucə" in OW EA, where "ə" -> "e" and "u" fronted to "ü" to harmonize with front "e", and "c" -> "s".
The final "ə" cannot be seen in the NW EA cognates, since "cə" never surfaces, only as "ci". Or if "uɣ" attaches to it, the "ə" drops out anyway.
Considering that greater extent of palatalization in the NW branch, these are cognates and from P-EA "ucə".
P-T "üńikün" = "alive, living", suffix "kün", root "üńi"
ReplyDeleteP-E "uŋilaɣ" = "to itch", "uŋuləg" = "to shake, tremble", "uŋummaʀ" = "to come alive again", "uŋuma" = "alive", "uŋuman" = "heart", P-Y "uŋuŋsiq" = "animal"
The Eskaleut root relates to "nervous system activity" to indicate a thing that is alive vs. dead, or animate vs. inanimate. When we use "nervous" as an adjective in English, it refers to an itchy or shaking / trembling state, as in the nervous system is overly active. Eskaleut has a similar connection. The link from the Eskaleut words to the P-T word is transparent, including the exact same sense of "alive".
Since the semantic basis is "nervous system activity", the most appropriate P-E words are "to itch" and "to shake, tremble". Their bases are nearly the same, "uŋi" vs. "uŋu". Clearly the P-T word is closer to the former, which we'll start from.
"ŋ" -> "ń" before "i", "u" fronts to "ü" to harmonize with front "i", yielding "üńi", QED.
In P-E, "ŋi" is far rarer than "ŋa", "ŋə", or "ŋu". It's another sign of palatalization, here targeting the velar nasal. Now we see the same concern in OW EA, where "ŋ" palatalized to "ń" before "i". Eskaleut doesn't have "ń", so I expect Eskimo mapped "ŋi" to "ni" instead, although I haven't checked that.
That's it for my exploration of vowel-initial words in P-T -- lots of Eskaleut origins there. And there are probably more, just seeing what I can easily figure out. I've done a lot with consonant-initial, so this filled out some other areas of the lexicon. Next, I'll get back to "x"-initial...
P-T "xabdansa" = "leaf", suffixes "n-sa", root "xabda"
ReplyDeleteP-YS "kavtak" = "to make a rattling or crunching noise, by walking over things on the ground"
The Eskimo verb refers to the sound made by walking over crisp crunchy stuff on the ground, usually ice or snow, or seashells, or frozen grass, etc. This is particular to their Arctic environment.
In a forested region like Manchuria, they have an additional source of natural stuff layered on the ground that makes a crisp crunchy noise when you walk over it -- leaves! In Tungusic, leaves are "the crispy-crunchy thing". Not the typical senses that are appealed to (texture and sound), when coining a word for "leaf" -- but very romantic and autumnal. How wonderfully Manchurian... ^_^
"k" -> "x", "v" -> "b", "t" -> "d", word-final C deleted since there's already standard Uralic / Tungusic word shape of CVCCV. That's "xabda", QED.
P-T "ʒuke" = "ice"
ReplyDeleteP-E "ciku" = "ice", "cikuŋiʀ" = "to break up (of ice)", P-I "cikaq" = "stiff", "kikkuləq" = "breathing hole in ice used by seals"
As with "ci" in New World Eskaleut, it could come from all sorts of origins, given the extensive palatalization of both "k" and "t" before "i" and sometimes "ə".
The clue that the NW EA word originally began with "k" is the Inuit word about the hole in the ice, which begins with "ki". The clue that this traces back to earlier "kə" is that this Inuit word did *not* undergo palatalization to "ci". NW EA distinguishes between "i" that was always "i", and "i" that came from "ə" -- the former are far more likely to trigger palatalization, and the latter far less likely.
So the P-EA word for "ice" was originally "kəku", which has a fossilized nominalizer suffix "u" attached to the verbal base "kəka" = "(to be) stiff". Ice is the "stiff" phase of H2O, not the liquid phase like water. Starting from "kəku":
"ə" -> "e", "k" -> "c" -> "ʒ" before "e", yielding "ʒeku", then vowels metathesized to dissimilate palatal "ʒ" from front vowel "e" ("ʒ" strongly prefers to be followed by back vowels). That's "ʒuke", QED.
This is probably a late-entering word, since it shows "k" -> "ʒ", suggesting there were already too many words derived by "k" -> "x". Both are part of the feature redundancy inversion -- voiced stops and voiceless fricatives. But the large amount of "x"-initial words, and smaller amount of "ʒ"-initial words, says the former was productive earlier, during the main stage of populating the lexicon, and the latter was resorted to only when the earlier container was full.
Also, both NW and OW Eskaleut are under threat of palatalization, and OW was more scrupulous about avoiding it. So when "k" -> "ʒ", that's contrary to their main strategy of keeping "k" in the back location (as "x"), and must have been a last-ditch effort, happening late.
P-T "ʒola" = "stone"
ReplyDeleteP-E "civa" = "to boil blubber", "kəvəɣ" = "to lift", P-I "civli" = "sediment in oil lamp", "kəvlək" = "to be effective", P-YS "civ(v)aʀ" = "downhill"
The Eskimo words all relate to "weight" or "heavy" -- heavy objects rolling downhill, heavier materials sinking when rendered (as in boiling blubber, or sediment that settles down in an oil lamp), heavy objects requiring lifting or carrying.
Being effective could relate to the sense of "heavy" as in "strong" or "resisting external pressure" or "weighing down on something to make its own pressure felt onto the patient". Alternatively, "effective" is about the literal material "stone" being the material for tools that make them the most effective -- unlike flexible or soft materials, which could be ineffective, stone makes them the most effective they can be.
The presence of "kə" suggests these all trace back to "kəv" in P-EA, with some degree of palatalization to "ci" in NW EA. The one most directly tied to "heaviness" is "sediment", about heavy things sinking, so the P-T word for "stone" must derive from an earlier Eskaleut word where "v" was followed by "l". The fact that "stone" can directly link to "effective" relating to tools, means we start from "kəvlək" (which was probably earlier "kəvləq" in P-E, but that final C is irrelevant anyway).
"ə" -> "e", "k" -> "c" -> "ʒ" before "e", word-final C deleted since there's already a word shape of CVCCV, yielding "ʒevle". To dissimilate "ʒ" from "e", vowel harmony is not possible since the other vowel is identical. So rather than harden "v" -> "b", it used the labial feature of "v" to round "e" into "o" -- albeit not front round "ö" as happened in "ʒȫr". Then 2nd "e" pulled back to harmonize with "o", while remaining low and unround, namely as "a". That yields "ʒola", QED.
This is another late-entering word, since "k" -> "ʒ" rather than "x". It also shows a slightly different outcome of dissimilating "ʒ" from "e", compared to fellow late-comer "ʒȫr". At such a late stage, rules were more of a last-ditch effort than a highly regular conversion of one sound to another.
Because "ʒ"-initial words are late, they're small in number, so it's hard to hear their subtle differences as "irregularities" -- which are more salient if they're opposing a regular pattern based on 100 examples. When there are so few comparable words, no single one of them sounds regular, and therefore, no single one sounds irregular. They're the oddball afterthought words, where uniform standards are no longer rigidly enforced, they're just filling in some remaining gaps in the lexicon, and quality control is less important than filling the gaps ASAP so the entire project of populating the lexicon can finally be done.
P-T "xilŋü" = "tongue", probably from earlier "xileŋü"
ReplyDeleteP-E "kələɣ" = "to inform", "cilaŋit" = "out of one's mind", "cilaŋiʀutə" = "to lose one's senses"
The Eskimo word for "inform" is the only one that retained the original "kəl", while the others were schwa-raised and palatalized to "cil", as is common in NW EA. They're semantically related since "to inform" uses "speech", and the others about losing one's mind or senses have the "deprived of X" postbase beginning with "ŋi", so the root of "intelligible speech" is "cila" in those two words, from earlier "kəla".
The Tungusic word is the body part used as a tool in order to "inform", that is by using speech, i.e. the tongue. So it attached the Eskaleut postbase "un", for tools or body parts, onto the verbal base. In OW EA, the original complex word was:
"kələɣun"
"ə" -> "e", yielding "keleɣun". I haven't worked out the conditioning or layer in the lexicon, where "e" is raised to "i", but it was in this case, as it was in NW EA, yielding "kileɣun".
"k" -> "x", yielding "xileɣun" (early word, not late where "k" -> "ʒ").
To prepare for typical CVCCV word shape, "un" metathesized to "nu", to bring consonant closer to middle of word, and make word end in a vowel, yielding "xileɣnu".
There are 4 C's, 1 too many, so sequence "ɣn" fused into "ŋ", preserving velar place, voicing, and continuant manner of "ɣ", just making the air-flow nasal instead of oral, to preserve "n". That yields "xileŋu".
We can tell there was probably a CC sequence between the 2nd and 3rd V's, since there's a consonant preserved there -- contrary to the tendency to delete voiced velar continuants between vowels. A sequence of CC during the application of that rule, would block them from being deleted, since neither C is intervocalic.
Then "u" harmonizes to front "i" and/or "e", yielding "xileŋü", QED. Or if you insist, deletion of 2nd V to produce CVCCV word shape, yielding "xilŋü".
I think the Proto-Mongolic word for "tongue" = "kelen" is from this Eskaleut source as well, since it doesn't look like the Proto-Turkic word "til". Turkic is the most Uralic of the 3 Altaic languages, and Tungusic the most Eskaleut of the 3. Mongolic is in between, to be judged case-by-case. Here, it looks like Tungusic.
There is a Proto-Uralic word for "tongue" = "kele", but the Samoyedic intermediate form, which Altaic would have to derive from, looks nothing like that. They mostly begin with "s", mostly lack "l", and no form combines both initial "k" and medial "l".
https://www.uralonet.nytud.hu/eintrag.cgi?id_eintrag=280
Probably this is a case of Uralic and Eskaleut sharing a word due to being half-siblings. But Mongolic and Tungusic got their word for "tongue" via Eskaleut, not its Proto-Uralic cognate (which is way too old to feed into Mongolic or Tungusic anyway).
BTW, the loan into Proto-Nivkh of "tongue" as "hilɣ" confirms that there used to be a distinct "ɣ" after the "l", not a velar nasal. Nivkh probably deleted the "n" syllable at the end, to fit its own monosyllabic word shape. Lacking the "n", there was no nasal to fuse with the "ɣ", which remained as is.
ReplyDeleteUnlike the Uralic word "kele", the Eskaleut form has final "ɣ", so for sure Tungusic got its word for "tongue" from an Eskaleut root. But I'll bet Mongolic did, too.
P-T "ʒep-" = "to eat"
ReplyDeleteP-E "kəpə" = "to cut, sever", "kəp(ə)kaʀ" = "to gnaw or remove meat from bone with teeth"
Semantically straightforward, since eating involves using the teeth to sever meat, and one derivative of the Eskimo root is about eating.
Since the P-T form is a root that usually has other material added, first we note that Eskaleut base-final "ə" tends to drop out when postbases attach, as in the form cited about gnawing a bone to remove meat. So we start phonetically from "kəp":
"ə" -> "e", "k" -> "c" -> "ʒ" before "e", yielding "ʒep-", QED.
Some of the derived Tungusic forms like "ʒepke-" may come from Eskaleut roots as well. That is the expected cognate of P-E "kəpkaʀ" = "to gnaw bone to remove meat", where the word-final C is deleted since there's already 3 C's, and "a" -> "e" to harmonize for frontness, while staying low and unround.
Maybe just a coincidence, but worth bearing in mind that Eskaleut can supply postbases / suffixes, not just bases.
This word, and "ʒuke", contradict the over-arching plan of inverting the feature redundancy of Eskaleut when inheriting words from that source. Eskaleut originally has voiceless stops and voiced fricatives, then Tungusic inverted that while building their lexicon. However, here we see medial "p" and/or "k", just like medial "k" in "ʒuke". Those are voiceless stops, not voiced.
What gives? I chalk this up to the late date of incorporation, when the rules stopped affecting medial C's and only targeted initial C's. We can tell they targeted initial C's more since the consonant "ʒ" almost only appears initially (16 entries), only a few times medially (and a couple more where it begins a suffix).
These initial "ʒ" come from Eskaleut "k" -- but clearly P-T didn't care to apply this rule to medial "k", else we'd see a bunch of medial "ʒ". The word "ʒuke" is revealing since medial "k" is also followed by "e", which would trigger palatalization to "ʒ" in initial position -- but not here, no change to "k" whatsoever.
By a certain point, they were mainly focused on distinguishing Eskaleut vs. Uralic origins using initial C's, not wherever they occur. And after a later point, they didn't even obsess about initial C's, as shown by "tuńga" and "töge", which are both of Eskaleut origin, yet begin with "t" instead of "d" or some voiceless fricative.
By this late point, most of the lexicon had already been constructed, and the division between Uralic and Eskaleut origins had already been made by the inversion of feature redundancy on the Eskaleut side. Any late-entering words were not subject to such strict rules to identify which origin they came from. They were too few in number to alter the overall pattern of Uralic keeping the original redundancy, and Eskaleut having the inverted one.
P-T "ʒuban" = "10"
ReplyDeleteP-E "kiput" = "to exchange or reverse", "kipułəɣ" = "to exchange places or alternate", "cipə(t)" = "to exceed or be left over", "cip(ə)nəʀ" = "excess"
Either "kipu" or "cipə" supply a semantic basis for "10". "Reverse / alternate" means you've counted on your 2 hands in a certain direction, now reverse it to get to 20. Or you're exchanging which part you're counting, as Eskimo speakers leave the hands after 10 and go to the toes / feet for 11-20. "Exceed, be left over" means the upper limit has been reached, you can't count past 10 on your two hands in a certain direction.
They seem to share a simple root "kip" about becoming full, which is clearly linked to "exceed", but also about reversing / alternating / etc. -- which you do when one way of doing things has reached its limit, then you start doing things some alternate way.
I don't think the final "n" in the P-T word is from Eskaleut -- just about all numerals in P-T end in "n", so it's part of a P-T numeral template, not meaningful. We just need to arrive at "ʒuba".
The P-EA root underlying both of the subsets of words in P-E, beginning with either "ki" or "ci", is "kipV", where the V doesn't matter here, as we'll see.
"k" -> "c" -> "ʒ" before "i", "p" -> "b", and postbase "a" added (passive / resultative participle), then "ʒ" dissimilated from "i" by shifting back to "u" to harmonize with back "a", while ignoring rounding harmony (not very strong, compared to front-back harmony). That yields "ʒuba", QED.
The passive / resultative postbase means "that which has been X'ed". In this case, 10 is the number where "the limit has been exceeded", the maximum has been arrived at. Or if it's about "reverse, alternate," 10 is the number where "the counting method has been reversed / alternated", i.e. you're going to switch from hands to feet, or undo the actions you made when counting from 1-10, in reverse order.
P-T "ʒügün" = "east; left"
ReplyDeleteP-E "ciqinəʀ" = "sun, sunlight"
This is now a doublet in Tungusic of the Eskaleut root for "sun, sunlight", along with "sïgūn" = "sun", which is an early-entering word. We can tell by the "ʒ" that "ʒügün" is late-entering, and may have a somewhat different set of rules operating.
The sun rises in the east, so that's where sunlight begins. And when you face south, your left hand is pointing toward the sun. Japanese speakers also faced south when coining their direction words, like how "hidari" (from "pidari") means "left" and contains the word "hi / pi" = "sun" in it. Facing south does not contradict them belonging to a solar cult -- they were simply orienting toward the sun *at its peak intensity* in the afternoon, not when it just begins to show itself in the morning.
The final "əʀ" is superfluous since there's already 3 C's.
"q" -> "g"
The P-EA root begins with "kə", and again I don't understand why schwa raises in this case, but it does. Maybe it raises at the stage when the preceding C is "c", which is the standard palatal consonant, and has the most "i"-like influence on surrounding segments.
"ə" -> "e", "k" -> "c" -> "ʒ" before "e". And "e" -> "i" before "c", let's say. That yields "ʒigin".
Then "ʒ" is dissimilated from "i", but vowel harmony can't do that since the other V is identical. The only way to impede palatalization without moving the vowel, is to round it, so that solution can also be used for dissimilating a palatal C from "i". They rounded the 2nd "i" just for good harmonic measure, yielding "ʒügün", QED.
Aside from "ʒ", the solution of rounding a front vowel rather than shifting it back, is another sign that this is a late word, unlike the early word "sïgūn", which shifted "i" back to "ï" to dissimilate "c" from "i", and which made "c" -> "s" (voiceless fricative) instead of "ʒ" (voiced stop / affricate).
"Sun" as in the celestial body is more fundamental to the lexicon than direction terms like "east" and "left", especially "left", which is an egocentric direction term, and not even conceived of until people begin thinking of themselves as the center of a set of coordinates, or like the center of their own little universe. Very introspective / rise of consciousness, not like geocentric direction terms.
So it's no surprise to see "sun, the celestial body" show early sound laws, and "east, left" show later sound laws, even though they're both adapting the same Eskaleut root about "sun, sunlight".
"e" -> "i" after "c"
ReplyDeleteTypo: I meant to say that P-EA "sun" originally began with "tə", not "kə". Both of those can ultimately palatalize to "ce" and then "ci" in P-E, where "sun, sunlight" is "ciqinəʀ". That is cognate with P-T "töge" = "winter", i.e. the "pouring-down" season, when sunlight is pouring down due to the Siberian High.
ReplyDeleteOnce the evolution of "sun" reaches the point where it begins with "ci", it can be adapted into Tungusic by either making "c" a voiceless fricative, like "sïgūn", or a voiced stop / affricate, like "ʒügün", to identify its Eskaleut origin.
So, "ʒ"-initial words in P-T don't only come from "k" being palatalized to "c", but "t" being palatalized to "c", and from there voiced to "ʒ". And of course, from words that were always "c"-initial back to P-EA, since "c" is not just a destination for other consonants that have become palatalized, but an original phoneme in its own right.
It doesn't look like many "ʒ" words came from earlier "t", since "ʒ" words are late-comers, and by that point, speakers were less avoidant of Eskaleut "t", like "töge" and "tuńga". But we'll see if some of the remaining "ʒ" words have an earlier "t" or "c" origin...
The ones whose 1st V is "a" are unlikely to be from "t" or "k", since they don't have palatalizing vowels -- that goes for the 2nd V as well, in those words. Probably those are from original "c", which can be followed by back vowels "a" or "u" with no problem, and doesn't need a palatalizing front vowel to wind up in palatal position -- that's it's original position!
P-T "ʒabi" = "boat"
ReplyDeleteP-E "caviɣ" = "knife", "cavaɣ" = "to work or craft", "caviʀuʀ" = "to whittle"
The root "cav" has to do with making crafts. The main tool used was a knife, but later in NW Eskaleut the word for "knife" came to refer to the material its blade is made from, "iron", and then "metal tool" in general, typically ones with a blade, though, like an axe or saw.
Large-scale crafting of wooden materials using a bladed tool was distinguished from fine-scale woodworking, as shown by the dedicated verb for "to whittle".
In NW EA, "caviɣ" can function as both a noun and a verb, i.e. "to use a knife", namely in craftmaking.
There's a old non-productive fossilized postbase "i" which means the result or state left over after the verb is completed. It even shows up in a verb for woodworking with bladed tools -- "cana" = "to carve", "cana + i" -> "cania" = "wood carvings, shavings, chips".
Attach that postbase to "caviɣ", used as a verb, meaning "the product of using a knife" i.e. during craftmaking. This refers to boats as crafts that are crafted using metal bladed tools, to dig out the interior, to plane the surface smooth, and so on. For forest and riverine cultures, dugout canoes are one of the main large-scale crafts that they make and use.
Also, P-T "ʒabi" is distinct from "temü" = "raft" or other floating platform sort of watercraft. That can be thrown together without too much crafting, whereas a dugout requires more labor and crafting and bladed tool work. Hence the emphasis on "metal blade tool" in its name.
From P-EA "caviɣi":
"c" -> "ʒ", "v" -> "b", "ɣ" -> 0 between vowels, 2 "i"s contract since long vowels banned outside of 1st syllable, yielding "ʒabi", QED.
Here's an example of "ʒ"-initial Tungusic words coming from original "c" in Eskaleut, not "t" or "k" that had been first palatalized to "c". The hint that it's original "c" is the following vowel being back. Technically, the vowels could have been "i-a" and metathesized to "a-i", but there are no relevant Eskaleut words "civa" or "cipa" relating to boats, or beginning with "k" or "t" either.
P-T "ińe" = "to laugh"
ReplyDeleteP-E "əŋlaʀ" = "to laugh"
"ə" -> "e", "ŋl" -> "ń" (fusing "nasal" and "palatal continuant" features, which we've seen before), then "e" -> "i" before palatal "ń", "a" fronted to "e" to harmonize with "e/i" (while remaining low and unround), meaningless word-final "ʀ" deleted. That's "ińe", QED.
Wiktionary notes that Doerfer said the P-T word was borrowed from Proto-Mongolic "hinixe" ("to laugh"), on the basis of Mongolian "ineex". These two words do seem cognate, but they must both be Eskaleutisms, possibly independently adopted into their lexicons. The Tungusic word does not have a final back fricative, which Mongolian does, and P-E does as well.
So perhaps Mongolic inherited this Eskaleut word with its own distinctive set of sound laws / phonotactics, which preserved the back fricative "ʀ" as "x", while Tungusic had different sound laws / phonotactics, which deleted it.
With further investigation of Mongolic words of Eskaleut origin -- adapted straight into Proto-Mongolic or Para-Mongolic, not borrowed from Tungusic -- maybe we can distinguish these different sound laws. Like, does Mongolic fuse "ŋl" into "n" or "ni", or is that fusion something that only Tungusic does, into "ń"?
If there's a borrowing, it would go from Mongolic to Tungusic, since there's a reason for Tungusic to delete final "x", but not a reason for Mongolic to append final "x". But again, Tungusic has the largest share of its lexicon -- and other phonological, morphological, and syntactic aspects -- coming from Eskaleut, compared to Mongolic.
So there's no need to assume a Mongolic source of this Tungusic Eskaleutism. I'm guessing Mongolic inherited this Eskaleut word as well, just applying slightly different phonotactics to it, than Tungusic did.
Turkic seems to lack it, though -- no surprise, it's too far away from Manchuria / Amur, to have Eskaleut as a major parent in its creolization. Maybe an animal name or tool name could be borrowed from the Eastern Altaic languages into the Western one, but "laugh" is not such a word.
P-T "ineŋī" = "day"
ReplyDeleteP-E "iŋət" = "to sit down", "iŋləʀ" = "sleeping platform", "iŋŋaʀ" = "to lie down to sleep"
Now that I'm getting familiar with Eskaleut postbases, I immediately spotted the privative one "ŋŋit" at the end of the Tungusic word for "day".
Actually, the Tungusic word originally referred to "dawn", the transition point between night and day, not "day" itself. Words for "dawn" are more likely to be morphologically complex than for "day" since "dawn" is more specialized -- a certain point or transition, not an open interval of time. And since "dawn" is more specific semantically, it tends to get more phonetic material to specify what that extra nuance of meaning is.
Later, it was extended to mean "day". Then, as it happens, the word for "day" was abbreviated of its "suffix", to give "ine" = "dawn", all over again. But by that point, Tungusic speakers were not so aware of what the original Eskaleut postbase meant. Tungusic did not begin with "ine" = "dawn" and then add a suffix "ŋī" to derive the word for "day" -- again, "dawn" must be more complex and derived, while "day" is simple and underived (at least, not derived from "dawn", maybe "sun" or something).
It's not clear which verb relating to "resting" was chosen here. So far, we've seen "ŋl" fuse to "ń", and yet "ineŋī" has alveolar "n". Maybe it's an irregularity. Or maybe it's from the verb for "to sit down", which metathesized "ət" -> "tə" to produce the standard medial CC cluster. Then the sequence "ŋt" fused to "n" -- fusing "nasal" and "alveolar" features.
Semantically, they mean slightly different things, but the same thing overall -- "day" is the time of the 24-hour cycle when you are deprived of sleeping in bed, or when you're deprived of sitting down, since your daily chores and activities require you to be up and moving about. Again, this makes more sense if the original sense was "dawn" -- the point at which you become deprived of sleeping in bed. "Day" lasts too long after that for you to still be thinking about "bed", you're already into daytime chore mode. "Dawn" is more relevant for "getting out of bed".
From "iŋət", which will get transformed into "iŋtə", and the postbase "ŋŋit", we start with:
"iŋtəŋŋit"
"ŋt" -> "n" (fusing "nasal" and "alveolar" features), "ə" -> "e", geminate "ŋŋ" simplified to "ŋ" since geminates are banned / rare in P-T, following "i" lengthened to compensate for deleted "ŋ" in former geminate pair, word-final C deleted cuz it's not "n" and there are already 2 C's after the 1st V, not to mention 3 V's, and the word is too long already.
That yields "ineŋī", QED.
The phonetic derivation from "iŋləʀ" is more problematic, requiring irregular "ŋl" -> "n" instead of "ń". Then, it's simple: resolving illegal CCC cluster "ʀŋŋ" -> "ŋŋ" (voiced back continuant being absorbed into a geminate voiced back continuant pair), then as before.
Either way, the Tungusic word is of Eskaleut origin, and originally referred to "dawn" rather than "day". At least in pure Eskaleut it did, perhaps when adopted into Tungusic, it shifted meaning to "day", if that Eskaleut privative postbase was not understandable as referring to a point in time when something was ended or deprived.
P-T "ʒalan" = "era, age, times; generation", suffix "n", root "ʒala"
ReplyDeleteP-E "cilakiɣ" = "to be good weather", "cilałuɣ" = "rain or bad weather", "cilaqliʀ" = "outside part or neighbor"
The P-E root is "cila" and refers to the external environmental conditions, like "weather" and "outside". The sense of "neighbor" ties it into a social kind of weather condition. Even in English, we refer to a zeitgeist / era / age / etc. as a "climate" or "atmosphere", using broad terms for the overall weather conditions in the environment, not what life is like within a private home. The zeitgeist pertains to everyone in the community, just like the weather.
I couldn't find any relevant Eskimo words beginning with "k" or "t", so this must have been "c" to begin with. However, P-E merges "cə" and "ci", so this could come -- and does come -- from "cə", judging from the Tungusic comparison. It must have been "cəla" in P-EA.
"c" -> "ʒ", "ə" -> "e", then to dissimilate "ʒ" from "e", "e" pulled back (while still low and unround) to harmonize with "a", becoming "a" as well. That yields "ʒala", QED.
P-T "ʒalum" = "full"
ReplyDeleteP-E "ciləɣ" = "full; thick, wide"
No relevant Eskimo words beginning with "k" or "t", so this is from "c" to begin with. The fact that the Tungusic cognate has a low vowel as the 1st V shows that the P-EA word was "cələɣ", which cannot be seen from P-E alone since it merged "cə" and "ci".
Contra Wiktionary, there's no Tungusic suffix "m", in fact this is the only P-T word that ends in "m", so it's very unusual and requires a special explanation. It does come from a suffix, but in Eskaleut -- the resultative / stative postbase "uma", which attaches to a verb and means "being in a state of having completed the verb". This contains the postbase "u", which is the copula "to be".
So, just "u" attaching to "cələɣ" means "to be full", and adding the "ma" means "being in a state of having been filled" -- that is, simply, "full". It just adds a connotation of some activity having taken place, which produced the full state of the object as a result -- like when you pack an empty suitcase with stuff, and it becomes full. That is one use of this word in Eskimo, a container being packed with stuff. Same as in English "to be full" after eating a meal, i.e. your formerly empty stomach was packed with food, and now it's in a state of having been filled.
Starting from "cələɣuma":
"c" -> "ʒ", "ə" -> "e", "ɣ" -> 0 between vowels, yielding "ʒeleuma". To resolve sequence "eu", "e" is from weak vowel "ə", so it simply drops out, leaving "ʒeluma". Then "ʒ" is dissimilated from "e" by "e" being pulled back (while still low and unround) to "a", to harmonize with back "u". Final "a" deleted -- since there are already 2 V's, suiting Uralic / Tungusic word shape, but creating the only strange occurrence of word-final "m", which may not have sounded too bad since it's at least a front-ish nasal like "n", the only common final C. That leaves "ʒalum", QED.
An even more standard word shape would have resulted from deleting the "u", leaving "ʒalma". Why was this not chosen? Perhaps they wanted to emphasize the Eskaleut postbase "uma", which requires keeping the "u", and maybe the "a" was dropped to emphasize the "m" (as the only example where "m" is final). IDK, but anyone's theory will have to explain this sole occurrence of final "m".
Notice that Tungusic and Eskimo share 2 near-homophones, for "climate" and "full", which are semantically unrelated. However, within both P-T and P-E, the roots of these words share 3 of 4 segments, in order -- P-T "ʒal" (then "a" vs. "u") and P-E "cil" (then "a" vs. "ə"). They also share one of the distinguishing final segments -- "a".
ReplyDeleteHomophones cannot be shared between unrelated languages, since, as rare as it is to match one word between languages, matching an unrelated sense with the same sound is equally rare, and multiplied since a homophone pair is a joint event. Rare times rare equals impossible.
These two aren't exactly homophones, but they're damn close. They are not due to coincidence, since it's too unlikely. Rather, their similarity reflects them sharing the first CVC, and only differing on the final V. That final V is not a derivational suffix, since they're not semantically related. They simply happen to sound so similar within each language, due to an accident of sharing the first CVC, and only differing on final V.
For two languages to share these accidental outcomes, means they must be related. Sure, two languages could have a false cognate by accident -- but for that word to also be part of a homophone pair, or near-homophone pair, where the unrelated 2nd sense is the same in both languages? Impossible.
Even if the supposed false cognate in each language did have a homophone or near-homophone, there's nothing driving the 2nd meaning to be the same, since homophony is accidental within all languages.
Suppose you think "climate" is a false cognate in both Tungusic and Eskaleut. Then it turns out there's a near-homophone for that word in both languages. And I tell you that in Tungusic, the near-homophone for "climate" is "full". What do you conclude about the near-homophone for "climate" in Eskaleut? Nothing whatsoever, if you think they're unrelated languages. It's just a member of a homophone in its own language -- the other member of the pair could be "foot", "mother", "to see", or anything else unrelated to "climate". And yet, it's "full" as well in Eskaleut.
Conclusion: they're related by shared ancestry.
P-T "ʒapa" = "to take, grab, hold"
ReplyDeleteWiktionary notes a close match to Korean "japda" (ditto), suggesting a 3rd consonant
P-E "cavət" = "to touch or feel", P-I "captək" = "to grope", P-YS "cavłəɣ" = "handle"
These verbs mean the same thing, and it appears in the P-YS word for "handle" ("cav", or maybe "cavł" reflects "cavt").
It's odd that the Tungusic word begins with "ʒ", which was a strategy to deal with Eskaleut roots (voicing a voiceless stop), yet also has "p", which should have been voiced to "b" for the same reason. But the Korean loan -- or inheritance? if Korean has Eskaleut roots as well? -- shows there used to be an alveolar stop after "p", and that the two medial C's had opposite voicing.
So the P-EA form was likely the same as P-E, "cavət". From there:
"c" -> "ʒ", "ət" metathesized to "tə" to fit standard Uralic / Tungusic word shape of CVCCV, and "ə" harmonized with "a" by moving back (while remaining low and unround) to "a", yielding "ʒavta".
It's unclear what precise route goes from "vt" -> "p", but the overall logic is clear -- "v" wanted to voice to "b", but was devoiced back to "p" or was prevented from voicing in the first place, by the following voiceless "t".
However, since there's no C after "p", it looks more like the two fused together, from either "vt" or "bt". The sequence "bt" only shows up once in P-T, "xebtilē" = "rib" (an early word, judging from "x" instead of "ʒ"). So it's very rare, but not impossible. The fusion must have taken place when the sequence was "vt", then -- labial place from "v", voicelessness and stop manner of "t".
In any case, that yields "ʒapa", QED.
Next is a real wild dogsled ride, hold on! P-T "ʒapkun" = "8", also reconstructed as "ʒabkun" with later devoicing of "b" -> "p" due to following voiceless "k".
ReplyDeleteThe expected cognate in Eskimo, with the least complication, would be "cavkun" or "capkun". So we scroll down the list of Eskimo reconstructions and find this humdinger of a riddle, as something somehow related to the number 8:
P-I "capək" = "drop (n)", "capkut" = "dropper tool", i.e. one that controls the drip of the liquid for precise aiming, as well as not wasting the liquid during pouring.
We'll start with the phonetic derivation, which is easy. Begin the base of "capək":
"c" -> "ʒ", "ək" methathesized to "kə" to fit standard word shape of CVCCV, yielding new base "ʒapkə". To that, add the P-E tool / body part suffix "un", "ʒapkəun", meaning "a dropper tool", now with the connotation of a body part as well.
Base-final "ə" drops out as usual in Eskaleut, yielding "ʒapkun", QED.
Now for the wild-and-crazy exploration of how "dropper tool / body part" relates to "8". Well, numerals come from counting the fingers on the hand, and no matter which direction you start with (pinky or thumb), 8 lands on the middle finger.
ReplyDeleteOK, so how is the middle finger a tool for dripping liquid?
The New World Eskaleut, as well as Arctic Tungusic speakers like Even and Evenki, presently-speaking Yakuts, and linguistically unrelated Chukchi, all share a cultural practice to ice the runners of their sleds, so that there's the least amount of friction as possible. With a thin coating of ice, they have a nice smooth glassy surface, not the rougher texture of the wood, bone, or antler runners themselves.
In the freezing conditions of Siberia and the Arctic, water is a scarce resource -- it freezes so easily. You don't want to waste it. To coat the sled runners with ice, though, you need some water to pour on them. Usually this comes from snow that is held in a storage pouch, then placed into the mouth to warm it up into water. But you only have a mouthful of water to work with, and don't want to waste it -- not to mention, there's not a ton of snow in your storage pouch.
If they just spat the water directly from the mouth onto the runners, that would release too much water at once, wasting most of it, and would not precisely target just the runners, also wasting it. The process would have to be repeated too many times, just to get a single coat of ice.
If they kept their lips only slightly open, like drooling the water out, that would control the speed of the flow, and not waste it as much as spitting. But their aim would still be very scattershot, since water dripping through a simple hole in the bottom of a container doesn't have any guidance or aim.
For a slow pouring speed, and for precise aiming, a spout is needed -- just like any teapot, oil lamp refiller, or gardening water can. Well, they didn't make dedicated vessels with spouts for this purpose, since they would be too bulky, require too much crafting for such a simple task, and so on.
So they improvised -- they used their fingers as a kind of water-slide to guide water away from their mouth. They placed the hand, palm-down, so that the main row of knuckles was directly under the lower lip. Then the lips were slightly opened, to let water flow down. Now, rather than falling straight down, they slid down the length of a finger and dripped off the fingertip. Since their finger is flexible, they can aim the drip for better precision.
And a fingertip is still pretty narrow, so it doesn't disperse and have bad aim, as spitting does. Regardless of where the water is flowing down the finger -- upside, downside, innerside, or outerside -- it all gets channeled back into the center of the long axis when it reaches the fingertip, making it more like a funnel or eyedropper, precise rather than scattershot.
Once they dripped it precisely and without missing / wasting any water, they immediately smoothed the water all over the runner, and just moments later, the freezing environment turned it into a thin coat of ice!
Imagine trying to discover this Tungusic word's origins, without knowing Eskaleut or Eskimo cultural practices, like icing their sled runners! You'd be hopeless! Leave it to the cliff-dwelling sage in the ruins of the blogosphere, I will wander down any secret passageway, as long as I know it will lead to some hidden treasure. ^_^
Why didn't "pk" voice in "ʒapkun", BTW, since this is of Eskaleut origin? Again, it could have begun as "bk", then devoiced back to "pk".
ReplyDeleteIt's a numeral, and not even a milestone numeral, so likely a late word, when they were less strict about inverting the feature redundancy of Eskaleut. By this point, they were mainly focused on the initial C for distinguishing Uralic vs. Eskaleut origin, not so much the medial C's.
Later still, they didn't even worry much about initial C's, but that's near the end of constructing the lexicon, when a few exceptions cannot alter the over-arching pattern of Eskaleut roots having voiced stops / affricates and voiceless fricatives, and the inverse for Uralic roots.
Why was the middle finger used for the makeshift spout? Sometimes the index was used, but the middle finger was preferred cuz it's in the center of the hand, so water is less likely to roll off either side of the hand and miss the target.
ReplyDeleteIt's also longer, and that provides more of a defined path for it to travel, before it leaves the channeling tool, so it has less chance of deviating once it leaves the vessel. Similar to a long-barrel vs. short-barrel gun or a watering can or oiling can with a long vs. short spout.
And it's still easy to utilize, since it's not the ring or pinky fingers, which are harder to move on their own. It's either index or middle finger, for dexterity -- and since the middle finger is longer and not on the edge of the hand, it's favored as the dropper finger.
BTW, returning to Funnel Beaker language sound-symbolism, which has been retained into Germanic and even modern English -- lots of "dr-" onsets relate to liquid falling down in a controlled manner from a source or container. Especially when the rime has "i":
drip, drizzle, dribble, drivel, drink, and I'm guessing "trickle" was probably "drickle" originally. ^_^
Before I finish up the "ʒ" words in a moment, it's worth emphasizing how bizarre they are phonotactically. There are 16 entries at Wiktionary, and only 7 fit the standard P-T word-ending of a vowel. The majority have non-standard endings, very odd.
ReplyDeleteEven the final consonants are strange. For words beginning with any other sound, the only allowed word-final C is "n", plus a plural suffix "l" added to the word for "child", in the word for "children" = "pöril".
But the "ʒ" words have one word each ending in "m", "p", "g", and "r", which have nothing in common with each other or the only common final C, "n".
This is another sign that these are late-entering words, when the standard word shape was no longer so vigorously enforced, and when the phonotactics of final C's was thrown out the window. And, to the main point, when words of Eskaleut origin were not systematically subjected to "inversion of feature redundancy", sometimes only targeting the initial C, sometimes not even that.
So if the sound laws used for "ʒ" words seems less uniform than in the other cases, that's true. The laws of the early, foundational stage had begun to weaken, probably cuz pure Eskaleut speakers were increasingly rare, and Tungusic speakers didn't need to so rigidly distinguish words of Eskaleut vs. Uralic origin. Just get them in the lexicon, and wrap it up already! Nobody's bilingual anymore anyway, so what does it matter if they can tell which words come from which source?
P-T "ʒukun" = "otter"
ReplyDeleteP-E "cikuʀ" = "needle or sharp tip", "cikiaʀ" = "pickaxe"
The best I can make of this name is that it refers to the otter's coat looking spikey when it's wet (which is how most people see them, since they're riverine or sea-based animals). It doesn't have that fine-grained look that dry land animals of a similar look have, like weasels, even though otters are also mustelids like weasels. The dry land ones don't have the big spikes that result from the fur matting into large spikes, after being soaked in water.
So, something like "the one with the spikey body", with the comparison to other mustelids being tacitly understood.
To the base of "cik" or maybe "ciku" = "spike", add the "un" postbase for tools or body parts, since the otter's spikes are formed by its fur, a body part. If the base is "ciku", then "ciku + un" shows a contraction of the 2 "u" vowels, since long vowels can't exist outside the 1st syllable (unless belonging wholly to a suffix).
"c" -> "ʒ", then "ʒ" dissimilated from "i" by "i" pulled back to harmonize with back "u", making it "u". That yields "ʒukun", QED.
P-T "ʒūg" = "house, dwelling, yurt, chum, tipi, etc."
ReplyDeleteP-E "civu" = "in front", "civuqaʀ" = "area or part in front"
The whole dwelling is referred to by one of its most salient parts -- the front part, before you even enter through the door / portal / entrance. This root, in the context of dwelling, refers to a porch or veranda or anteroom / mud-room, etc. For pit-homes, it refers to the long approaching passage before the entrance to the main residential room.
The original word could have been the same as P-E "civuqaʀ". Starting from there:
"c" -> "ʒ", "q" -> "g", word-final C deleted since it's just a dictionary form case ending and there's already 3 C's, yielding "ʒivuga".
"v" -> 0 between vowels, and "ʒ" dissimilated from "i" by "i" being pulled back to harmonize with back "u", becoming "u", and the "uu" sequence becomes "ū" since long vowels are allowed in 1st syllable. That yields "ʒūga", which requires an irregular or unmotivated deletion of the final vowel to yield the required "ʒūg", giving it the very odd word-ending of "g" (the only word in all of P-T).
But anyone's theory will have to accept the totally irregular ending of this word.
It's a late word, and clearly the rules about word-final V's had relaxed (at least for "ʒ" words). So perhaps final "a" shaded into a schwa, then deleted due to weakness and lack of stress.
At least there's no voiceless stops in this Eskaleut-origin word, as in "ʒukun" just before.
And that's every "ʒ" word in P-T -- every single one from Eskaleut origins! ^_^
Interlude on backwards hats in American cultural history and ethnogenesis. I'll be writing more, but first just want to file an IRL report. I tried to just ignore it, or sleep on it, but she's still in my thoughts a week later, so...
ReplyDeleteLast week in the thrift store, out of nowhere a girl appeared who looked like she'd transported from the '90s. Light-to-medium blue loose-fitting jeans, her only top was a dark magenta bandeau wrapped around her flat chest, tan skin on full display, and a teal ballcap ("dad hat") worn backwards. Not a fitted hat, but the strap-closing kind, where the strap goes across the forehead and/or bangs.
What I'll write about later, is the place of this kind of hat in the history of distinctly American clothing styles. It's one of the least Euro-looking or Old World-looking things we have ever come up with. But it actually ties into a longer history of distinctly American headwear, in fact it was the last new variation on the theme, before American culture died in the 21st century.
Of course the color combo of purple and green instantly took me back to the '90s, but it was more than that with her. She looked just like a girl I went to middle school with, when those styles were new and exciting. Tall, slender, tan, medium-length light-to-medium brown hair parted in the middle, cute girl-next-door facial features... I had to double-check my yearbooks from that time, and yes, this girl was a reincarnation in more ways than the sartorial. ^_^
She noticed me back, and we danced around each other in nearby aisles, but I don't think she wanted to be approached, just admired for her retro '90s / y2k look that she probably researched on TikTok (whoever taught her did a good job). Plus she was in her late teens or early 20s, might not want to hear some literal '90s teen talk about the good ol' days, which she never got to experience...
However, a few songs later, the speaker system started playing "There She Goes", the '90s cover by Sixpence None the Richer. OK, I can take a hint from the music gods. I would have been whistling along to that song anyway, but I made an effort to try and find her, and she found me, too, both walking down a main corridor toward each other, while I kept whistling the tune, and gave her some eye-contact. She didn't return it that time, but she didn't swerve away either, she was swaying her head side to side to the melody, smiling, wide-eyed, etc., just looking forward instead of locking eyes like when we first saw each other.
ReplyDeleteIf it had been the actual '90s, or even the 2000s, we would've chatted each other up a bit, but since it's the 2020s and all social connections have been severed permanently, no such luck. Not even being a hot and stylish guy will provoke a verbal response from them anymore... I wasn't trying to get her number or anything, I just wanted to gush about what a rejuvenating and welcome breath of fresh air she was for a '90s teen. But I think that message got across to her nonetheless, since I was whistling an iconic '80s-then-'90s tune right at her as we passed by each other.
She didn't have a device out, wasn't staring down at anything or off to the side. It was just that semi-nervous thing girls do where they look straight ahead, and hope to catch you looking at them out of the corner of their eye. They don't want to come on too strong, seem desperate or eager, etc. We didn't do that in the '90s, it was a more socially harmonious and cooperative time. But since the late '90s in fact, and especially during the discordant phase of the 50-year civil discord cycle, which peaked in the early 2020s, every social interaction is a zero-sum contest or game of who can be the most aloof and get the most attention.
She really wasn't that bad of an example of that trend, it just kept it from being an authentic '90s moment. But I guarantee she's never had a hot guy whistle her a melody face-to-face, let alone to such a romantic pining song, from her preferred zeitgeist, no less. So maybe that psyched her up enough, so that next time, or the time after, she'll be more comfortable around boys in public.
She was pretty comfortable, but she still had her guard up somewhat... and the only thing that can counteract that, is for random guys to lower their guard around her, to show her that fair is fair, guys are willing to drop their guard if she's willing, and it's not a zero-sum contest, social interaction should be a cooperative game of exchanging eye-contact, a few pleasantries, maybe flirting, and then going on with the rest of your day feeling nice and warm and fuzzy inside.
I know, what am I complaining about, if it was 80% a good experience? I'm not complaining, I'm a chronicler, and I'm just noting that although we seem to be out of the absolute abysmal depths of the 2010s, regarding the "staring down at devices" trend, we're still not even close to being out of the woods.
ReplyDeleteAnd she was just at the right age to be going through a social growth spurt, when the Covid hysteria shut everything down, depriving Zoomers of socialization at a critical developmental period. It's a miracle she isn't bedrotmaxxing like most girls her age.
I almost wanted to thank her out loud, for leaving the house and showing up in public, and putting in the effort to look cool and fun-loving. It really was a sight for sore eyes, girls of all types are just totally under self-imposed lockdown for the past several years, after Covid was over. The only time you catch them out and about is going to the grocery store, to avoid starvation. Otherwise, they just won't go out, not to stores, parks, or anywhere.
Even 10 years ago, they and their friends would be out to some degree, normies, alt girls, and everyone in between. They were not as socially open while in public, as their counterparts from the '80s or early '90s, but they at least did leave the house and want to be in public -- just with their guard up. So this girl was at least a breath of fresh air from 2015.
But for that first moment when we locked eyes, she did seem like she was straight outta the '90s, and we were in a high school hallway on the first day of school. Which recently began, didn't it? Maybe she was in literal first-day-of-school mode.
Everyone always made such an effort to make a statement on the first day of school... I doubt her school let her show up wearing only a bandeau on top, but maybe she had on a baggy shirt over top, semi-unbuttoned, and then took it off once school was over. Or she's in college, and there's no such dress code.
What is it about a girl wearing a backwards hat? Well, it's a distinctly American style, so it's like seeing "American folk costume" on display IRL, and that's refreshing, so that American culture isn't just some online museum exhibit, but something you can experience IRL.
ReplyDeleteBut it's more than that, it's revealing a character trait, like she's confident and socially independent, not insecure, not hive-minded, but still wanting to engage with others rather than wall herself off.
You can't make a fashion statement without wanting some kind of back-and-forth social engagement. But if you're doing something a little outside the majority, you're also showing you're confident and not obsessed with peer-pressure and so on. And yet, it's not "outside the mainstream" in the sense of attacking American culture like a resentful iconoclast or destroyer or corroder, just wearing something that not everybody else wears, while still being recognizably all-American.
"Why is your hat backwards?" IDK, it's just quirky and fun -- what's wrong with good-natured fun? There was nothing confrontational or in-yer-face about the backwards hat of the '90s. It was just cool, and lighthearted and good-natured.
It's a little provocative, but in a cheerleading kind of way, provoking a positive response. Livening up the mood a little. A one-woman cheer squad...
That was a major part of the appeal about the backwards-hat-girl -- she didn't have a clique of girl friends who would act hive-mindedly toward you. She was her own person -- as shown by the backwards hat! Not everybody wears their hat that way, y'know. Nothing wrong with girls being hive-minded and clique-y, but it *is* a breath of fresh air when one of them is comfortable and confident on her own, so you two can just interact on your own, without her friends trying to block you, criticize you, and so on.
Most girls HATE girls like that, since she's cutting them out of their little social managerial role as gatekeeper of beings friends with her. She has decided to be her own gatekeeper, and if she likes someone, she likes them, and if not, not, but bowing to a clique has nothing to do with it.
As guys mature socially, it's less and less terrifying to be subjected to the inspection of a girl's clique-mates. You get used to it. But when you're just going through adolescence, you're already nervous around girls -- what's she going to think about you? But when there's a clique that she's a part of, now you have to be 2nd-order nervous -- what are her clique-mates going to think, about what she thinks about you? It's not just you they're inspecting, they're inspecting her and gatekeeping her.
She may be fine with interacting with you, but her clique-mates think that attitude is too cavalier, and they're going to insert themselves between you, in her best interest, or so they say. Maybe if she's the type who needs protecting, that's fine. But it's always a breath of fresh air when she isn't that type, and can just cut out the gatekeeping middlemen, or middle-girls as it were.
It's the same vibe that a girl gives off who's wearing only a bandeau on top, while having a small chest. It just doesn't phase her at all -- "who cares how big my boobs are?" Also, girls who say that usually have great buns anyway. ^_^ But even if she weren't a sex bomb somewhere else, it just doesn't phase her. She's too comfortable in her own body, whatever it happens to be shaped like.
ReplyDeleteAnd remember, girls usually don't dress to seduce random guys in the population at large. They mainly dress to impress other girls, or cattily compete with them, or whatever. It's mainly the opinions of other girls that they have in mind, not what guys think.
Hardly any other girls would egg her on to wear only a bandeau on top. That's showing way too much skin, and they don't like feeling like they're now pressured to show that much skin, or else lose the contest for male attention.
But she's not like them -- she's not showing that much skin in order to gain the upper-hand in the contest for male attention. She's just comfortable in her body, wants to RETVRN to nature, and is a free-spirited aspiring noble savage. Why not go out looking au naturel?
Maybe her clique-mates would try to prevent her from wearing that, in her own interests so that the small size of her chest wasn't so obvious. What will the boys think? "You should at least stuff your bra, if you're going to wear just a bandeau!"
But she doesn't really care what the boys think either -- not if they're that obsessed with big boobs. She only wants to interact with the ones who appreciate her gestalt attractiveness, not zooming in on where da juggzzz are supposed to be. Lithe, tan, cute, smiley -- what's not to be attracted to, juggzzz or no juggzzz?
I didn't spin around to check out her buns, but I'll bet they were great anyway.
Even if they weren't, she was too cute to ignore, she really lit the place up with her mood-lifting fun-loving energy or charisma (or vibe or aura or whatever they're calling it these days) -- something that a heaving chest or a bubble butt can never do on its own.
So, whether it's the opinion of the girl hive-mind or the dude hive-mind, she just couldn't care less -- not hating them or antagonizing them, just merrily marching on to the beat of a different drum. ^_^
https://www.youtube.com/watch?v=JmM-mKss54c
That kind of attitude is not just liberating in an individual sense, for the girl herself -- it puts the guy at ease, too, freeing him from the dread that he's being confronted by the girl-hive, instead of her herself. It's impossible not to cheer up around girls like her...
No one ever wrote a love song about, "I think we're part of a hive now". It goes, "I think we're alone now" -- "Just the two of us" -- and so on.
ReplyDeleteThat's how the "girl who's her own girl" makes you feel, right away, and the whole time you're with her. ^_^
https://www.youtube.com/watch?v=w6Q3mHyzn78
And now for a few style icons of the backwards hat girl heyday of the '90s. An early crush of mine, Al (the tomboy form of Alicia) from Step by Step. Nothing says down-to-earth like a forest green backwards hat...
ReplyDeletehttps://www.the-sun.com/wp-content/uploads/sites/6/2023/09/august-16-1993-842924466.jpg?resize=789,960
The protag from The Secret World of Alex Mack, flashing an engaging smile rather than a haughty or competitive look like her 2010s descendants would look. My younger brother had a huge crush on her.
https://www.imdb.com/title/tt0108921/mediaviewer/rm1673525761/
I went through my middle school yearbooks, spanning fall of '92 to spring of '95, and it seems like *the* year of the backwards hat girl was 1994. There's a collage of the trip that us 7th-graders took to the zoo, and there's a candid portrait of two BFFs, both wearing backwards hats! One of them even has a bit of her bangs dangling down over her eye -- and she wasn't even emo! Both of them were the pretty, popular preppy type. They idolized Alicia Silverstone from Clueless, before that movie even came out. It really was an all-American look, not just rocker chicks or rapper chicks or whatever else -- the popular girls who dressed preppy wore backwards hats.
In fact, one of the science teacher chaperons is wearing his hat backwards, too! He was definitely NOT the "I'm still cool" kind of teacher. He was the consciously fuddy-duddy, absent-minded professor type. That's what made it so funny, even he was joining the trend, which he must have seen the kids doing during the bus ride to the zoo. As I said when reviewing old yearbook videos, during the '80s and first half of the '90s, the school environment really was one great big happy community and extended family...
Also from around '94, more of a rocker or indie babe, manning the high school radio station, with not only a backwards hat, but a flannel shirt, and several rings with prominent stones in them. As hinted at by the Wayne's World soundtrack CD nearby, there was a resurgence of interest in the '60s and '70s during the '90s. Mood rings and other hippie / boho kind of jewelry was part of that, even if hippies didn't wear backwards hats.
https://www.taboracademy.org/the-rock-of-marion
Even Playboy covers memorialized the free-spirited backwards hat girl, quite earlier than other forms of pop culture, in fact. Here's the semi-SFW 1991 cover girl Julie Clarke, whose hot pink backwards hat is amplified by quirky large earrings, multiple bracelets and watches, multicolor gloves, and a pair of blue rollerblades -- the ultimate symbol of charting her own nomadic gypsy course through life, just cruising around.
ReplyDeletehttps://www.flickr.com/photos/63969880@N02/5872111739
More than a pretty face and hot bod, this kind of girl brings a spontaneous spark of joy to wherever her meandering ways carry her. In the good ol' days they really made an effort to make their models play characters or roles, by using wardrobe and set design, to humanize the models, not just "hey, getta loada dis chick's juggzzz..." There used to be style and glamour to it, and a good-natured relationship with the audience, too, not dehumanizing them by saying, "you're just subhuman scum that only wants to oogle da juggzzz, well here ya go, just some random naked chick with no context, showing off da goodzzz..."
This was many years before Rollergirl in Boogie Nights, and would have felt more like Olivia Newton-John's rollerskating goddess from Xanadu. Neither of whom wore backwards hats, though. Rollerskating was the perfect athletic connotation so that she wasn't too tomboyish, but still wasn't an ordinary girl. Surfer girl would have been another perfect choice. Or skateboarding, like in the video for "Free Fallin'" by Tom Petty.
Rollerskating has more of a graceful, dancing motion, like an angel swooping in to cheer up the lives of us mere mortals. You don't get that effect with other sports that girls play.
So graceful and angelic that the cover of Playboy hardly seems pornographic at all. She was a backwards hat girl, after all, impossible to objectify and treat like some random bod. She's so down-to-earth and relatable, it's very disarming to guys who would otherwise leer at an adult magazine cover. She's hot, but you don't feel like leering. You want to be her friend so that her joyful spontaneity is a constant positive presence in your life, even though you're aware these backwards-hat-wearing rollerblading types are like gypsies, and may wander into and out of your life in the blink of an eye...
That dizzying spell they cast on you leaves no time for you to objectify them. You're too wonderstruck, and that feels better than cheaply leering at some random bod. They make you appreciative, not greedy. Everyday angels, although not dedicated guardian types, more like Santa Claus who has to zip all around the world, lighting up everyone's lives.
Anyone who lived it, is still grateful for the free-spirited backwards hat honeys of the '90s. It's not just a costume you can LARP, it has to be an expression of your personality or character type. Back then, the girl next door had such a personality. Now it's vanishingly rare, but it's still just as appreciated when it swoops into your life -- and then, just like that, right back out of it...
The headband as distinctly American headwear, adopted from the Indians. Well, before getting started on this journey through American ethnogenesis and folk costume, this is what the backwards hat of the '80s and '90s was fitting into, the very last original variation on a long ongoing theme -- the headband.
ReplyDeleteSpecifically, a headband that crosses horizontally over the forehead rather than rise over the top of the crown, sitting lower rather than higher. The vertical kind is pretty American, too, but it's more of a variation on the original horizontal kind, and there's a continuous gray area where it's sitting at an angle. And the vertical kind is not borrowed from the Indians, so it's less iconic of our New World cultural milieu.
I was struggling to think about where this offbeat backwards hat trend came from, what it was supposed to mean... then seeing all those free-spirited girls sporting them in '90s pop culture (and IRL pictures preserved in yearbooks), it hit me -- they were tapping into the hippie / boho headband. It's a piece of headwear, with a prominent horizontal band running across the forehead, maybe showing the forehead skin, maybe the bangs instead, sitting lower rather than higher on the head, and the horizontal orientation is reinforced by the bill in the back side also being more horizontal than vertical.
That's why the fitted hats don't strike quite the same iconic all-American note -- it just looks like a skull cap, not a band or bandeau or bandana. That adjusting strap in the "back" of the hat was certainly not intended to be a visual band or bandeau, and certainly not intended to be seen from the front. It was just some utilitarian gadget to make them one-size-fits-all, and cost-efficient to mass-produce.
And yet, leave it to the American sense of defining a new style for a new nation, to repurpose that utilitarian / efficiency gadget for a purely aesthetic purposes, as an in-group marker and shibboleth, connected in a long chain to other across-the-forehead bands that root it deep in American culture.
There was a huge revival of late '60s and '70s culture during the '90s, but it wasn't just LARP-ing and reviving. Well, some of it was LARP-ing, like That '70s Show and Austin Powers and Boogie Nights. There were non-parodying earnest retrospectives like Dazed and Confused, too. And then there were totally original variations on the same underlying theme, since America still had a little gas left in its creative tank. And so, the backwards hat was born -- the kind with a strap! Otherwise, it just wouldn't look like a headband...
The '90s were coming off of the insane energy levels of the neon New Wave '80s, to a more lowkey and laid-back vibe. And Gen X began replacing the go-go cocaine-fueled careerist yuppie Boomers as the hip new generation. The era of the unpretentious slacker had arrived, the perfect opportunity to reincarnate the stylistic markers of their bohemian ancestors from the '60s and '70s. Emphasis on "reincarnate" -- same spirit, novel yet recognizable form.
As I said, everyone aspired to this style, and lived this lowkey anti-yuppie life. Yes, including the pretty popular preppy girls. Those two with the backwards hats were famous for being ditzy, not overly serious competitive strivers. Our 9th grade English (woman) teacher affectionately called them Tweedle Dee and Tweedle Dum -- adorable, ditzy, and definitely *not* model students.
ReplyDeleteBeing a study-nerd was still totally uncool in the '90s, so the popular girls, too, upheld the banner of the slacker, even if they weren't into alternative rock, skateboarding, smoking pot, or whatever other stereotypes of the slacker from the media.
They were cool with those kinds of kids, they just didn't need to be into grunge in order to feel like coasting and being cool rather than making all of life into a high-pressure status contest. I keep trying to emphasize for Millennials and Zoomers who only grew up with frenemies and mean girls, there was no such thing in the '80s or first half of the '90s. And that pair was one of the last to be referred to affectionately as "twins" rather than sarcastically as frenemies etc.
And they didn't team up in order to wage combat against the rest of their peers, like a mean girls clique. Girls in the good ol' days simply did not have a bitchy or catty bone in their body. Much like Tweedle Dee and Tweedle Dum, they were comic relief and enjoyable to be around.
It was fun being around just one of them, but if they were in a class together, their twin-brain went into overdrive, and their good-natured ditzy mischief could not be stopped -- not even by the teacher calling them Tweedle Dee and Tweedle Dum.
I really miss those two... ^_^
Starting with the mature phase of the headband's adoption as a marker of American cultural identity, the form that everyone now thinks of -- the '60s and '70s hippie / boho kind.
ReplyDeleteAs a sign of how much creative spark was still left in America, that zeitgeist produced several variations on the theme, just within its own time. There was the simple leather or suede strap, the one with a geometric bead grid design running across the front (and maybe the sides and back for good measure), the kind with tiny seashells across the front, the kind with a print on fabric, the really au naturel garland type made from plant stems, grass, and flowers... and that's just the basic material!
The way of securing them around the head varied as well -- keeping it as close to a ring as possible, or letting one or both ends dangle down dramatically, or fixing separate tails to dangle down from around the temple area, and whether the knot was in the center of the back or off to one side...
And if there were tails of any kind, were they simply pieces of fringe material, or did they have little feathers or shells or beads or other ornaments on the ends?
To get into the proper mood, rather than linking a whole bunch of separate images, see the many slideshows of the era, set to "San Francisco" by Scott McKenzie, with the memorable refrain, "If you're going to San Francisco / Be sure to wear some flowers in your hair".
https://www.youtube.com/watch?v=7I0vkKy504U
https://www.youtube.com/watch?v=PQHb4z0prjo
https://www.youtube.com/watch?v=pVrpqOzAbmQ
That last one is preparing for the Monterey International Pop Festival of 1967, where the song debuted. The event's promo poster features a repurposed photo of Canadian dancer Maud Allan, from her 1906 portrayal of Salome.
https://en.wikipedia.org/wiki/File:Monterey_International_Pop_Music_Festival_poster.jpg
This not only rooted the psychedelic style in Art Nouveau, it was part of American ethnogenesis, whereby we ditched our Euro and even Indo-Euro origins, and claimed descent from various Saharo-Arabian cultures instead, mainly Egypt but the Levant in general, the ziggurat-builders of Mesopotamia, camel nomads of the Sahara and Arabian deserts, and so on. And sure enough, the American (or Canadian) portrayal of Salome requires a headband rather than a full headscarf or veil. The jargon is "diadem" or "circlet", but it's what we recognize as a headband. It even has dangly tails with ornaments!
The 1906 portrayal of Salome shows that the exotic Saharo-Arabian version of this theme did not have to wait for the Egyptian mania that the flappers indulged in during the 1920s, after King Tut's tomb was discovered. It goes right back to the beginning of American ethnogenesis, after our integrative civil war wrapped up, during the late 19th and early 20th centuries.
We were so desperate to sever our links to Euro and Indo-Euro cultures, that we erected the world's tallest structure of its day -- an Egyptian obelisk, the Washington Monument, which remains iconic of American architecture long after its completion in the 1880s.
The heavily jeweled circlet never caught on as the main form of the headband in America, since it was considered too royal and decadent, not for everyday Americans who were still naive and earnest and innocent.
Don't let that Art Nouveau Salome fool you, the '60s / '70s scene was well aware of the true origins and inspiration for the headband -- the New World Indians. Here's a contemporaneous ad for hippie / boho headbands and similar accessories, consciously invoking the Indian roots of the style:
ReplyDeletehttps://www.reddit.com/r/vintageads/comments/iojehr/land_o_lakes_native_design_jewelry_and_leather_ad/
Girls -- what's stopping you from looking like this...? ^_^
That hippie fashion offer was operated by none other than Land O'Lakes, the dairy producer from Minnesota's Twin Cities, whose own logo has featured a Native American babe since its founding in the 1920s. Naturally, she wears a prominent decorated headband worn across the forehead.
https://1000logos.net/land-olakes-logo
One of my coolest chepie thrift store finds (for only 50 cents, back in the golden age of thrifting, during the 2010s), is a metal recipe box painted to look just like the original packaging, where it's a genre landscape painting with the woman in profile and off-center. I don't use it to hold recipe cards, but as its inspiration was intended -- to hold 4 sticks of butter in the fridge! It's so cool... ^_^
This also reveals the role of the Midwest as an incubator of American ethnogenesis, where Indians were still valued and sought as sources of inspiration for our new, New World culture. Architecturally, that would culminate in Frank Lloyd Wright's visit to California, where he adapted Central American step pyramids into the newly born American style.
The hippies and boho's of the '60s and '70s had seen that Land O'Lakes babe for their entire lives growing up, not to mention similar portrayals during the Western movie and TV series craze of the '50s and '60s.
It goes back right to our first encounters with the Indians themselves, who were always noted for wearing headbands of various types -- the leather / suede strap, some kind of colorful fabric wrapped around and tied or tucked in, the colorful beaded patterns from out West, the kinds with leather fringe and feathers attached to the ends, and everything in between. That was aside from the highly ornamental war bonnets or headdresses for high-ranking figures. Ordinary members of their culture wore headbands as well, just not the high-falutin' ones.
Paintings by American artists of Indians in their native costume showed them wearing headbands since the beginning. See several by George Catlin:
https://commons.wikimedia.org/wiki/File:George_Catlin_-_The_White_Cloud,_Head_Chief_of_the_Iowas_-_Google_Art_Project.jpg
https://commons.wikimedia.org/wiki/File:C%C3%B3l-lee,_a_Band_Chief_Cherokee_by_George_Catlin_(1985.66.285)_Smithsonian.jpg
https://sv.wikipedia.org/wiki/Fil:Ten-sq%C3%BAat-a-way.jpg
Although the medium shifted to photography, portrayals of Indians kept showing them wearing headbands into the 20th C, such as this series by Emma Freeman, shot during the 1910s. Among the tribes of Northern California, it was already headband city *well* in advance of the Summer of Love!
https://www.boredpanda.com/freeman-collection-native-americans
And that is the true origin of the headband as an distinctly American piece of headwear -- adopting something clearly New World from the native cultures, to fit in better with our new environment, and to make it clear we're no longer Europeans or Indo-Europeans. We're headband-wearers. ^_^
In 2004, there was a techno-dance song that sampled "San Francisco", "The Sound of San Francisco" by Austria's Global Deejays. As you can see from the music video, headbands became cemented in the public mind with that era.
ReplyDeletehttps://www.youtube.com/watch?v=8kjpPgmidUM
Although I never got to experience the Summer of Love, I did get to dance to that song during the last heyday of danceclub culture, during the 2000s and early 2010s, before it bit the dust over the course of the 2010s, and will never return ever again, despite Covid hysteria having been over for years now. You need to like people and trust people to dance around them, and Americans hate each other's guts and look at each other suspiciously, so no more dancing forever.
That song was an example of the 2000s techno songs that began with "Sandstorm" by Darude from 1999, where there was a periodic chorus or phase where the thump-thump, un-tsss beat stopped dead in its tracks, while a peaceful, atmospheric, often acoustic passage played to fill the void. I normally hate that kind of dance song -- periodically crippling the dancers, instead of just letting them do their thing.
But in this case, it was allowable since it gave it that wistful nostalgic feel -- simpler times, days gone by, before the arrival of un-tsss techno music. It was a nice emotional contrast, even if it slowed the movement down multiple times during the song. That's when you just start gliding by doing light hops on one foot several times, then switching feet. That allows you to keep in motion, but slowing it down somewhat, and that gliding motion makes it feel like a slow dance at the roller rink from the good ol' days... pretty fun, if done right.
"Feel Good Inc." by Gorillaz, from the same year, had the same dynamic. Gliding through those drum-free acoustic atmospheric passages, then getting more jumpy and aggressive when it switched back to the regular danceclub part.
I never liked "Sandstorm" for that reason -- you can't do much of anything corporeally during those silent atmospheric sections. It's just dead silent, then the regular chord progressions begin again, first quiet, then turning up the volume. So it's not like a different instrument or emotional tone or anything, just literally freezing the dancers dead in their tracks. I hated that -- let us dance, dammit! In the Global Deejays and Gorillaz songs, the acoustic guitar is used as a rhythm instrument, albeit just light strumming, but it was some kind of beat rather than no beat whatsoever in the Darude song.
Man, I really miss packed danceclubs no matter when you went out... I was acrobatically dancing non-stop for 4 hours straight, at least twice per weekend, sometimes 3 times if I didn't take a break on Friday (in between '80s night on Thursday, and the main Saturday night crowd 2 days later). In the summer of 2008, I was going out 4 nights a week, to an off-night Tuesday session, where there were only 10 people max, but still, getting some dancing time in on a week night.
ReplyDeleteElder Millennials remember that final dying gasp of danceclub culture -- they were most of the crowd I was surrounded by, in fact. But I don't think late Millennials, and certainly Zoomers, will ever know what it's like to pick any random club -- or establish a favorite / regular spot -- and every Saturday night, just go melt into the superorganic crowd, shutting off the cerebral part of your brain, and just being a totally corporeal animal, in a crowd though, not just going crazy by yourself.
That was one of the earliest victims of woketard iconoclasm during the 2010s -- dance clubs. If you don't fully trust everyone there, you will NOT be moving around, letting your guard down to cut loose, and in fact you probably won't even go there to begin with. Too "toxic" and "problematic".
I'll never forgive woketard Millennials for killing off dance clubs, not just for selfish reasons cuz I still want to go out dancing, but cuz it deprived the entire next generation of that rite of passage and weekend leisure activity.
Yeah, Zoomers don't trust each other, and are socially awkward -- cuz all the environments that are designed to take in socially and corporeally awkward people, and transform them into fun-lovers and mischief-makers, were systematically shut down by Millennials, who pulled the dance club ladder up after they aged out of the scene at 27, just in time to snag a fake email job that Zoomers won't even get as a consolation prize.
To foreshadow the new wave of headband styles after the hippie / boho stage, and sticking with the 2000s dance club revival of retro sounds, also from 2004, when '80s revival was in full swing, here's "Call on Me" by Eric Prydz.
ReplyDeletehttps://www.youtube.com/watch?v=qetW6R9Jxs4
It samples "Valerie" by Steve Winwood from 1982, and the music video leans heavily into the aerobics mania, and its obligatory fashion statement -- the sweatband, naturally worn across the forehead, not vertically. It just wouldn't be so quintessentially American otherwise!
Like "Sandstorm", I hated dancing to this one, cuz the quiet parts are just totally dead rhythmically. One of the quiet passages keeps the thump-thump bass on the on-beats, but the hi-hats on the off-beats are gone. And in the final quiet passage, which should build toward the climax, there's no rhythm whatsoever! Just that keyboard sequence playing over and over, which does not have enough of a percussive timbre to get your body moving to a steady beat. A pick striking guitar strings has some kind of impact to it, but synth warbles do not, they're too blobby and mushy. Might as well sit on the sidelines...
You see what the overall purpose was, though, right? To take the listener out of the crazy thumping modern-day craziness, hush them up, and tell them, "Shhh, you hear that? If you're quiet enough, you can still hear 'Valerie' faintly echo-ing across the ruins of Danceteria, like a visit from the Ghost of Disco Past..."
Yeah, I get it -- but don't totally strip out all the rhythm during those quiet passages!
Dance clubs used to be sites for male bonding, too (no homo). If you think you've seen guys hype each other up in a gym -- you've never been to a dance club. It gets a lot rowdier than a controlled environment like a gym.
ReplyDeleteI almost always went alone, since I was the life of the party, and could make friends with literally anyone and everyone, assuming they weren't creepy. And sure, college babes or high school girls with fake ID's would routinely ask for me to pull them up on stage to dance with me, surround me from all sides and grind away... the usual horniness from being 22.
But guys who are 22 also have crazy hormones, they just express it differently when you're a dude. They want to join the life of the party, too, not just be a wallflower. There was always a group of frat dudes or whoever they were, who joined me on the stage, hyping each other up.
And although my fondest memories naturally involve the dudettes, the dudes remembering me and giving me nicknames etc. brings back good memories of when guys were friends with each other as well. I'll never forget one came up to me asking if my name was "Dan" -- I said, no, why? "Oh, well some of the guys call you Dan, Dan, the '80s Man!" I literally LOL'd. Glad to be a positive role model for all the young dudes out there... ^_^
It was the same with tutoring and teaching -- the girls all had a huge crush on me, but the guys looked up to me, too. I was one of the only people they'd ever met who was a math wiz, but also verbally and socially intelligent enough to convey that to the math-averse. And guys and girls both appreciate someone who's really into what they do, while being socially atuned as well, not just endlessly rambling about their obsessions...
That's what this blog is for! ^_^
Also, here's a better link for the "Sound of San Francisco" video, that other one was 240 trash.
https://www.youtube.com/watch?v=WNnD2MWoPNI
A widespread source of Zoomer male indignation is older guys treating them like crap, ignoring them, competing with them sexually (stealing their would-be gf's with their ill-gotten corporate parasite income), jealously guarding all the cool stuff that some previous generation of guys passed along to them, and now is no longer being paid forward, etc.
ReplyDeleteMaybe it's cuz I never sold my soul to the corporate devil parasite, but I've never been that way. If anything, I do try to egg on -- er, encourage -- young guys, if they're feeling risk-averse. During Covid, I drove around blasting "Electric Love" when that was the popular TikTok trend, "letting my friend know I have a crush on them by kissing them to this song". School dances were dead, everything was dead -- the least I could do was be a music-mobile to egg them on about confessing their feelings to each other. Hehe. They all smiled, too, they knew it was good-natured prodding by the unc-mobile driver. ^_^
And I tutored and mentored them just as much as I did their female age-mates. It was just whoever's parents wanted to pay me, not me trying to take on only female students so I could flirt with them, and getting annoyed at male students asking me to tutor them. I had no problem with guys.
But then, that's why I'm not at the top of the self-destructing institutions, whose meritless elites are blowing them up from inside. I would have kept them standing and thriving, but it was the wrong time in the imperial lifespan, by the time I would be rising up them. By the 21st century, the only thing rising is garbage. If you're a psycho, parasite, scammer, nepo baby, etc., is all that counts, or protected ethnicity or generation (Boomer).
I don't mind being a cliff-dwelling sage in the ruins of the blogosphere, I can only follow my calling according to the realities of the time and place I'm living in. But it would have been better in the good ol' days. Evil talentless garbage did not parasitize its way to the top of every institution in society -- that's why it was the good ol' days. They had good elites, we have garbage elites.
But I was also never a "bros before hoes" guy. That's a loser Millennial trait. Gen X-ers like girls.
ReplyDeleteIf you're a Zoomer and use the term "foid", you are aping the worst part of your Millennial oppressors. They're using their fake email job income to steal your potential gf's, and then gassing you up about foid this and foid that, so that you stay functionally gay and pose minimal threat to their hunt for Zoomette pussy.
If the Millennial bros snagged a fake email job and/or a cheap mortgage prior to 2023 or whenever, they can afford to talk shit about foids -- they have the material resources to entice those foids to overlook their use of the word "foid". You, a Zoomer with a bullshit job or no job and never owning real estate -- canNOT afford to use the word "foid", or even think it.
And no, you can't use "foid" ironically. When you make it a habit, you become it. The more you talk like a de facto gay incel, the more you seal your fate to being exactly that.
And that's assuming the bros gassing you up about "foids" aren't just literally gay and trying to get you to hate girls so you'll desperately show your dick to them in a group chat -- 90% probability if it's an online social media platform. In their day, they learned the hard way that that type of guy was just a gay creep, but they won't pass that knowledge on to you, so you don't learn the hard way.
Gen X-ers never had to worry about this online crap, but we also didn't have male frenemies or gay creeps trying to manipulate us. And gerontocracy was bad, but not as vicious as now.
The point is: if you insist on being a girl-hater, as a Zoomer, you only get the shitty part of that deal -- being untouched and unnoticed forever. You don't get the cushy fake email job, or its ability to buy off some "foid" who will put up with your girl-hating cuz you're her sugar daddy. A Zoomer sugar daddy? They will never exist. For the love of God, quit thinking you're going to have it as good as previous generations, if you just act like them.
If all we need to own multiple real estate properties was "just act like Jeff Spicoli", all generations would be wealthy landowners, not only the Boomers.
The false consciousness across the generational divide is insane. Just cuz you performatively hate on girls like a Millennial, doesn't mean you'll get his fake email job or ability to buy off a sugar baby. Aside from being a traitor to your generation, you'll fail in your attempt to betray your generation! You can't leave your gen just by aping some other one. You can't be trans-generational any more than you can be transgender.
"Gen X-ers like girls."
DeleteEspecially girls like these. Fit yet feminine:
https://m.youtube.com/watch?v=KrdNDBlbkxU&list=RDKrdNDBlbkxU&start_radio=1&pp=ygUdc2F2ZWQgYnkgdGhlIGJlbGwgcGVyZm9ybWFuY2WgBwE%3D
Black Americans adopted the headband as well in the '60s and '70s, lest this be thought of as "white cultural appropriation of indigenous bla bla bla bla bla...." If you were American, you became a headband-wearer.
ReplyDeleteJimi Hendrix wore headbands regularly, so much so that they're just as valuable as memorabilia as guitar picks and other purely musical items.
https://www.tmz.com/2024/02/03/jimi-hendrix-headband-auction-bidding-guitar
Sly Stone was more of a big-ass-hat wearer, but he donned headbands as well.
https://en.wikipedia.org/wiki/File:Sly_Stone.jpg
George Clinton and Bootsie Collins:
https://gatheringofthevibes.com/wp-content/uploads/george-clinton-01-4-31.jpg
https://images.fineartamerica.com/images/artworkimages/medium/3/1-bootsy-collins-concert-photos.jpg
Even after the psychedelic funk heyday was over, La Toya Jackson kept the free-spirited flame alive in the '80s and '90s, making headbands her signature look, including Jazz Age revival types with studded jewels, but mainly the new type of thin twisted fabric in alternating color bands, for the New Wave / aerobics look.
https://www.gettyimages.com/detail/news-photo/american-singer-la-toya-jackson-1986-news-photo/89831324
Needless to say, the backwards hat was popular with black Americans in the '80s and '90s as well.
But headbands worn across the forehead remained staples as well, most notably worn by Tupac Shakur, connecting '90s gangsta rap with the '60s and '70s counter-culture era. Much like the ballcap, it was worn backwards, with the knot in the front.
https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcQmeAxGwlp1ii3FCXopOkBq86UYURAtInz6OlRcm-tnOg&s=10
None of these were carried over from Africa, they were invented or adopted from Native Americans during the early 20th C., and anyone who was American had to adopt them to some degree, maybe in a sub-cultural version, but still taking part in an all-American identity, as headband-wearers.
A wonderful gallery & history of 1920s headbands, from the heyday of blogs, and written during the Great Gatsby craze of the early 2010s. We'll see headbands of other types being revived around that time, in a bit.
ReplyDeletehttps://vintagedancer.com/1920s/six-1920s-headband-styles-you-can-wear-today
There were fabric types, not just metal & gemstone jewelry types. Mostly they went across the forehead. And they typically featured a prominent feather -- just like the Indian style.
It might have only been one instead of a full arrangement of feathers, and perhaps from an exotic species like ostrich or peacock, but the overall point was clear. Namely, we're no longer "Europeans living in America", we're Americans, and we're adapting the local culture, with our own little spin on it to make it our own, but even more so to distinguish ourselves from the Euros.
Euros at the time were also into flower garlands worn around the forehead, as part of Art Nouveau, but feathers? That's not part of the ancient heritage of Europe, like flowers and other plants worn as a crown.
True -- we're not in Europe anymore, we're adapting a different tradition of natural headbands -- from a fauna source, rather than flora. The Indians here used feathers rather than flowers -- well, then, feathers it is! (And flowers, too, but especially feathers.)
Style icon Debbie Harry from Blondie tying together, as it were, several strands of the headband tradition, during a performance on Midnight Special in 1979.
ReplyDeletehttps://www.youtube.com/watch?v=1SYKNq3oYUg
First, an homage to 1920s Jazz Age styles, by sporting a bob hairdo and wearing a tube-shaped dress (or rather, a romper).
Second, taking part in au courant disco trends, like the tube or column dresses by Yves Saint Laurent and Halston, as part of a Jazz Age revival, which began around '74 with Hollywood productions of The Great Gatsby and Murder on the Orient Express. The curvaceous Midcentury shapes were followed by straight-down silhouettes during the disco era, which revived the wild-and-crazy dance-fever nightlife of the Roaring Twenties. The headband was the perfect item to adapt to the neo-Jazz Age.
Third, paving the way for New Wave in the '80s, as Blondie were more on the rock side than R&B, funk, and other genres with big-band levels of instrumentation that characterized the disco sound. Her costume is fairly simplified, according to the stripped-down aesthetic of punk and New Wave, not as elaborate and embellished as other disco-era outfits.
Finally, preserving the hippie / boho spirit of the late '60s and early '70s, even though the counter-culture was just about dead and buried by '79. She emphasized this connection with the spoken-word message, delivered in the middle of the song, against nuclear power. (This was the same years as the Three Mile Island disaster.)
Not everyone in punk and New Wave was a hippie-hater. Joey Ramone was a hippie at heart, as was Debbie Harry, and even miss "Bad Reputation" herself, Joan Jett recorded a hard-edged yet plaintive and pining cover of "Crimson and Clover" in '81.
Speaking of which, Joan Jett with sweatband and dual wristbands... hachi machi!
https://i.pinimg.com/736x/fd/56/45/fd5645387825f28aae9571dd81398b55.jpg
You can't be a style icon in America without wearing a headband at some point. ^_^
Debbie Harry was a literal hippie! I never knew that before Blondie, she was in a psychedelic folk band, The Wind in the Willows -- how's *that* for a late '60s name? ^_^ She was born in 1945, before the Boomers, so she was the prime age for the hippie scene (Boomers only watched it on TV, being too young).
ReplyDeletehttps://www.artsy.net/artwork/steven-schlesinger-debbie-harry-and-wind-in-the-willows-1966
https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEirUtVw5Wbv4GtFBU99AYyK3r_Vu4E4EKI6aD49K98eJPNqEHJSzc9JONHDwKzpsujoXBGIfWbJixXUvj0nq6FwY6MBiDFWbVEQwphbg2HJFiecS4musRz4tjKcRl2vid5vg_4-F7ikrWo/s1600/2775c3f6e9e03def1deb656c4994c329.jpg
Well, her musical sensibilities may have changed, but she was clearly felt some other part of the Summer of Love was worth preserving, and tended the eternal flame of the headband, long after the counter-culture had bitten the dust.
We've already touched on the '80s, but here are some more from the headband's transition into a more athletic / sporty accessory. It may not have been as recognizably Native American anymore, but that's part of the inevitable evolution of it from a pure borrowing at the outset, to something that's uniquely its own thing in a mature stage.
ReplyDeleteThe most iconic of the aerobics-era headbands, the 1981 music video for "Physical" by Olivia Newton-John.
https://www.youtube.com/watch?v=vWz9VN40nCA
1982 ad for Adidas, "Love your body, flash Adidas". Showing a very free-spirited girl being carried on a guy's shoulders, while she covers his eyes, both of them being trust games and testifying to how cooperative the nation still was. They're in an outdoor setting, not some hi-tech gym, they're moving their bodies around and being active, not being decadent bedrotters or doomscrollers, and the whole message about "love your body" and baring it to others -- very hippy-dippy. ^_^
https://www.alamy.com/1982-adidas-ad-love-your-body-flash-adidas-image604845921.html
Yet again, Debbie Harry, in Videodrome from 1983, predicting a future fashion trend of 2 headbands worn at the same time. Also, the H.R. Giger cover of her 1981 solo album, KooKoo, channeling the Jazz Age circlet and slick hair.
https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcRzeQqMJ-jqjJpmntRzvOV2n0jaO2Li-4Xip4Woh3a1uQ
https://i0.wp.com/longlivevinyl.net/wp-content/uploads/2018/09/Screen-Shot-2018-09-17-at-11.42.52.png
Was this just a figment of the imagination of show business and advertisers? No, movies and TV and pop music are documentaries of their zeitgeist. Here's a news clip of the opening night of Return of the Jedi in San Diego, 1983, where one of the patrons is a real-life blonde babe with a headband prominently worn across her forehead. Some attendees showed their creative side by constructing a Star Wars costume to wear -- others just needed to throw on their headband to express their free-spirited, all-American identity. ^_^
https://www.youtube.com/watch?v=5uOt7pPluLg
American Japanophilia intersected with our adoption of the headband, whereby we came to view the Japanese headband ("hachimaki") as *the* definitive Far East martial arts headband. Carl Douglas popularized the headband in promo videos for his disco hit song "Kung Fu Fighting", but China and kung fu only go so far in America -- it really has to be about Japan and karate, for it to catch on here.
ReplyDeleteAnd so it did in 1984 with The Karate Kid, in which Mr. Miyagi gives Daniel an Okinawan-style hachimaki, white cloth with a blue flower design in the center -- tying back, in the American mind, to "wear some flowers in your hair". Back to nature, flower child, the wisdom of the Far East, Zen Buddhism... somehow it all tied together, and was distinct from the modern sweatband for the aerobicizers working out to their Buns of Steel videotape in the VCR.
https://www.slashfilm.com/img/gallery/the-karate-kid-headband-came-from-an-improvised-moment-during-an-audition-with-pat-morita/intro-1666817149.jpg
The love of headbands in Glorious Nippon actually goes back to their ethnogenesis a couple thousand years ago. Long before the red headband with long tails worn by Ryu from Street Fighter II, and centuries before the samurai used them inside their helmets, women in the Kofun era (mid-1st millennium AD) were sporting them:
https://note.com/art_history_ch/n/n0625a421866b?hl=en
And a 3rd C AD Chinese chronicle about the Japanese culture, Wajinden, noted that men wore cloth headbands that they tied around their heads in order to keep their hair back (which means they were worn horizontally across the forehead), and did not wear other hats or caps on top. So those would have looked exactly like the Southwest and Pacific Coast Native American styles that we gravitated to in America.
Japanese ultimately belongs to the Wa-Dene-Yeniseian family of languages, and it makes me wonder if their penchant for headbands as a cultural in-group marker is part of that shared heritage across the Pacific as well. At least on the northern side of the Western Pacific, i.e. Siberia, where the Wa people were originally from before settling into the Japanese islands.
Regardless of whether the Japanese and Pacific Coast Native American headband traditions share a common ancestor, or are independent developments, Americans treated them the same -- something that we could borrow and adapt to be distinctly non-European, New World, and Pacific.
Much as Mexicans feel awkward donning the geometric beaded headbands of their Aztec ancestors, Japanese people don't feel like wearing a hachimaki on a typical day being out and about. Maybe for a special occasion, where traditional clothing is called for, but not on a typical day.
They could totally adapt it to their current tastes, though, just like Americans did. It may not look exactly like the traditional style, but it will clearly resemble it, remind themselves of the traditional style. But it'll be more like the carefree Americans of the '70s and '80s wearing a cloth band across their forehead -- something to add a little pizzazz to the daily outfit. Something fun and cool.
Flower-children in the Land of the Rising Sun -- I can totally see it! ^_^
Japan has also preserved the backwards hat better than its American inventor has -- sadly typical of our collapsing shithole empire.
ReplyDeleteAnd Japan immediately adopted the backwards hat in the '90s, with Ness from Earthbound wearing one, to signal his independent / free-spirited / adventurous curiosity.
And Ash Ketchum from this minor series you might've heard of called POKEMON, who used his backwards hat as a modern-day karate hachimaki, to signal he was about to kick some serious ass when he turned his hat backwards.
The protag from Harvest Moon 64 wears a backwards hat, tying into the "back to nature" theme and hippie / Native American headbands. I was reminded of this from Nerissa's stream of the game last night -- and she even used a thumbnail image of herself wearing a backwards hat! ^_^
Speaking of vtubers, female characters, too, continue to sport backwards hats right up through the present vtuber phenomenon -- most notably on Subaru, to support her tomboy image.
But it's more than just a visual or fashion choice, it's meant to reassure the male audience that she isn't a typical girl who has a clique of girl friends who are going to subject you to a multi-round interview process, just to see if you're good enough to be her friend. She will interact with you one-on-one, and be her own judge of your character, since she's socially independent. The backwards hat is a signal of that social independence.
She has girl friends, of course, but a girl who wears a backwards hat is comfortable leaving her girl group for awhile, and just chatting and having fun with guys. That is very anxiety-reducing for guys who get nervous in the presence of too much estrogen.
In America over the past couple decades, a girl wearing a backwards hat, on the rare occasion that it does happen, is meant to support an image of her being haughty, superior, looking down on other girls, and being hostile and confrontational toward guys. It's anti-social, and is revealed by the smug, smirking, stone-faced facial expression. Back in the socially harmonious '90s, and still today in Glorious Nippon, a backwards hat girl wears a smile on her face, is excited to be around others, and is extraverted rather than stand-off-ish. She's just socially independent and free-spirited, not anti-social and condescending.
Leave it to Japan to act as the Noah's Ark for anything good that America ever created. But like I said, Japan has been wearing headbands for 2000 years or longer, so they probably viewed the American backwards hat as us being late to the party, but a welcome guest, hehe. And they eagerly adopted our version of the "band across the forehead". Especially after we adopted their karate hachimaki. One fun-loving borrowing deserves another. ^_^
Now for the various types of men's headbands that became popular in the wake of the '60s and early '70s counter-culture. First was the outlaw biker look, which grew right out of that environment.
ReplyDeleteFrom there, it joined the rock music uniform, usually red and usually taller than the hippie type, but still not covering the top of the head. Bruce Springsteen was an early adopter in '84, as part of his "Born in the USA" era look:
https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcS00nExIKtqG4I39rXV6p2Phwk0N7212MZMrkp097dr7g&s=10
Mark Knopfler from Dire Straits with a nod to the '80s sweatband:
https://www.gettyimages.com/detail/news-photo/mark-knopfler-performing-with-dire-straits-wearing-headband-news-photo/85840991
Steven Tyler from Aerosmith with a more wild-and-crazy color / pattern combo:
https://media.gettyimages.com/id/528656728/photo/steven-tyler-wearing-a-headband-and-parrot-pin.jpg?s=594x594&w=gi&k=20&c=cUj32jyhWHWEnI6bkQKjeYncOccyMdMnX8fYCqpJXAk=
Axl Rose from Guns N' Roses:
https://defendingaxlrose.com/wp-content/uploads/2011/12/axlroseaxlrose5.jpeg
Bon Jovi:
https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcQ_SD3JWKXfkf-uIqnQqRpjgRiAk5q7Aq44F70CexPfJA
Bret Michaels from Poison:
https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcTRJ5ExNZKnvkxIZCyVpjjVrE8awbdLmhAp6fnxSRks5A
And on and on and on...
Brand from The Goonies, as part of his '80s exercise look (gym shorts over sweatpants), did not opt for the standard sweatband, but mixed in a rolled red bandana that is more from the biker / rocker costume:
ReplyDeletehttps://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcSY8CXP3qmMIiZkYC6rnC6-LMxKTY40Rd_Ue-aIXd4hJw
The other major men's headband was the Vietnam-era soldier, which is anachronistic since Americans did not wear headbands in the actual war. This was a case of poetic license from Hollywood movies from the late '70s and '80s, when the Vietnam War became a recurring piece of subject matter.
ReplyDeleteThe idea was to mix in the counter-cultural uniform with the soldier's uniform, to suggest that at least some of the soldiers were not just unwilling but perhaps anti-war and drafted against their will. And to show without telling that the disciplinary system had so fractured apart, being on the other side of the world, that their superiors were incapable of enforcing strict uniform appearance, to the extent that they were openly wearing items from the enemy's uniform -- domestic enemies from the anti-war counter-culture.
The earliest example of the headband-in-Vietnam is one of the earliest examples of the new Vietnam genre as a whole, The Deer Hunter from 1978:
https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcTL6aKG0EI7niuDg6inKAnDu2k69ZTye5GCzidW3KwmpQ
https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcRoC3yzOrgJugIqF0sFS9VK8YcK2fQIvxLuYiO867JBPw
Apocalypse Now has a photojournalist wearing a red headband, but he's not a soldier, so I don't think that really counts.
The characters in Alien were not military, but once it's them vs. the alien, it does take on a military feel. Dennis wears a blue headband, not the usual red, perhaps to match the cool and subdued color palette of the ship overall.
https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcQAWsvuHO8lOTPW_1UTjCEVX7SkuHiayaYyK_zmgOHFBg
The sequel, Aliens, was a Vietnam movie set in outer space, and to show she's as badass as the men, Vasquez wears a rolled red bandana as though she were from a biker gang.
https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcReKgfA9UqJIq40oioVp6NGc_EtdgDIhQo54mi6jCLy_A
Platoon shows Charlie Sheen's character wearing the rolled red bandana, and Willem Dafoe's character wearing an olive drab one.
https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcSBOS6Tgw9_1b3fzGrlRvmfNe2EEL5uWb4K6UmWHBE-tA
https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcTYy5dxIGuLffQZwwMnnxwUKgpRGpM1-nSviWGW7At0RA
Full Metal Jacket aimed for verisimilitude, for the "reality is more surreal than fiction" effect. No headbands are worn in it.
Probably the most iconic of the '80s Vietnam-related movies is First Blood and its sequel, where Rambo's transformation from drifter to revenger involves a costume change where he fashions an impromptu headband with dangling tails, to suggest the karate hachimaki (not Vietnamese, but it's poetic license). In the first movie, it's made from olive drab canvas, which is what he had available, whereas in the sequel where he has time to prepare for a return to Vietnam, it's a blood-red color for dramatic emphasis.
https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcQoUEHiTy-krGIs07Le8AqXSGCR4fyMak7RQ3iKIembPQ
https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcTy07yh78ZQQTfAaCaAMvVNyRwxkySZp-jQD7N4VVKKSQ
Given the Far Eastern setting of the Vietnam War, the Japanese hachimaki from the martial arts trend crossed over into the soldier trend. Vietnamese don't actually wear headbands like that, but it's the Far East, so anything evocative of the broad Pacific coast of Asia will naturally call to mind others like it.
Can't forget Willie Nelson, who brought the headband into the Country / Western scene:
ReplyDeletehttps://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcSF47_Rhwqi956UzeSsA4dIHTzf12xJ-jy-aZzesEU4WQ
Epilogue: Scene kid headbands, Coachella LARP-ing in the 21st century. After the last bit of creative juice was squeezed out of American culture in the '90s, there really was nothing new in the headband's evolution.
ReplyDeleteThe main place they still exist are in the '60s / '70s hippie / boho LARP that Coachella and other such festivals rely on for their image, as though it's a neverending latter-day Woodstock. Well, the *real* Woodstock revival was in good ol' 1994, during the peak of the backwards hat craze and '60s / '70s revival in general.
Now they don't even bother with the music trying to sound like it's from that time -- and there were LOTS of different sounds to choose from. It's just the fact that Coachella is an outdoor music-ish festival, so why not LARP as bohos from the '70s?
I'm all for preserving styles after their creative evolution is over -- that's what makes them part of the standard folk costume. I just don't like that only a tiny fraction of the population does it, and only to score status-striving points in a zero-sum contest for clicks and clout, after posting pics to social media about going to Coachella.
Those festivals are lame, and have been lame for decades, so no, you don't get cool points just for going -- you have to contribute to constructing an atmosphere of coolness, of collective effervescence that will outlive its fleeting moment, just like Woodstock, which people still remember and talk about over half a century later.
There was a headband revival in the late 2000s among scene kids, although I hesitate to call it an innovation or new contribution. Regardless, at least they were keeping the free-spirited flame alive, unlike their dour mopey anti-social foils, the emos.
ReplyDeleteScene kids were so bright and colorful, extraverted, smiley, giggly, always hanging out and being spontaneous in public spaces, and were the last sub-culture to valorize and embody young innocent heterosexual love. Not hooking up, not faux sexual edginess, no trannies, and minimal gay presence (they joined the emos).
Much like the '60s / '70s counter-culture, they were not valorizing hedonism or promiscuity, but back-to-nature Edenic RETVRN -- Adam & Eve, with bangs over their eyes. The Blue Lagoon -- in suburbia. That was it!
No one truly misses the emos, other than emos themselves, but all sorts of people who were not part of the scene scene, will always wax nostalgic about the scene kids. They were adorable and likeable. ^_^
So, sure enough, whether intending to or not, they revived the headband for one last go-round in au courant American culture. The item itself was no different, just a standard plastic headband that normally rose vertically over the top of the head, to keep hair back. They simply lowered its angle so it ran across the forehead, like any honest-to-goodness American headwear should! It was thin, like the '80s version, since this was also peak '80s revival, and did not have any ornamentation on it. But the overall point was clear -- "we're part of that whole headband tradition... y'know?"
As one of the last unique looks in American cultural history, scene still has its adherents in the 2020s. What else are kids going to do -- something new? Heh, that requires cooperation, and everyone hates everyone else's guts. But you can adopt what previous cooperative cultures created, so scene it is! And far from being a marginal item, the headband is something today's scene devotees still clearly remember as part of the overall look:
https://www.reddit.com/r/scene/comments/1j6iz0c/can_we_bring_back_scene_kids_wearing
Once upon a time, before social media, a large chunk of Web 2.0 was simply re-posting pictures of scene girls and pining away for their Manic Pixie Dream Girl cuteness and spontaneity. I won't add a sarcastic Millennial note about "whether that was real or only imagined" -- the scene girls really were fun-loving extraverted types. Just cuz they wore skinny jeans like emos, didn't make them resentful and anti-social. They were the last reincarnation of the hippies...
https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcQXoQ_DZvk0R-J_P8RfHFH7fYPilfS-FzYtLVtYTcuifg
The early 2010s also saw a Native American revival for the first time since the Southwest craze of the '90s. And feathered headbands, and humungo war bonnets, were part of that image -- although hardly anyone wore them on their head, except for the occasional Coachella appearance. They at least showed their likeness on t-shirts, especially if there was an Aztec print involved, which was another part of that Native American trend.
ReplyDeleteAs mentioned before, there was a Jazz Age revival of that kind of headband, with the release of The Great Gatsby in movie theaters.
And '80s revival was still going strong, so the American Apparel uniform might have still had a sweatband in it.
All of these, including the Native American headdress with a full row of large feathers, are shown in the music video for the indie dance-rock hit "What You Know" by Two Door Dance Club (from 2010 / '11).
https://www.youtube.com/watch?v=YXwYJyrKK5A
As the woketard iconoclasm of the 2010s erupted soon afterward, all of these images were not only halted, they were singled out for destruction and shaming of anyone taking part in their American heritage. Let alone enjoying their heritage -- there was such a puritanical witch-hunting drive behind woketard-ism, that the sinners were not merely sinners, they were gleeful sinners and therefore doubly sinful and requiring double the punishment.
Then when American cultural production ground to a halt due to the BLM / Antifa riots and Covid hysteria, it never recovered, as now everyone hates everyone else's guts. The woketard Eye of Sauron is no longer necessary, since nobody will participate in a group with anyone else for any reason whatsoever -- all the fun has been stamped out of American society, that there is no need to police the denizens of the panopticon anymore.
Not for cultural reasons, anyway. Now the iconoclasts are right-wingers searching for any hint of criticism of Israel or Jews, as the Pentagon and Israel and 90% of Jewish opinion have plunged America into its final death-spiral through a pre-ordained catastrophic failure in the Persian Gulf (and the Houthis, and the Taliban, and Hezbollah, and....).
Ten years after 2016, it's hilarious to look back and see what the witch-hunters were so obsessed with destroying -- people having fun and taking part in their own cultural heritage? Talk about "First World problems"... now we've plunged off the cliff, rapidly devolved into a Third World shithole, and suffer from Third World problems, like censorship over criticism of the elites' compounding and cascading failures.
I think people will look back at the late 2000s and early 2010s as the American Belle Epoque -- decadent, widening inequality, godless, etc., but at least not plunging right over the cliff *just yet*. They're quickly approaching the precipice, but have not yet flung themselves over it, which took place during the 1910s, not the Gay Nineties or the 1900s.
And much as people still LARP as Edwardians from the Belle Epoque, people will still LARP as scene kids from 2008, a century into the future. What else is there going to be in the meantime?
*Two Door Cinema Club
ReplyDeleteIs the Japanese headband from Ainu / Emishi origins? I looked through the other Siberian cultures, and those near the Pacific coast wear headbands, regardless of their present linguistic group -- Evenki (Tungusic, formerly Eskaleut), Chukchi and Itelmen and Koryak, not to mention Eskaleut on the New World side of the Pacific. However, their style of headband features elaborate beadwork, often with dangling pieces.
ReplyDeleteThe hachimaki from Japan does not feature any beadwork or other applied ornaments like shells, fur, feathers, etc. And it's made from cloth, not a leather strap. But it does feature some kind of embroidery for ornamentation.
Whaddaya know? That's just like the traditional Ainu headband, the matanpushi.
https://ds-archive.haverford.edu/japanesemodernism/exhibits/show/transition-of-ainu-culture-thr/introduction
https://japanfans.nl/wp-content/uploads/2021/04/ainu4-1024x768.jpg
The surviving Yeniseian speakers like the Ket don't have headbands that wrap around the forehead and temples only. Their headwear is more like a scarf that covers most of the head, including the top. The Pacific Rim headbands leave the top of the hair and head exposed, although in Arctic regions, there may be a separate hood on top of the headband.
So I'm not sure if the Wa people brought the hachimaki with them into Japan. It could very well be an adoption of a local sartorial shibboleth -- exactly like the American adoption of the Native American headband. ^_^
The Emishi were not genocided, they were gradually incorporated into the Yamato culture that fused newly arriving Yayoi and native Jomon cultures.
Some of their religious beliefs were incorporated into the Japanese religious system -- like Sanzu no Kawa, the River of Three Paths, which is present in Ainu folk religion, and belongs to the broader Uralic religious sytem about the afterlife and preparation for it, a kind of karma system.
Other parts of Emishi religion were not adopted into Yamato religion, like the bear cult.
Some elements of Emishi clothing and grooming were left out of the Yamato style, such as heavy untrimmed beards.
But other Emishi elements were adopted into the Yamato, like the embroidered cloth headband.
Neat!
Koreans don't have a traditional headband like the Japanese or Ainu, which confirms that the hachimaki is native to the Japanese islands, originating with the Jomon / Emishi / Ainu.
ReplyDeleteMaybe the Koreans, or pre-Koreans, used to have a headband of some kind, since they're from the Pacific coast. But if they did, they lost it, probably due to the Koreans imitating China so heavily, whereas Japan preserved much of its non-Chinese culture.
Koreans still have shamans in some form, but they don't have an elaborate hierarchical / institutional native religion like Shinto in Glorious Nippon. They gave it up for Buddhism, imported from China. Japan also imported Buddhism, but did not jettison its pre-Buddhist religion at the same time, they syncretized them or kept them parallel.
Is Korean a creole of Wa / Japonic and some other Paleosiberian language? I promised I would not delve into Korean too deeply until South Korean woketards surrender their culture-destroying crusade against Japanese pop culture (mostly involving the Rising Sun Flag). And I am keeping to that.
ReplyDeleteBut I got a little curious about their word for "sun", since that word unlocked two Paleosiberian mysteries for me -- the recent one about Tungusic being heavily Eskaleut, and the earlier one about Japonic being Dene-Yeniseian.
Well well well... Korean appears to have the same word for "sun" as Japonic, cognate with Na-Dene and Yeniseian. In Modern Korean it's "hae", and was first recorded in Old Korean as "hoy":
https://en.wiktionary.org/wiki/%ED%95%B4#Korean
This is clearly cognate with the Proto-Koreanic word for "day", which is "hol".
https://en.wiktionary.org/wiki/Reconstruction:Proto-Koreanic/hol
The first 2 segments are identical, and the final consonants are very similar -- voiced palatal continuants, with one being more of a semi-vowel ("y") and the other being a bit more consonantal ("l"). Very little is known about Old Korean, and almost nothing about Proto-Koreanic, since they're a young "family" with only 2 branches, one of which (Jeju) has hardly any speakers. And their writing system during Old Korean obscured most of the word, except for the final section.
So for all we know, in OK the final "l" was closer to a palatalized "l", "lʲ". If so, then late OK and MK final "y" is just a simplification of "lʲ", by making the palatalization feature primary rather than secondary.
Or maybe there is some subtle but meaningful suffix "l" that derives "hol" = "day" from "hoy" = "sun". They're still clearly cognates. That's not too surprising, since many Siberian languages have the same, or mostly the same, word for both "sun" and "day". And for "sun god(dess)". In Japanese, the central sun goddess has a simple name "sun deity" -- "Pi no Kami" in OJ (better known as "Amaterasu"). Likewise in Korean mythology, the founder of the ancient ruling dynasty is the sun god with the basic word for "sun" in his name -- "Hae Mo-su", meaning "Sun Ruler / Lord".
So where does Korean "hoy" come from? I instantly recognized it as descended from Dene-Yeniseian "xʷaj" (P-ND and P-Y), the same source of Proto-Japonic "pi".
ReplyDeletehttps://en.wiktionary.org/wiki/Reconstruction:Proto-Na-Dene/x%CA%B7aj
https://en.wiktionary.org/wiki/Reconstruction:Proto-Yeniseian/x%CA%B7aj
P-J resolved the difficult consonant "xʷ" by making the labialization primary rather than secondary, and preserving voicelessness, by making it "p". This "xʷ" -> "p" sound change typifies the transition from Wa (the direct sister to Na-Dene and Yeniseian) to Japonic (belonging to southern Korea and Japan, as they incorporated former Emishi / Ainu speakers). They vocalized coda glide "j" -> "i", since coda consonants are banned in P-J, leaving "pai". P-J does not allow vowel sequences, and typically they resolve that by keeping the 2nd and deleting the 1st, which they did in this case, yielding "pi", QED.
Well, Korean didn't use the same strategies to resolve the Wa-Dene-Yeniseian word, cuz Korean must have had a non-Emishi/Ainu element to it, unlike Japonic. What that other Paleosiberian language is, IDK -- by geography, it was probably Nivkh, but maybe Eskaleut, IDK right now. The point is, Korean had a different set of phonotactic constraints and a different phoneme inventory, compared to Japonic. So they didn't resolve "xʷaj" in the same way -- but it was still beyond their ordinary phonology, so they still had to resolve it somehow.
Korean does not ban coda consonants, so final "j" was not a problem. This is the same sound written "y" in Korean. They left that as is.
Korean does not have labialized velars, and neither does Nivkh or Eskaleut, even though both of them have velar fricatives. So "xʷ" was a problem, but only the labialization feature, not the velar fricative.
Japonic does not have velar fricatives, probably owing to Ainu being Uralic, which mostly lacks velar fricatives, and certainly not in initial position. So when the Wa incorporated Emishi / Ainu speakers, they had to get rid of velar fricatives.
But whatever Paleosiberian language that Korean speakers used to have, aside from Wa-Dene-Yeniseian, must have had velar fricatives, like Nivkh or Eskaleut, since they did not resolve "xʷ" as "p" -- they used "h" instead.
Again, very little is known about OK or P-K, especially for initial position of the word. But this "h" is not a weakening of earlier "p", as Middle / Modern Japanese "h" is a weakening of P-J and OJ "p". OK has plenty of words beginning with "p", and "h" as well, meaning they're distinct. Korean also has lots of words beginning with "k", so "h" is not a weakening of "k" either.
I conclude that the OK or P-K "h" was a voicless velar fricative, "x", and reflects the phoneme inventory of the precursor to Korean that was spoken by future Korean speakers. Such as Nivkh or Eskaleut. So, in resolving "xʷ", they kept the primary consonant "x", which is transcribed as "h" in Korean studies.
Well, they still don't have "xʷ", though -- so what to do with that secondary labialization feature? I think they off-loaded that onto the following vowel, which was originally "a". The transcription "o" in Korean studies is not the usual "o" as in English "boat" -- it's "ʌ" as in English "but". It's mid-to-low in height, back, and unround.
ReplyDeleteSo, clearly they did not treat "ʷ" as a rounding feature, since "ʌ" is unround. However, they could still treat it as a "velar location" feature, since "w" has not only labial place of articulation but velar co-articulation. In other words, they just took original "a" and moved it to the back position. The only low-ish, back, unround vowel in Korean is "ʌ" -- so that's how they preserved "ʷ", rather than just deleting it and losing all information about it, as if they had made it "hay".
Notice also, that this avoids a vowel merger, as if they had altered "a" -> "wo" or "wu", i.e. the back ROUND vowels. Korean has a ton of vowels, which is unusual for its geography, and is one crucial way in which it's different from Japonic.
But as I detailed in my exploration of Japonic being Wa-Dene-Yeniseian, Japonic dealt with the compression of its consonant inventory, from several dozens down to just 9, by opening up a new word-final vowel slot, onto which the lost information about the original 2nd C could be preserved.
Well, perhaps Korean took a similar approach, but preserved lost info on the sole nuclear vowel. However, to avoid an insane vowel merger in that sole vowel slot, they doubled the number of vowels in their inventory. The original vowels indicated no lost info being preserved on them. The new exotic vowels signaled, "some lost info is being preserved here".
In the case of "o" / "ʌ", perhaps it meant the preceding consonant used to be labialized, and the lost labialization feature has been preserved as a back feature instead. And since "ʌ" is unround, the original vowel must have been more central or front, while also being low-ish, in order for a back feature to shift it to "ʌ". That is, it used to be "a", and preceded by a labialized consonant.
I have a few more Wa-Dene-Yeniseian etymologies for P-K / OK words, but the word for "sun; day" is so important, I wanted to start with that one. It's always the one that gets the ball rolling...
I hope I can find a suitable North Korean babe to dedicate this little piece of scholarship to, since South Korea is still under my boycott. But the Korean language is the same in the North, so it's not so irrevocably poisoned by South Korean woketards. That's how I will rationalize this. ^_^
Korean word for "world; era" from Wa-Dene-Yeniseian origin? There's an Old Korean word "NWUli" = "world", and later took on the sense of "era, time, generation" in Middle Korean. The capital "NWU" shows we don't know exactly what it was pronounced like, but it's that sound in MK, but the lower-case "li" means we do know it was that sound -- or "ri", but still, that ending is known.
ReplyDeletehttps://en.wiktionary.org/wiki/%E4%B8%96%E7%90%86
I'm not positive of this, especially since I can't explain what the "li / ri" final part means, but I would assume is some kind of noun-to-noun suffix that extends the meaning in a figurative way, not so literally.
There's a Proto-Na-Dene word "nu" = "island", which is "noow" in Tlingit (also means "hill") and "nuˑ" in Proto-Athabaskan.
https://en.wiktionary.org/wiki/Reconstruction:Proto-Athabaskan/nu%CB%91
In the first few centuries AD, Chinese records refer to Japan as "Wa Nu", where "Wa" is the endonym of the Eastern Steppe people who seeded the Yamato culture in Japan. Sometimes the Chinese referred to "the Wa country of Nu". In either case, they seem to be referring to Kyushu, which is an island, and it would make total sense to refer to Kyushu as "Wa Island".
Although the word "nu" does not survive as a standalone word for "island" in P-J or later, it did survive inside of a compound. Namely "nusi" = "master", where "nu" refers to the domain over which someone rules, which in Japan is an island.
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/nusi
I already went over this discovery during my journey into the Dene-Yeniseian origins of Japonic / Wa.
It does not exist in the Yeniseian branch of the mega-family, since Yeniseian bans initial nasals, but Na-Dene allows them, so does Wa / Japonic, and so does Korean. Yeniseian is the odd one out in this respect.
The semantic connection to Korean "nuli" or "nuri" is clear, where "island" is a kind of self-contained world or land, with boundaries that define it. That also applies to "age, era, generation", which is the temporal equivalent -- it has certain boundaries that define it as an island within the overall flow of time.
It's not "world" in the sense of earth, soil, ground, etc. But in the sense of a large geological landmass that people inhabit, that has an ecosystem that sustains them, that has clear boundaries of where their world begins and other worlds end.
Again, the "li / ri" element is some kind of generic nominalizer that gives it a more figurative sense, not a literal landmass in the middle of a large water body.
Korean words for "1" are cognate with Yeniseian word for "2" and maybe "1". Numerals are not very diagnostic of shared ancestry, since they are ditched and new ones adopted, without changing languages. They are just a kind of technology. Japanese ditched its native system (aside from 4 and 7) and adopted Chinese-inspired words instead. Nevertheless, something caught my eye.
ReplyDeleteIn Middle Korean, "1" is "hona(h)" (where "o" is unround back low-ish "ʌ").
https://en.wiktionary.org/wiki/%E1%84%92%E1%86%9E%EB%82%B3#Middle_Korean
In Proto-Yeniseian, "2" is "xʷin", several of whose descendants have a vowel that is unround, high, and central. Vajda compares it to Proto-Athabaskan "χonʔ" = "half (of something)".
https://en.wiktionary.org/wiki/Reconstruction:Proto-Yeniseian/x%CA%B7in
Phonetically, I invoke the sound change for "sun; day", where Wa-Dene-Yeniseian "xʷ" -> "ho". That is, the velar fricative remains, and the secondary labialization becomes a back vowel afterward. But not one of the old back vowels, which were round -- one of the new back vowels, which are meant to preserve lost info, namely that the preceding consonant used to be labialized.
In this case, the original vowel "i" loses info about its height, since "ʌ" is mid-to-low rather than high, but it retains the info that it is unround.
So, based on this case and the case of "sun; day", the "ʷ" -> "o" change replaces the nuclear vowel, rather than shifting the original vowel backward at the same height. I assumed it did the latter, based on the original vowel being low "a" in "sun; day" = "xʷaj". But here, the original vowel is high "i", yet it is also transformed into mid-low "ʌ", not a high back unround vowel like "ɯ".
Eerily, the final "a" in the Korean word "hona" also appears in several branches of Yeniseian (Arin, Kott, and Pumpokol), where it is the "pseudo-copulative numeral suffix" (...duh). IDK if it serves the same function in Korean, but it shares this final "a" with several Yeniseian languages as well.
Semantically, though, aren't these words incompatible? The Korean word means "1" and the Yeniseian word means "2". Doesn't math class teach that there is no overlap between these two sets? Well, maybe, but primitive people aren't nerds, let alone quant nerds. They name numbers after concrete things, like fingers or something else.
ReplyDeleteFor example, suppose two languages are definitely related, and have cognate names for "pointer finger". But suppose that, since number-counting systems are a kind of tech and somewhat arbitrarily constructed, one language begins counting on the pointer finger, while the other language begins counting on the thumb. Then their cognate word for "pointer finger" will refer to "1" in the first language, and "2" in the second language. Just one way that can happen, despite the abstract numbers 1 and 2 not overlapping at all.
Another way could be at work here, given the connection to a Na-Dene word for "half", which involves the concept of "part vs. whole", or fractions, or dividing / cutting / separating the whole into parts. The simplest way to divide, cut, or separate a whole is to make just 1 cut, leaving 2 parts, which are called halves.
But then, when looking at the outcome, you still see the two halves as belonging to a single whole, just being separated or cut apart from each other. Still, they form a matching pair that fits together to reconstruct the single whole.
In Latin, it's suspected that the words for "part", "portion", and "pair, mate" are the same. And yet, "part" and "portion" are single fragments, while matching or joining them together to form a "pair" leads to the number 2, not 1. But you can see how they're all related.
https://en.wiktionary.org/wiki/pars#Latin
I think something similar is going on with Korean and Yeniseian numerals, where the words are cognates. But it refers to "1" in Korean, in the sense of "1 part of a whole" or "1 part of a pair". And it refers to "2" in Yeniseian, in the sense of "pair of 2 parts that fit together to reconstruct the whole". Just like Latin cognates for "part" and "pair", where "part" wants to refer to "1" and "pair" wants to refer to "2".
There's another word for "1" in Old Korean, which is "HAton(h)" or "HOton(h)".
ReplyDeletehttps://en.wiktionary.org/wiki/%E4%B8%80%E7%AD%89#Old_Korean
The confusion comes from "HAton" having such a poor phonetic relation to Middle Korean "hona", so the OK form is assumed to be "HOton" to make that connection easier.
But maybe that is uncalled for, and there are two distinct roots relating to "1" in the history of Korean. In a 12th-C Korean word-list, this OK word is clearly phonetically listed as "haton", *and* it is given no Middle Korean reflex. So it's from a different root from "hona".
We've already seen a Yeniseian cognate for "hona" -- what about for "haton"? Yet again IDK what the final element "ton" means, but I do know what the initial "HA" element means -- it's cognate with the Yeniseian word for "1", which is "qus" for inanimate things and "qog" for animate things, both of which are presumed to stem from the P-Y distal demonstrative pronoun "qa", i.e. "that".
https://en.wiktionary.org/wiki/Reconstruction:Proto-Yeniseian/qus
Why the distal pronoun, "that"? Probably cuz you use your pointer finger to indicate distal objects, since there's so much space that's far away. For proximal objects, you don't need to point with your finger, since the listener can figure out which one you mean, given the small range of space it's confined to.
So the semantic development is "distal pronoun" -> "finger used to point at distal objects, while using distal pronoun" -> "number falling on this finger when counting", which is 1 if you start with the pointer finger, as they do.
By the time of P-Y, this relationship to the distal pronoun had become obscured and fossilized, but that is the assumed origin of the P-Y words for "1", which both begin with "q" and have different suffixes for animate vs. inanimate.
If Korean has a Wa-Dene-Yeniseian origin, then it used to have uvular consonants, like "q". Japonic largely deleted these, and are the source of so many vowel-initial words in P-J. But perhaps Korean shifted them forward to velar location. In this case, earlier "qa" -> "xa", transcribed as "ha". In Japonic, the distal pronoun was originally "ka", which is just Yeniseian "qa" shifted forward to velar location.
ReplyDeleteOne good sign for OK "ha" vs. Yeniseian "qa" is the vowel is the same, one of the old vowels, "a" -- not an exotic new vowel like "ʌ". That means the preceding consonant was *not* labialized -- and sure enough, in Yeniseian it's "q" rather than "qʷ" or any other labialized uvular. Indeed, P-Y does not have any labialized uvulars, only labialized velars.
Both "q" and "x" are voiceless.
The only difference is "q" is a stop and "x" is a fricative. Why not map "q" -> "k" and be even more faithful? IDK, maybe "x" became a catch-all container for the exotic back consonants that were lost between Wa and Korean, regardless of whether they were stops or fricatives.
Japonic did not have this option for dealing with uvulars, since they had not velar fricatives, since Ainu did not, being Uralic. So they either moved them to velar stops or deleted them altogether, where the initial vowel in a language that wants words to begin with a consonant, was a clear signal that something used to be there. And what other class of consonants were deleted, aside from the exotic uvulars? None, they must have begun with uvulars.
Whatever the L2 / filtering language was in the Korean creolization, it did have velar fricatives, so perhaps Nivkh or Eskaleut. Therefore it did have "x" as a catch-all container for lost back consonants.
To be clear, "ha" was never a distal pronoun in Korean proper -- it would have been a fossilized reflex of "qa" from Wa, surviving in a compound "haton".
To speculate about that "ton" element, in Old Korean it was used to mean "when" retrospectively. So perhaps the logic is that "ha" reflected the Wa distal pronoun, and by extension, the finger used when pointing out far-away objects, i.e. the pointer finger. Then "ton" means "when" in the context of counting on the fingers. Putting that together, "haton" means "the number WHEN it's the DISTAL PRONOUN finger", i.e. the number "when" it's the pointer finger, i.e. "1".
If so, then both Korean and Yeniseian have a word for "1" that derives from a cognate "qa" / "ha", the distal demonstrative pronoun, linked to the pointer finger. They put different suffixes onto it, and in both cases this morphology is obscure and fossilized rather than transparent in the proto-languages. But still, they do seem to share an ancestor.
And it's a different shared root than the previous one about "part vs. whole". I deny the link between OK "HAton" and MK "hona". Rather, they are two distinct roots -- "HAton" referring to the pointer finger, and "hona" referring to "1 part of a pair".
So this is a case where one way of expressing a number replaced an earlier and different way of expressing that same number. The two Korean words are not cognate, but each one does have a cognate in Yeniseian -- in one case, the Yeniseian word also means "1" (the "distal pronoun" form), while in the other case, the Yeniseian word means "2" or "pair of 2 parts" (the "part vs. whole" form).
To summarize the overall view of the relationship between Korean and Japonic, my hunch is that both are descended from Wa -- that is, the precursor to Japonic, before heavy incorporation of Emishi / Ainu speakers in the Japanese islands. Wa is the Eastern Steppe continental language that is sister to Yeniseian and Na-Dene.
ReplyDeleteThat was the L1 target, in a polyglot social / cultural / political environment.
In Japan, the L2 was Emishi / Ainu. The result of "speaking Wa with an Emishi accent" is Japonic.
In Korea, the L2 was Nivkh or Eskaleut or something like that, not Emishi or Ainu. The result of "speaking Wa with a Nivkh / Eskaleut accent" is Korean.
Since both Japonic and Korean come from attempts to speak a single language, Wa, they are very similar to each other. But because they reflect two different sources of L2 speakers -- Emishi / Ainu vs. Nivkh or Eskaleut -- they also have their differences.
It's possible that as Korean became its own language, it later absorbed former Japonic speakers who were left in Korea. According to Janhunen, those were in SW Korea, Baekje, whereas Korean proper comes from Silla in SE Korea.
But during the formation of Korean, I still think they had an L1 target of Wa, which was spoken in the Korean peninsula before the arrival of Nivkh / Eskaleut / Whoever else speakers during the nomadic Dark Ages. Probably after 300 AD.
It doesn't look like they were trying to speak Japonic, but Wa. Perfect example, "sun" is "pi" in P-J -- very easy to speak in Nivkh, Eskaleut, etc., as just "pi". And yet, Proto-Korean used "hoy", which is much closer to the P-Yeniseian and P-Na-Dene word for "sun", and therefore the Wa word as well, from which P-J "pi" descends, "xʷaj".
In fact, I'm not sure the language of SW Korea was Japonic proper, i.e. with the heavy Emishi / Ainu influences. Maybe it was still Wa in Korea, unless there were also Emishi / Ainu speakers in SW Korea, and that's where Japonic began, as "Wa filtered through an Emishi / Ainu accent".
But outside of SW Korea, it was not necessarily Japonic, possibly Wa instead. And when the migrants from Manchuria or the Amur region entered Korea, they tried learning Wa, not Japonic, when they first arrived in the northern and central parts of Korea. By the time some of them settled into SE Korea, they formed a new creole, "Wa filtered through a Nivkh / Eskaleut / Etc. accent", which became Korean.
The non-Wa part of Korean looks like Eskaleut, not Nivkh. Well, tying into the previous journey -- which is not over yet, just exploring some side quests, like Korean -- Tungusic has a mainly Eskaleut origin. So perhaps it was Tungusic, or the stage when Eskaleut was creolizing into Tungusic. But at any rate, *not* Nivkh, and seemingly not any of the other usual suspects for Siberia. So the best bet is Eskaleut.
ReplyDeleteNow for a look at the phoneme inventory of Proto-Koreanic, to see what it looks most like.
https://en.wikipedia.org/wiki/Koreanic_languages#Phonology
First, it's heavily compressed, with no voiced obstruents, much like Proto-Japonic. However, I don't think this is due to the two inventories being shared. Rather, it's that any 2 heavily compressed phoneme inventories will look like each other, by convergent evolution -- namely, heavily compressed inventories will weed out all sorts of marked phonemes, and retain only the most unmarked and basic phonemes in the world. And they won't have any kind of contrasting series, just voiceless stops.
And in fact, P-J and P-K don't share very much, when you account for them both being heavily compressed. P-J has 9 consonants, P-K has 12 (the chart above leaves out "l" for some reason). And it's not like one is a subset of the other. P-J has 1 consonant that P-K doesn't -- "w" -- and P-K has 4 consonants that P-J doesn't -- "c", "ŋ", "h" (velar fricative), and "l". Of P-K's 12 consonants, 3 are fairly marked -- "c", "ŋ", and "h".
They were both targeting Wa as the L1, but they came from different L2's, which explains why their phoneme inventories look fairly different, apart from both being compressed.
So whose inventory does P-K's look like, then? Not Nivkh, a common suggestiong since it's in the region, but Eskaleut, which is rarely considered since people think it's only an Arctic family, but my discovery of Proto-Tungusic being mainly Eskaleut places Eskaleut way down into Manchuria in antiquity and perhaps the early Medieval era.
Proto-Nivkh has "w", but Eskaleut does not -- and neither does P-K. Eskaleut has a voiced labial fricative, written "v", but it doesn't have a labial approximant like "w". P-J has "w", so does Proto-Yeniseian. The Altaic foundation, derived from late Samoyedic, does not have "w", which hardened into "b" -- but P-K doesn't have "b" either.
P-K has nasals "m", "n" and "ŋ", which perfectly matches the nasals from Eskaleut, but not Proto-Tungusic or Proto-Turkic or Proto-Samoyedic or Nivkh, which have palatal "ń" as well, whereas both Korean and Eskaleut lack "ń". Also, Nivkh and Tungusic can begin words with "ŋ", but Korean and Eskaleut cannot. Yeniseian has labialized "ŋʷ", but none of the others do, including Korean.
The palatal "c" in P-K is compatible with Eskaleut, although with others as well.
Likewise the "h" in P-K, being the same as "ɣ" from Eskaleut, but devoiced since P-K went through a compression that removed all voiced obstruents.
P-K has "l", unlike P-J or P-Y (which has lateral fricatives, but not the approximant). Eskaleut has "l" as well.
Eskaleut does not have "r", only "l", and in that sense is like P-J and Middle Korean, where there's only 1 liquid, not 2. But Old Korean and P-K have a second "liquid", written "r" in Korean studies, and distinct from "l". This is one potential difference between Korean and Eskaleut.
ReplyDeleteBut the evidence that the OK consonant was specifically "r" is not very strong. It's based on the Middle Chinese character that Korean chose to represent its own "r", but this word in Middle Chinese was "it", ending in "t", not "r".
So then the theory goes that in Northeast China, the dialect had weakened "t" to an "r", so it really was "r" after all that the Koreans heard. And when Old Tibetan recorded Middle Chinese words ending in "t", it also used the Tibetan letter for "r", not "t", so it was "r" after all.
Since it looks like Korean came from a partly Eskaleut background, I challenge that interpretation, and say the OK "r" was more like the voiced interdental fricative from Eskaleut, "ð", although perhaps it had been devoiced as part of the ban on voiced obstruents in P-K, i.e. "θ". It is very common for "ð" to become "r" over a language's history, otherwise hardening to "d", but "ð" is an unstable consonant in most languages.
Standard Middle Chinese did not have interdental fricatives, and neither did Old Tibetan. So if the NE Chinese dialects had begun to weaken "t", it's possible it softened to "θ", not "r", at least that's the natural first stage, before perhaps a second stage where it softens further to "r". If so, Chinese and Tibetan orthography would have no way to represent the stage where it's an interdental fricative, since their standard dialects do not have such consonants.
Perhaps it was just such an interdental fricative that the Koreans heard, and used that character to represent their own Korean interdental fricative -- which was not "r" originally. Over time, it softened to "r", but originally it was an interdental fricative, in P-K or Pre-PK.
The status of "l" in P-K is secure, since the Chinese character they borrowed began with "l". But the precise nature of the "r"-like phoneme in P-K and OK is not secure. It could have been "r", but it could have also been an interdental fricative, which later softened to "r".
In any case, the Korean "r"-like phoneme eventually merged into "l" during Middle Korean, so the "r"-like phoneme was not very stable. So perhaps the "r" -> "l" change was just the last in a longer series of changes to an unstable phoneme, which was originally "ð" as in Eskaleut, then probably devoiced to "θ" when P-K lost voiced obstruents, then softened further to "r", and finally merged into "l".
When this unstable phoneme became "r", IDK. But it came from earlier "ð" or "θ", just like Eskaleut.
Having "l", lacking "r", but having something "r"-like i.e. "ð" or "θ", is very Eskaleut. And it's quite possible that's what P-K had as well.
As for vowels, P-K or OK had many more vowels than Eskaleut, but as I said, I think that was their strategy to preserve lost info from their compressed consonant inventory, such as losing labialization on velar consonants. Japonic opened up a 2nd V slot at the end of the word, formerly of shape CVC, now CVCV. Chinese and other unrelated Southeast Asian languages dealt with deleted word-final consonants by preserving that info onto the nuclear vowel, by means of new tones.
ReplyDeleteSo it's natural to think that the heavily compressed consonant inventory of P-K lead to an expansion in the vowel system somehow. It wasn't a whole new vowel slot, it wasn't tone on the sole vowel. Well, it must have been an expansion in the number of vowel phonemes, where Korean has around 10, depending on exactly which stage.
Having so many vowels is unlike any other language family in the region, so it's natural to think that it was not always like that, but was the result of a major expansion to the vowel inventory. And why else would you do that, unless it was to compensate for a compression in the consonant inventory, at least in some positions of the word?
P-K does have all 4 vowels of Eskaleut -- "a", "i", "u", and "ə". It also has "e" and "o" -- just like Tungusic, which came from Eskaleut origins. So if Korean has an Eskaleut parent, it would not be impossible to get "e" and "o" -- so did Tungusic, another language with an Eskaleut parent. It also has a high central unround vowel "ɨ", unlike Tungusic or Eskaleut. Probably just to expand the vowel space into a former vacuum.
I deny that P-K or OK had any kind of vowel harmony, which I'm not alone in believing. The main reason for thinking vowel harmony existed originally, was supposing that the vowel harmony of Middle Korean was merely preserving what came before. But now some believe it arose from the collapse of weak vowels during MK, which led to initial consonant clusters, an unusual thing, and then later led to vowel harmony as a way to re-stabilize the vowel system.
IDK about all the intricate arguments, I just know that there are hardly any words reconstructed for P-K, and even OK may lack info about the main vowel, if it's in the logogram part and not the phonogram part.
If most of the intuition that Korean had vowel harmony was based on grouping it with Altaic or specifically Tungusic, maybe that grouping is wrong. Tungusic only has vowel harmony cuz it adopted the Altaic foundation, which is Samoyedic (with a Yeniseian accent), and Uralic has more vowel harmony than it knows what to do with.
But Tungusic's Eskaleut precursor did not have vowel harmony, since there are only 4 vowels in Eskaleut -- not enough to set up a clear system of oppositions. If Korean also has an Eskaleut parent, then it may not have had vowel harmony either, even though it had sufficient vowels to set up oppositions. If it was targeting an L1 of Wa / Japonic, neither of those have vowel harmony, nor does Yeniseian or Na-Dene.
At most, there's Arisaka's Law for P-J, but that's not vowel harmony. And that may have only existed in Japonic proper, not Wa, since Yeniseian and Na-Dene lack vowel harmony. So if Korean's target was Wa, not Japonic, there was no vowel harmony in the target.
Since Korean seems to have no Uralic ancestry, I doubt it had vowel harmony at birth. Lack of vowel harmony is compatible with a Wa-Dene-Yeniseian parent, as well as an Eskaleut parent. It is not compatible with a Uralic or Altaic parent, nor with a Nivkh parent, since Nivkh has vowel harmony (high vs. non-high, as opposed to the front vs. back system of Uralic / Altaic).
I was very glad to see skepticism of vowel harmony in Old Korean or Proto-Korean -- not just cuz it supports my view that Korean has Wa-Dene-Yeniseian and Eskaleut mixed ancestry, but cuz I'm sick of seeing all these overly complicated vowel harmony systems, where they're clearly trying to force vowel harmony into existence, and the evidence keeps resisting it.
Korean doesn't allow initial "l", which is like the Altaic languages -- but also like Eskaleut! Proto-Nivkh does allow initial "l", another strike against Nivkh being part of the ancestry for Korean. But yet another way in which Korean is compatible with Eskaleut ancestry.
ReplyDeleteThere was some speculation that Korean used to have an ergative-absolutive case system, based on the case marker "-i" being used for subjects of intransitive verbs and the objects, rather than agents, of transitive verbs. Later this was challenged, and re-interpreted to reflect topic vs. focus in the domain of pragmatics rather than morphosyntax.
ReplyDeleteIDK about the status of that debate -- all I know is, there's something at least superficially ergative-absolutive about the earliest stages of Korean, and that's very strange for its region. Especially compared to the Altaic languages and Japonic, which are as rigidly nominative-accusative as languages can be.
But who else in the region has an ergative case system, or at least split ergativity? Eskaleut!
Perhaps the bizarre use of the "-i" case marker in earlier stages of Korean reflected its gradual transition from a somewhat ergative system to a fully nominative-accusative system, which is where Korean is at now, just like the Altaic and Japonic languages.
That means that, before ample attestation was available, perhaps Korean had an ergative or split-ergative system -- which no one supposes about Japonic, or the Altaic languages. The fact that it's even a topic for debate, points to Eskaleut ancestry for Korean. Not entirely, it's a creole. But who else could bring ergativity to Korean, even in a small fleeting dose? Eskaleut.
That's it for the typological comparisons. Tomorrow I'll write about the roots of early Koreanic speakers in Manchuria, to relate back to the topic of Tungusic being of heavily Eskaleut origin. Some Eskaleut speakers joined the Altaic creolization, and that became Tungusic. But other Eskaleut speakers from Manchuria migrated away from the crowded and rowdy Steppe, into Korea, and they fused instead with Wa speakers, whose resulting creole was Korean.
ReplyDeleteWhen Korean had 2 "liquids", both were banned word-initially. The ban on "l" is compatible with Eskaleut phonotactics, but so is the ban on the "r"-like consonant -- which was probably "ð" originally in Pre-Proto-Korean. Eskaleut bans the voiced / fricative consonants word-initially, including "ð". So even though "ð" later changed to "r" in Korean, it would have reflected the same phonotactic rules about where it could appear -- like not initially.
ReplyDeleteKorean *does* allow voiceless fricatives initially, like "s" and "h", but not the voiced liquids "r" or "l". So perhaps Korean phonotactics focused more on the voicing feature, and not the fricative feature, from their earlier Eskaleut phonotactics, where these features are redundant. Korean eliminated voiced obstruents, so now stops and fricatives are both voiceless.
Since Korean allows "s" and "h" initially, but not "r", that means the change from "ð" to "r" (perhaps through an intermediate stage of "θ") must have happened early. Otherwise, if it simply lost its voicing, becoming "θ", then it should be allowed initially, just like "s" and "h", the other voiceless fricatives. But it was not. That means it had already changed to "r" before the appearance of these rules about word-initial consonants.
The alternative model is that Proto-Korean did not eliminate all voiced obstruents at once, but allowed some, like "ð", to exist as phonemes. However, the voiced obstruents were still banned initially, as in Eskaleut. That means the change from "ð" -> "r" could have happened much later.
IDK about that, since "h" was allowed initially way back in P-K, as shown by the word for "sun; day" = "hoy". If that was still voiced as in Eskaleut "ɣ", it would not be allowed initially according to Eskaleut sound rules. Also, it would be a bad choice for the Wa-Dene-Yeniseian root, which begins with a voiceless fricative, "xʷ".
This alternative model says that only "ð" was allowed to persist as a voiced obstruent, whereas "ɣ" was devoiced very early. That seems not symmetric or systematic, and ad-hoc.
So, Korean "r" has its origin in Eskaleut "ð", but it changed to "r" fairly early, and was never allowed initially, either as "ð" or as "r". Whereas "h" has its origin in Eskaleut "ɣ", but was devoiced early, and *was* allowed initially, since it's voiceless.
If "ð" was devoiced early to "θ", it was also softened to "r" right after that, at an early stage, since it was not allowed initially. If that 2-step change, "ð" -> "θ" -> "r", would have taken too long to happen early, then there was no intermediate stage, and "ð" changed directly to "r" at the early stage, so that neither stage of its life could ever appear initially.
Now for the cultural migration history that led to Korean ethnogenesis, and its link to the Eskaleut language family.
ReplyDeleteMuch like other cultures of Siberia, during the late 1st millennium BC and early 1st millennium AD, there was an insane game of musical chairs in Eastern Siberia, which led to linguistic creolization like there has never been before or since.
That's what makes it so hard to assign the languages of NE Asia to large families, and why they all seem like one-language isolates -- many of them are half-siblings that share one parent, but also have another parent, due to being creoles.
The standard model of linguistic parentage and descent, assumes 1 parent only, not 2 or more. Well, by that model, there are hardly any language families in NE Asia. And certainly no families with more than 2 branches. But this misses the reality of the birth of new language families through creolization, not only uniparental descent.
Korea was no different from the rest of NE Asia in this respect. There was a widespread language family already there by the late 1st millennium BC -- Wa (sister to Na-Dene and Yeniseian), or perhaps in some parts, its creolized form Japonic (after incorporation of Emishi / Ainu speakers, if there were any in southern Korea).
Japonic would have been confined to southern Korea, the so-called "Han" regions of ancient Korea (no relation to the Han ethnicity of China), and Wa would have been spoken in northern and central Korea, and who knows, perhaps in parts of southern Korea where they had not incorporated Emishi / Ainu speakers and become Japonic.
Sometime around 0 AD, Manchurian migrants began entering Korea. Some traversed the mountain range the separates northern Korea from Manchuria, and then headed straight south, settling the western parts of Korea. Others avoided the mountains by heading east until they reached the coast, and traveled south along the coast until they bypassed the mountains, and settled the eastern parts of Korea.
All of them came from a kingdom called Buyeo (AKA Puyŏ), which was in the exact region that I discussed before, regarding the Eskaleut origin of the name Xianbei ("Sarpi", meaning "forked" in Eskaleut, cognate with "whale tail" in Eskimo). Namely, the fertile plain that lies nestled between the Greater and Lesser Khingan Mountain ranges, in Manchuria.
https://en.wikipedia.org/wiki/Buyeo
Since I've already established that Eskaleut was spoken there, during the genesis of the Tungusic family, it stands to reason that Eskaleut was the language of the Buyeo migrants into Korea, since that was the same time period as the genesis of Tungusic -- late 1st mill BC, early 1st mill AD.
To speculate about the meaning of the name Buyeo, which was originally "puyʌ", I found a good match in Proto-Eskimo "puvyaʀ" = "crop", from the root "puvə" = "to swell", in the sense of crops growing or expanding up from the ground. That region in Manchuria was the most fertile plain for growing crops, unlike the mountain slopes, heavy forests, or tundra regions to the north. So it's semantically fitting.
Phonetically, "pu" is fine, and we've already seen "a" -> "ʌ" to preserve a lost secondary labialization feature "ʷ" from an adjacent phoneme. Here, "v" is labial or labio-dental, and since it's voiced, it cannot exist in Korean, nor was "w" an option either, since that never existed in Korean. So a consonant with a labial feature was deleted, and Korean preserved that lost info by shifting "a" -> "ʌ", as in the word for "sun; day". This would be even easier if the "v" and "a" were adjacent, so perhaps there was a metathesis of "vy" -> "yv", before "v" was deleted and its labialization preserved on the following vowel. Final "ʀ" is just a case marker for the dictionary form of a noun, so it could have been deleted with no semantic loss.
That yields "puyʌ", QED. ^_^
Korean locative particles from Dene-Yeniseian origin. So far, it looks like the Korean lexicon is mainly Wa-Dene-Yeniseian, and the Eskaleut influence is mainly attempting to speak Wa with an Eskaleut accent. So there's an Eskaleut influence on phoneme inventory, phonotactics, etc. But lexically, they seem to be targeting Wa.
ReplyDeleteMoving from standalone words to bound morphemes, which are never borrowed, there are two Old Korean locative particles (which later fused into a single particle), "-a" and "-kuy".
https://en.wiktionary.org/wiki/%E8%89%AF#Old_Korean
https://en.wiktionary.org/wiki/%E4%B8%AD#Old_Korean
Proto-Yeniseian has a locative suffix "qej", which must be the target of "kuy". Korean lost uvulars, and they fronted "q" -> "k".
Coda C's are fine, especially a glide like "y" / "j", so that was fine too.
The vowel transcribed as "u" in Korean studies is the unround, high, central vowel (between "i" and "u", and unround). This is not the same as "e", but they are both unround. Perhaps the lost info about uvular place in "q" was re-encoded onto the vowel by pulling it back, since uvular is further back than velar. That still leaves a raising, to high height -- perhaps influenced by "y" being "i"-like, which is high, or there was some kind of vowel-raising in the history of Korean, as there was in Old Japanese.
Pretty close match, overall.
Vajda also reconstructs a locative particle for both P-Y, "-aq", and P-Na-Dene, "aχ(e)", whose nuclear vowel is "a", has no C before the V, and whose coda C is a uvular.
To go from this to OK "a", all that is needed is deleting final uvulars rather than trying to render them. Perhaps first rendering them as a glottal stop (a far-back stop C), but later re-analyzing that as a zero phoneme -- just as happened in the history of Japonic, with many of its former uvulars from its Wa stage. In this model, uvulars were only rendered in initial, maybe medial, position.
Neither of these Yeniseian locative particles appear in Japonic, so once again we see that Korean was not based on Japonic, but on Wa, the sister to Na-Dene and Yeniseian, an earlier stage of Japonic before the incorporation of Emishi / Ainu speakers.
What is the logic of rendering uvulars in initial and medial position, but not final? If they wanted to preserve the lost info about uvular being further-back than velar. In "qej", the lost info about "q" was re-encoded onto the *following* vowel, pulling it back to central. But if "q" is final, as in "-aq", it has no *following* vowel to re-encode any info onto.
ReplyDeleteIf preserving lost info was crucial, then the inability to do so meant not rendering the old phoneme at all, just deleting it. Again, maybe through an intermediate stage where it became a glottal stop as a place-holder "far-back stop C", then becoming a zero phoneme. The glottal stop was not a phoneme in this model, just an allophonic variant that is easier to pronounce than a uvular stop, in a language that has no uvulars.
English does not have uvulars, but we can make a glottal stop perfectly fine, sometimes even in final position, as an allophonic variant -- like most contempo British dialects, even including RP, pronouncing word-final "t" as a glottal stop, such as in "what?" They're glottalizing a front consonant, "t", but glottalizing "q" is even more phonetically motivated, since the locations are nearly the same.
Korean proximal pronoun is from Wa origin, possibly Dene-Yeniseian as well. In OK, it's "i". In Proto-Japonic, there's a word "ima" = "now", which Frellesvig interprets as "i + ma" = "proximal pronoun + place", whose "i" is cognate with the Korean proximal pronoun.
ReplyDeletehttps://en.wiktionary.org/wiki/%E6%98%AF#Old_Korean
https://en.wiktionary.org/wiki/%E4%BB%8A#Japanese
Previously I said this "i" in P-J "ima" was cognate with Yeniseian "qej", since Japonic prefers to simply delete uvulars (especially if they're not labialized, where the labialization could be preserved as a labial C). Then the "j" vocalizes to "i", the V sequence is banned, and the 2nd V replaces the 1st, yielding simply "i".
That may still be possible -- but considering that Korean is targeting Wa for its lexicon, perhaps the "i" in P-J "ima" was always just "i", as in OK, both reflecting "i" from Wa.
The proximal demonstrative "pronoun" in Yeniseian is "ki-", and in Proto-Athabaskan it's "jii" (affricate, not approximant). So the "i" in Korean and Japonic would seem to be only from Wa, not cognate with Yeniseian or Na-Dene forms.
However, perhaps Wa dissimilated palatal affricate "j" from palatal vowel "i" -- which is just like the Japonic ban on "yi". Tungusic also dissimilated palatal approximants from "i", likely reflecting Eskaleut phonotactics. Since Korean came from an Eskaleut-accented background, maybe they did something similar, at least for "ji". Korean has always banned "yi", like Japonic, although it does allow palatal affricates followed by "i", just like Modern Japanese does.
But at some earlier stage, perhaps the Wa proximal pronoun was very close to the Athabaskan / Na-Dene one, "ji". Both of those might be cognate with Yeniseian "ki", with palatalization of "k" -> "j", although IDK if that's a systematic correspondence. Maybe just in some limited cases like bound morphemes.
Then suppose Wa "ji" softened to "yi". Yeniseian does not allow "yi", so probably that was true for Wa as well, judging from Korean and Japonic. Na-Dene allows "yi", though. At least for Wa, though, "yi" -> "i", and that was inherited into both Korean and Japonic.
Otherwise, there would need to be 2 separate changes of Wa "ji" -> Korean "i" and Japonic "i". It's simpler if there was only 1 change within Wa, where "ji" softened toward "yi", a banned sequence, resulting in "i". Then since both Korean and Japonic are based on Wa's lexicon, they both adopted "i" as the proximal pronoun -- always for Korean, only in the archaic fossilized component "i" of "ima" for Japonic.
Returning to the big-picture of Korean ethnogenesis, this model means that the foreigners from Manchuria tried to adopt the local language of the commoners, which was Wa by the time the Manchurians began settling Korea, circa 0 AD and after. It was not the locals adopting the language of the elite foreigners, as though it were "Eskaleut / Tungusic spoken with a Wa accent".
ReplyDeleteThe phoneme inventory of P-K looks more like a compressed version of Eskaleut than a compressed version of Wa-Dene-Yeniseian. It's an Eskaleut accent, not a Wa accent. The target must have been Wa, not Eskaleut.
This is a common pattern in inter-cultural history, where the elite foreigners adopt the local commoners' language rather than vice versa. In East Asia, several Steppe groups assimilated into the Chinese norm, not vice versa, such as the Xianbei and the Manchus, after the Steppe group conquered the Chinese.
In Europe, the Bulgars spoke Turkic, then adopted Slavic from the commoners who they led. Franks switched from Germanic to Romance when they took over France, Normans switched from Norse to French when they controlled Normandy, and an off-shoot of the Normandy-based Normans then switched again to English when they controlled England after liberating it from the Viking Empire's occupation.
And so on and so forth.
In Korea, the elite migrants from Manchuria (Buyeo) adopted the language of the local commoners, Wa. It was not a straightforward conversion, since Korean is not literally a dialect of Wa. It was more of a creolization, but with Wa playing a similar role to L1, and Eskaleut / Tungusic playing the role of L2, in a model of language conversion.
Korean and Manchurian ethnonyms. Beginning with "Manchu", who give "Manchuria" its name. In their own language, it's "manju", supposedly from P-T "maŋgu", but also reconstructed as "mangu". This may or may not be cognate with the Tungusic name for the Amur River, which may be more like "mamgo". Here, I'm only focusing on the endonym of the culture from Manchuria, and am not looking into the name of the Amur River.
ReplyDeletehttps://en.wiktionary.org/wiki/Reconstruction:Proto-Tungusic/Ma%C5%8Bgu
Naturally this name has an Eskaleut origin, and it turns out the nasal is originally "n", not "ŋ". Perhaps some dialects later shifted "n" back to "ŋ" to agree in location with the following "g". Not in Manchu itself, though, where it is still "n".
P-E "maniɣ" = "to be smooth or flat", "maniɣuʀutə" = "to humble oneself before someone", P-I "maniʀaq" = "plain".
The root is "maniɣ", and here refers to a geographic "plain", since that is where Manchuria is. It's not the northern tundra, not the heavy forests, nor the mountain slopes of the Khingan Mountains, let alone the open Steppe, or the Gobi Desert, or anything else. It's the fertile plain nestled within the forking Khingan ranges, making it suitable to sedenary agriculture rather than other subsistence modes.
That's what sets Manchurians apart from other nearby cultures -- they sedentarized and adopted grain agriculture, unlike the nomadic pastoralists of the Steppe, reindeer herders of the tundra, or hunter-gatherers from the forests.
Probably it was their local name for the geographic region -- "the plains" as opposed to the mountains, forests, etc. Then they adopted the name for their endonym when interacting with others, whether the others were Tungusic-speaking or not. They were the "plains" people, not the reindeer-herding Tungusic speakers, not the nomadic Steppe Mongolic speakers, etc. They made sedentary grain agriculture so central to their identity, they put it in their own ethnonym.
"ɣ" -> "g", "ig" metathesized to "gi" to fit standard Uralic / Tungusic word-shape of CVCCV, and "i" pulled back to "u" (at same height) to harmonize with back "a", yielding "mangu", QED.
The epicenter for grain agriculture in East Asia, China, noticed that the Manchus and similar cultures were not like the other barbarians to the north -- these ones were sedentary and cultivated multiple grain species, even if they also raised some livestock, including horses, and still wore pants instead of robes, and were archers, reflecting their deeper origins as Eastern Steppe horse-riding nomadic pastoralists.
That was in the 2nd or 1st millennium BC -- now it's the 1st and 2nd millennium AD, and at least some of the northerners have left horse-based pastoralism behind, and adopted sedentary agriculture instead. Y'know, maybe we Chinese could assimilate them into our way of life? That's exactly what they did, even though the Manchus conquered China militarily and politically.
Very few descendants of the Manchus still speak a Tungusic language, or even live in Manchuria -- they willingly joined the Chinese, and by this point only a DNA test (or perhaps a look at their facial features) could reveal that their ancestors came from somewhere outside of China.
Korean ethnonyms Ye, Maek, and Yemaek. One of the main precursors to Korean culture was the Yemaek, a culture from Manchuria and North Korea.
ReplyDeletehttps://en.wikipedia.org/wiki/Yemaek
They are contrasted with the "Three Hans" or "Samhan" cultures from South Korea, which were more like Wa or Japonic. The Yemaek and Samhan eventually syncretized into a single culture, Korean.
Well, even the northern group, Yemaek, may not have been so homogeneous throughout their entire history. Their very endonym is likely a compound of two ethnonyms, representing an earlier cultural fusion. Namely, Ye and Maek.
"Ye" also appears in the Old Korean name for Japan and the Japanese, which was "YEri", whose final element "ri" is that same generic nominalizer we saw before with "NWUri". It seems to make something more abstract or extended in meaning, in this case from a toponym to the people who live there, or perhaps from one person of a certain culture to their entire culture. The root of this name is simply "Ye". This Korean name for the Japanese continued through Middle Korean, but was later abandoned in favor of a variant of "Wa" in the 20th C.
https://en.wiktionary.org/wiki/%E5%80%AD%E7%90%86
I'm not the first to notice this similarity, nor the first to infer that there was a Wa / Japonic culture still thriving in North Korea or perhaps Manchuria, during the time when Ye and Maek were still distinct from each other. The Ye would have spoken Wa, or some transition stage between Wa and Japonic.
Where I do have something new to contribute, is about the ethnonym "Maek" (the letters "ae" represent a single vowel, like "e" in English "met"). The existing theories bizarrely try to link it to other ethnonyms with a high(-ish) back round vowel like "o" or "u", when this vowel is front, low-mid, and unround.
For instance, one of Korea's names among outsiders was based on "mukri", where that final "ri" is the standard nominalizer and is unimportant, the root being "muk". Sorry, but the vowel in "muk" is waaaayyy too different from "maek" -- literally the opposite on every dimension, height, frontness, and roundness. And there's only 3 sounds in the entire name, so getting that wrong is damning. Other than that, a perfect fit...
Instead, I think "Maek" is a rendering of "Mangu", the toponym referring to the Manchurian plain. The various cultures from those plains were insistent on being different from the wild-and-crazy horse-riding nomads, or the backwards hunter-gatherers, or reindeer herders. They were sedentary, cultivated multiple crops, were capable of centralized states, stratified societies, civilization, and all that other stuff.
ReplyDeleteSince we know that the Maek were inter-mixing with the Ye, who spoke Wa or Japonic, let's run a Tungusic word like "mangu" through a Wa or Japonic phonotactic filter.
If it's Wa, the preferred word-shape is CVC, just like its sisters Na-Dene and Yeniseian. Certainly the Ye were fine with monosyllable words, and had not adopted the CVCV word-shape of Japonic -- cuz their own name is only CV, much like "Wa". The CVCV shape came later. So, that means they'll delete the final V, leaving "mang".
Eventually, Wa will lose its voiced obstruents by the time of P-J, so it must have happened between Wa and P-J. Perhaps that had already begun during the time of the fusion between Ye and Maek. That will devoice "g" -> "k", leaving "mank".
Wa, like Na-Dene and Yeniseian, bans coda clusters unless the 1st is a wimpy little approximant, which "n" is not. Japonic will enforce a stronger ban -- no coda consonants at all. Well, the evolution had not reached P-J yet, so single consonants were allowed in coda position, but not clusters like "nk". To preserve the original borders of the word, "m" and "k" were kept, and medial "n" deleted. That would leave "mak"...
But it's not "mak", it's "maek". I think they preserved some lost info about the deleted "n" onto the vowel. The most salient trait about what was deleted, was that it was a nasal consonant. Nasal consonant air-flow has similar phonetic properties to high vowels, so the best way to encode nasality onto a vowel is to raise its height. Assuming you aren't going to literally nasalize the vowel, which was not an option in this case, since vowels could not carry phonemic nasality.
That means raise "a" -- to what? Well, it should stay unround, and be central or front, not back (since back vowels tend to be round). They chose "ae" to show that it was close to "a", but had a nasal after it. If they raised it too high, to "e" or "i", how could they tell that it was originally "a" rather than "ae"? By making the result "ae", there is only 1 possible vowel it could have been originally -- something lower than "ae", namely "a". If they made it "i", it could have been any number of vowels below "i", like "e", "ae", or "a".
With that twist on the vowel to show it used to be followed by a nasal, that yielded "maek", QED.
I'm not sure whether the original sense was "plains people", or specifically the ethnic group later known as "manju" (AKA "Manchu"). But given how early the name "Maek" was in use (early 1st millennium AD), and how many centuries pass until "manju" appears (in the 2nd millennium AD), I doubt they referred to the same group, just at different time-periods. Probably they are independent derivations or doublets from the same Tungusic root (of Eskaleut origin), meaning "plains people".
Next is "Korea" itself, whose earliest form is "Goguryeo", where "go" = "high, big" and "guryeo" = "castle, fortress". Since there were no voiced obstruents in P-K or OK, the original pronunciation was "kuryʌ". This referred to a polity that controlled central + northern Korea and southern + central Manchuria, from roughly 0 to 700 AD. A successor polity abbreviated the name to "Goryeo", pronounced "koryʌ", which is where "Korea" comes from (with "o" rather than "u" as the 1st vowel).
ReplyDeletehttps://en.wikipedia.org/wiki/Goguryeo#Names_and_etymology
It is already known that the only other language family with a cognate for "castle" is Old Japanese "ki". So we ignore an Eskaleut or Tungusic origin, and treat Goguryeo as Wa / Japonic, at least for this term. Instead we look for Wa-Dene-Yeniseian origins.
"Ko" is a loanword from Chinese, so we ignore that, and focus on the "castle" word.
Proto-Yeniseian has a spatial / positional word "kewɬ" = "outside", as well as a form with a noun nominalizer suffix "ja" added on, "kewɬja", meaning "outside" but with an even stronger noun connotation. In Yeniseian, adjectives and spatial words behave like nouns, not verbs.
There is also a suffix "ʌ" that has to do with collective nouns, sometimes also followed by "ŋ", whose combined form is listed as "Vŋ" = "plural / collective noun suffix". Because the V there is unpredictable, but usually a back one, "Vŋ" is treated as a fusion of an earlier vowel suffix and a "ŋ" suffix. The earlier vowel-only suffix is presumed to be "ʌ".
I think Goguryeo "kuryʌ" = "castle" is cognate with P-Y "kewɬ" or "kewɬja" = "outside", plus the collective noun suffix "ʌ".
Semantically, a castle or fortress or other defensive compound is defined by being surrounded by walls, moats, palisades, fences, earthen mounds, and other large physical barriers that lie along a periphery, which protect the central or core part of the compound. It's not just a single building, but one with thick impenetrable walls, and also being part of a larger compound that has nested layers -- a defensive periphery, and a core where all the main activities take place.
ReplyDelete"Outside" is related to "outer" or "exterior", and in the context of a building site, with "boundary" and "border" and "periphery", all contrasted with "inside", "inner", "interior", "enclosure", and "core".
These spatial terms can be given more concrete form by a nominalizer, and especially a collective noun suffix. An "outer side" or "exterior" can refer to a fairly thin dividing structure, like the sides of a tent. Technically, a tent has an interior and exterior part, and the material making up the tent is a kind of wall or boundary. But it's pretty thin, and no one would use "facade" to describe a tent's exterior -- it has to be bulkier and more solid and thick than a tent's sides.
A border or boundary or periphery is, in the simplest form, a wispy-thin 1-dimensional line that loops around the space it encloses. In some cases, it's not tangible at all, just some socially agreed-upon convention about where one plot of land ends and another begins, with no physical barrier separating the 2 plots whatsoever.
But, place a bunch of those 1-dimensional loops near / in contact with each other, and suddenly they form a 2-dimensional or 3-dimensional thick solid barrier between the 2 spaces. In the same way that a bunch of individual strands or fibers can collectively form a single rope, which has more solidity than all of the individual strands, if they were not joined together. 100 wispy-thin strands, which are not in contact with each other, are still wispy-thin in their totality, since 100 times 0 is still 0. But joined together, they have a collective solidity.
So, in the Goguryeo word "kuryʌ", "castle" is referred to in a pars pro toto way -- by its massive, solid, defensive walls or other barriers around the periphery of the compound. Possibly referring to the thick walls of a single building, but usually there's another word for "wall". This word comes from a more abstract spatial term about "outer, exterior, periphery", which usually implies it lies some distance away from the core, and sounds more like the defensive barriers surrounding the entire compound.
Phonetically, P-Y "kewɬ" has the right initial C. Wa / Japonic evolved toward eliminating coda C's, especially if they were part of a cluster. So "w" was vocalized into "u", leaving a vowel sequence "eu", which is also banned in Japonic, and probably in Wa as well. As in Japonic, the 2nd V replaced the 1st V, leaving only "u" as the vowel.
ReplyDeleteThere are no lateral fricatives in the late stage of Wa, as it transitioned into Japonic. In P-J, "s" is the corresponding sound for "ɬ" in Na-Dene and Yeniseian. There is no "l" in P-J, so the only place that "ɬ" can go to, is "s". But before P-J, there may have been a stage where "ɬ" -> "l", at least in word-final position (probably not word-initially, since liquids are banned initially in Yeniseian and Japonic, therefore probably in Wa as well).
Maybe there were several branches or dialects of Wa, and the branch leading to P-J changed "ɬ" -> "s" unconditionally, while the branch leading to Goguryeo changed "ɬ" -> "l" word-finally, probably "ɬ" -> "s" or "y" word-initially.
The yields "kul" so far.
Attaching the P-Y nominalizer "ja", which would be written "ya" in Korean studies, yields "kulya".
That is sufficient semantically to derive a concrete, tangible "surrounding barrier" from an abstract spatial term for "outer". But phonetically, the final V is somewhat lacking, and should be "ʌ". So far, the only change from "a" -> "ʌ" involved a deleted labialized consonant, and no such thing is going on here.
So, tack on that P-Y collective noun suffix "ʌ", which strengthens it semantically, not just a noun but a collective one like a defensive barrier should be, but also gives the proper final vowel. The sequence "aʌ" is resolved as usual in Wa / Japonic, where the 2nd V replaces the 1st V, yielding "kulyʌ", QED.
Why the liquid was later altered to yield "kuryʌ", IDK, perhaps to dissimilate "l" from "y", while keeping it liquid.
As for the cognate in Old Japanese "ki", that is modeled as cognate with "kuy" in the Baekje language, which is either cognate with or a loan from "kul" in the Goguryeo language, with "l" -> "y" word-finally.
Perhaps, though the Baekje word simply wanted a monosyllabic word-shape, as is standard for Na-Dene and Yeniseian and Wa. That would result in some loss of semantic transparency, but if they were borrowing the Goguryeo word, they already knew what it meant semantically, so they could eliminate some of the suffixes that had been used to construct the full meaning in the first place.
So, they Baekje would delete the final V, which leaves an illegal coda-cluster, which is resolved by preserving the border sounds of the word and eliminating the 1st sound in the cluster, yielding "kuy", QED.
From there, its loan-form into P-J or OJ would vocalize coda "y" to "i", since coda C's are banned, and the vowel sequence "ui" is resolved by keeping the 2nd and deleting the 1st, yielding "ki", QED.
To emphasize why a barrier wants a collective connotation, it's always constructed from a bunch of smaller pieces, which together form a collective solid barrier. Whereas if they were not joined together, all the individual pieces would not be even 1% of a barrier -- like a bunch of bricks strewn widely around the ground, rather than joined together by masonry. Or piles of dirt strewn about, rather than compacted into a large mound that ran around the periphery. Or a bunch of sticks strewn widely around the ground, rather than tied together to make a fence or palisade.
ReplyDeleteEven forming them into a geometrically regular arrangement is not enough -- they have to be close together, so that nothing can pass between them. A periphery that has 1 sharp stick spaced every 5 feet apart, is no palisade. When they're 5 inches or 5 millimeters apart, *then* they make a palisade.
Same with a watery moat -- all those drops of water form a collective moat, as opposed to the same total volume of water being separated into 1000 little pools.
No matter what the defensive barrier is, it's made up of building blocks, sometimes literally. A bunch of identical componenets, placed close together to form an entire whole, invites a collective noun suffix, not just any ol' noun suffix.
Finally, "Joseon" also has a Wa-Dene-Yeniseian origin. This is the earliest name of a polity from Korea, which flourished in North Korea and ended around 100 BC, beginning who knows when, but several centuries before then at least. The name was later used for the polity that controlled Korea from about 1400 to 1900 AD, and the ancient one received the title "Gojoseon" meaning "Old Joseon" to distinguish it from the Early Modern one. North Korea continues to use "Joseon" in its full name, while South Korea does not.
ReplyDeletehttps://en.wikipedia.org/wiki/Old_Chos%C5%8Fn
I immediately noticed that this name sounds awfully similar to "Jurchen", the Tungusic-speaking culture that expanded to control Manchuria and NE China during the Jin Dynasty, and later seeded the Manchu culture who conquered China as the Qing Dynasty. In Manchu, the name is "Jušen", no "r" or other liquid in the middle, nearly identical to the pronunciation of "Joseon", which is "Josʌn".
Well, great minds think alike -- Janhunen and others have already explored the idea that ethnonyms have remained similar in NE Asia, throughout the centuries. And Korean "Joseon" being cognate with Manchu "Jušen" is one such example they cite.
What I have to contribute is the deeper origins of this name. At first, you might think it's Tungusic (or Eskaleut), since it is used by the Jurchens, who were Tungusic speakers, as were their successors the Manchus. If true, that would place a Tungusic or Eskaleut-named group in control of North Korea during the 1st millennium BC, as (Go)Joseon.
But here's where my model of the Eskaleut origins of Tungusic can help out. Namely, it absolutely rules out a Tungusic origin of this name. Why? Well, regardless of what deeper origins you think Tungusic has, just focusing on Proto-Tungusic itself, there are zero (0) words that begin with "ʒ" and have "s" or indeed any other fricative (like "x") anywhere else in the word. Words beginning with "ʒ" only have stops, liquids, and nasals as other consonants.
And in reverse order, of the 32 words beginning with "s" (at Wiktionary), only 1 has "ʒ" elsewhere -- "sugʒansa", from root "sugʒa", meaning "fish".
So, in Proto-Tungusic, "ʒ" and "s" really do NOT want to occur in the same word, and they appear to be fully banned from co-occuring if the "ʒ" is word-initial.
Regardless of *why* you think this rule exists, the fact is, it does exist. And that rules out a Tungusic origin of the name "Jušen", which begins with "ʒ" and has a sibilant elsewhere.
My model, however, does explain why these consonants do not occur in the same word. Both "ʒ" and "s" are not native to the Eskaleut phoneme inventory, and they only arose as alterations of Eskaleut "c". But they are separate strategies to deal with "c", and were not deployed at the same stage of history. The earlier strategy was "s", since there are tons of P-T words with "s" in every position of the word, whereas "ʒ" must be a later strategy since it's far less common and mostly occurs initially.
A Tungusic origin of "Jušen" would have to be a deeper Eskaleut origin, but neither "ʒ" nor "s" / "š" exist in Eskaleut, so both would have to be derived from "c". However, why would word-initial "c" be altered to "ʒ", while medial "c" was altered to "s / š"? That is not motivated, only ad-hoc, and is contradicted by the Tungusic lexicon, where no word exists with both of these consonants.
Furthermore, there is no Eskaleut root that has "c" as the 1st and 2nd consonants, regardless of the vowels. The closest is "cuki(ʀ)" = "splinter or thorn", where the "ki" could be palatalized to "ci", yielding "cuci". There is no semantic grounds for supposing that the name "Jušen" had anything to do with "splinter or thorn".
Phonetically and semantically, we reach a dead end for a Tungusic / Eskaleut origin of "Jušen" and "Joseon".
So instead we pursue Wa-Dene-Yeniseian...
Before that, I'll establish a Wa-Dene-Yeniseian origin for the Japonic suffix "si", which appears at the end of adjectives in Old Japanese, and in a few tool / instrument words. This will play a role in the etymology of "Joseon", but it's worth demonstrating on its own, to further link Japonic with Dene-Yeniseian.
ReplyDeleteProto-Yeniseian has a suffix "še" which is a nominalizer, and usually attaches to verbs. When it attaches to a verb, it is used in 2 ways. First, as a relativizer, like deriving a noun phrase "he who hunts" from the verb phrase "he hunts". Second, as the agent or do-er of the verb, and sometimes the patient -- both of these being participles, or noun forms of a verb.
However, it also attaches to adjectives to produce a noun form of the adjective. This is then used as the predicate of a subject noun, to express that that the subject is characterized by some trait. Crucially, no separate copula verb is used -- just "subject trait". Both of these words are nouns, a subject noun and a nominalized form of an adjective. The morpheme that turns the adjective into its nominalized form, is "še".
The reflexes of P-Y "še" include "si". And that is what I think the Japonic suffix "si" is from, phonetically. P-J strongly favors high over mid vowels, so it raised "e" to "i". It does not have a separate palatal sibilant, only "s". So as elsewhere in the Dene-Yeniseian family, it became "si" in Japonic.
The Dene-Yeniseian framework highlights a structurally similar suffix in Na-Dene, although it has had its consonant eroded over time, and mainly shows up as "e" or "i".
How does this suffix function in Japonic, though? Well, Japonic is not polysynthetic like Na-Dene and Yeniseian, so it can't just stick a suffix onto an entire verb phrase, to relativize it. That usage does not happen.
But it does appear as a suffix for at least one tool or instrument -- "kusi" = "comb; skewer, spit". It attaches to a root "ku" which relates to teeth and biting. E.g., "kupu" = "to bite, eat", "kupa" = "mulberry", "kutui" = "mouth", "kutira" = "whale", "kuki" = "nail" (spike for driving through something to fasten it, much like a tooth piercing food), and possibly "kusa" = "grass" (where a blade of grass looks like a tooth, same semantic development for "pa" meaning both "tooth" and "blade" in P-J).
A comb has rows of teeth, and a skewer or spit is like a nail that pierces through food, much like a tooth does. In many languages, the word for "tooth" is used to derive the word for "peg" or "nail" or "spike" or other fastener.
So, "kusi" is "the thing that bites" or "the thing that uses its teeth", and suffix "si" acts as a nominalizer attaching to a verbal root. It's not attaching to a noun root about "tooth", since "tooth" = "pa" in P-J. Here, "ku" is more about the verb "to bite", not the noun "tooth".
But the most important role of "si" in P-J and OJ was at the end of an adjective. In OJ there are 2 types of adjective endings, "si" when the adjective is used as a predicate -- "subject is X" -- and "ki" when used to modify a noun -- "X noun".
ReplyDeleteOver time, "si" was dropped in favor of "ki" for both usages. Then "k" was eroded, leaving only "i" as the adjective ending for Modern Japanese -- very much like its cognate in Na-Dene, with its deleted consonant, and only a high-front unround vowel.
In OJ, adjectives used as predicates of a subject, functioned like nouns, not verbs. The sentence construction was "noun noun" or maybe with a topical particle after the 1st noun, meaning that the 1st noun is the 2nd noun. Crucially, no copula verb was used a la "da / desu" in Modern Japanese, just "noun-1 (topic particle) noun-2", as in Russian, Arabic, and many other zero-copula languages.
For example, "kare pa yamato-pito" means "he is a Japanese person". Here, "kare" is the pronoun "he", "pa" is the topic particle, "yamato" is the old name for Japan, and "pito" means "person". There is no copula verb, just an equating of one noun with another noun.
Well, adjectives worked the same way in OJ, and likely in P-J as well. But instead of a bare root referring to the trait, the suffix "si" was added. For example, the root we'll see relating to "Joseon", P-J "tuyo" = "strong", but when used as a predicate of some subject, it's "tuyosi".
So, "he is strong" is "kare wa tuyosi", again with no copula verb, but with "si" added to the adjectival root. What this means is that "si" is a nominalizer, since the standard construction for subjects and predicates is "noun noun". So, if the predicate is an adjective, it must first be turned into a noun to fit the "noun noun" format. And the suffix "si" is what turns the adjective root into its noun form, so it can be used as the predicate of a subject.
Earlier linguists of Japanese have theorized that this "si" at the end of adjectives, was an earlier copula verb that fossilized onto the adjective root. Well, that's pretty close -- but it was actually a nominalizer suffix, not a copula verb.
The use of a copula verb is much later in Japanese history -- at the earliest stages, it was a zero-copula language. But in such languages, they insist on a closer match between word categories, like "noun noun", maybe with a topic particle to mark the subject / topic, and the unmarked one as the predicate.
So, P-J / OJ "si", the nominalizing suffix involved in deriving stative predicates, is not only a phonetic match for P-Y "še", it shares 2 crucial functions as well. Rarely, it was used to derive a tool or instrument from a verbal root, like "kusi" = "comb", from a root about "to bite". But far more productively, it was used to derive stative predicates, for use in a zero-copula construction to say that a subject is or has some adjective trait -- exactly like in Yeniseian.
The only function that it lacks, compared to Yeniseian and Na-Dene, is the relativizer, but that's due to Japonic no longer being polysynthetic like its distant ancestor, so it can't package a million shades of meaning onto a single verb root.
Another crucial link between Japonic and Dene-Yeniseian! ^_^
Now, onto the Wa-Dene-Yenisian origin of "Joseon" and "Jurchen" ("Jušen"). This ethnonym is not Japonic, but Wa, much more similar to Yeniseian than to the later stage or branch of Wa, where Emishi / Ainu speakers were incorporated into a Wa-speaking group, whose synthesis became Japonic.
ReplyDeleteFirst, the boring part about suffixes, since I just spent all that time on P-Y "še" and P-J "si". That suffix is the "sV" part of the ethnonyms. It derives a noun form of an adjective root, for use as a predicate of some subject.
The only phonetic remark is that Wa did not have more than 1 sibilant, so "š" -> "s". Later, in Jurchen / Manchu, the "s" became palatalized by the high-ish front vowel "e", to become "š" once again.
The vowel will get deleted during resolution of a vowel sequence for "Joseon", but judging from "Jušen", it could have been "e" originally, just like in P-Y.
After that derivational suffix, the inflectional suffix is cognate with P-Y "Vŋ", the plural / collective noun suffix, which is fitting for an ethnonym, referring to an entire society or culture, not isolated individuals.
That vowel is unpredictable, but usually "ʌ", and that is what it was for "Joseon". There is no such vowel in Tungusic, so they appear to have simply deleted it, and preserved only the vowel from the "še" suffix.
This assumes, however, that the Jurchens did not merely borrow the older ethnonym "Joseon", since there is no "e" vowel there, and it's phonetically implausible that they would deal with problematic "ʌ" by making it "e" of all things. They would probably make it "a" or "o" or low-mid "o".
Also, the vowels are not harmonic for "Jušen" -- back "u" and front "e" -- which reinforces that it's not Tungusic. If they heard "jusʌn", that would be the perfect opportunity to make the vowels harmonize, by changing "ʌ" to back vowels "a" or "o" -- not sending it forward to "e", which would create a disharmonic root.
So, the ethnonym was probably coined with the two suffixes being "še" and "ʌŋ", combined without deleting either vowel, "šeʌŋ", maybe or maybe not altering "š" to "s" (impossible to tell).
ReplyDeleteThe earliest attested form, "Joseon", behaved like a Wa / Japonic language, by banning the vowel sequence, and replacing the 1st with the 2nd.
But the original form must have remained in circulation somewhere in a more purely Yeniseian or early-Wa form, with both vowels. This 2-vowel form is what the Jurchens based their name on, and they were not Wa or Japonic speakers, so they did not resolve the sequence "eʌ" in the same way as the Joseon people. Rather, as Tungusic speakers, they didn't have "ʌ", so they simply deleted that vowel and left the other unproblematic one.
The last phonetic change is that "ŋ" did not exist in Proto-Korean or Old Korean. I was misinformed by the Wikipedia articles on those languages. It simply does not exist, only "m" and "n" are reconstructed, based on Korean evidence. Only the efforts to connect Korean to Altaic languages suppose it existed in P-K. But even external comparisons rule that out, like when Korean scribes borrowed Chinese words that had final "ŋ" -- it was "n" to Korean ears. They did not have a separate character for "ŋ", only "m" and "n", exactly like Japonic, and unlike the Altaic languages.
When did "ŋ" enter Korean? Only during Middle Korean, or the Early Modern era, as a softening of "h" (which itself was a softening of "k" from OK). Very strange, but not without precedents elsewhere in the world. You'd think if they wanted to soften "h" further, they would just make it a semi-vowel like "y", which they did in some cases. But in many others, they softened it to a nasal in the same place, "ŋ". Before then, Old Korean borrowed Chinese words with "ŋ", during the late 1st millennium AD. But "Joseon" was not a Sino-Korean word, and it was attested long before the massive borrowing of Chinese words into Korean.
So in "Joseon", final "ŋ" -> "n", the nearest nasal.
But even in Tungusic "Jušen", they mapped "ŋ" -> "n" word-finally, despite P-T having "ŋ" as phoneme. In Tungusic, it can only appear initially and medially, not finally. In fact, just about the only word-final consonant allowed is "n". So for different reasons, Tungusic also altered "ŋ" to "n" in this ethnonym.
That's just the suffixes! They mean "the X ones" or "the ones who are X", and we'll see what that root X is in the next comment.
Last is the root of the ethnonym, which is "strong", and is cognate with the Yeniseian word for "stone". So the full name means "the strong ones" or "the ones who are strong".
ReplyDeleteNot just like having big muscles and looking buff, which is a gay obsession. Nor about lifting heavy weights, which is also just about sports and contests.
When connected to "stone", it's about being formidable, unmoveable, impenetrable, impossible to shatter or fragment, incapable of being droopy or limpy, and serving as the perfect wall or barrier against the hordes from China or the Eastern Steppe, the perfect protector of the Korean peninsula. Nobody can get passed us.
In P-Y, the root for "stone" is "cew", and the P-J adjective "strong" is "tuyə", which are cognate with each other. There's a further cognate in P-Na-Dene "caj" = "stone". IDK if I discussed this Dene-Yeniseian origin of P-J "tuyə" before, or not, but it is.
https://en.wiktionary.org/wiki/Reconstruction:Proto-Yeniseian/cew%C3%A7
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/tuy%C9%99
https://en.wiktionary.org/wiki/Reconstruction:Proto-Na-Dene/caj
At this stage of Wa, or maybe in this regional dialect, in northern Korea away from the Japonic core in southern Korea and Japan, the palatal stop "c" was not eliminated as it was in Japonic. This is the sound transcribed as "j" in "Joseon", for its time (no voiced obstruents), more like English "ch" in "chose".
The branch of Wa leading to Korean will lose "w", under the influence of incorpoating Eskaleut / Tungusic speakers, who don't have "w". But even before that, when it was pure Wa, they still did not tolerate coda "w", as shown by its fate in Japonic. So they vocalized "w" to "o" or "u", and it replaced the preceding vowel in a V sequence, rendering "eo" or "eu" as "o" or "u". Here, they chose "o", which is more faithful to the height of the deleted "e" vowel.
Tungusic "Jušen" rendered it as "u", closer to the vowel in Japonic "tuyə", which is muddied by the fact that there are 2 vowels at different heights, plus a semi-vowel between them. Nevertheless, Tungusic "u" is still a reasonable reflex of earlier "ew".
Phonetically, that's the end of the story, P-Y "cew" corresponds to "co" or "cu" in "Joseon" and "Jušen". Semantically, it originally meant "stone", but acquired an adjectival sense to mean "strong, formidable, unbreakable", as also attested in the Japonic cognate "tuyə". Again, nouns and adjectives were more fluid, since they were the word classes that were not verbs, which had their own full suite of derivational and inflectional morphology.
Hold on, that's not the end of the story for the Tungusic form of the name. P-T does not have "c", since it's mainly Eskaleut in origin, and inverted Eskaleut's feature redundancy. So it wanted voiced stops and voiceless fricatives, for native words.
ReplyDeleteThe "cu" in the Wa-Dene-Yeniseian ethnonym, meaning "strong", cannot survive with "c". It must become either a voiced stop, "ʒ", or a voiceless fricative, "s". Recall that the early stage of Tungusic showed the "s" strategy, and only later was the "ʒ" strategy used -- and even then, mostly in initial position.
Well well well, whaddaya know, it all works out. "Jurchen", pronounced "Jušen", is the name of a fairly late group, who flourished close to 1000 AD. They show the "ʒ" strategy, but also "s", which means they must have based their name on a form like "cosen", which is what the Korean form of the name was -- with both "c" and "s", since it's of Wa-Dene-Yeniseian origin, and these consonants are allowed in the same word, unlike Tungusic.
By 1000 AD, Tungusic languages had both "s" and "ʒ", so borrowing a word with "s" was no problem. But they still did not like "c", so that required resolution. Since it's at a late stage of Tungusic, they opted for the late strategy, "ʒ", yiedling "Jušen".
However, Chinese records note an earlier group of Tungusic speakers in the same Manchurian / North Korean region, who were called "Susen" or "Sušen" -- suspciously close to "Jušen", the group who were their later successors in the region.
https://en.wikipedia.org/wiki/Sushen
Well, my Eskaleut approach to Tungusic solves that riddle, too -- since that was an earlier stage of Tungusic, they favored changing "c" to "s" instead of "ʒ".
There are two groups with the name "Sushen" in Chinese records. One is definitely covered by this cognate of Joseon model -- the alias of the Yilou, active in Manchuria during the early-to-mid 1st millennium AD.
But there was another such group who flourished much earlier, in the mid-to-late 1st millennium BC. That is not so far back that a Tungusic rendering of a Wa-Dene-Yeniseian word is out of the question... but it's cutting it close, as far as I see things. But maybe their name is cognate with "Joseon" as well. Certainly the culture from the mid-1st millennium AD is, and they have an early form of the name, where "c" -> "s", unlike the later form, where "c" -> "ʒ".
This also shows that the Jurchens did not borrow or preserve the name of the Sushen, since otherwise it would begin with "s" as well. Rather, they were rendering a name that began with "c", like the name from North Korea, "Joseon".
So, Tungusic treated the ethnonym as native or local, even though it was not Eskaleut. Still, it was local, albeit of Wa-Dene-Yeniseian origin. The main split in the Tungusic lexicon is Uralic vs. non-Uralic, to distinguish the foreign Altaic foundation from the local Eskaleut tradition.
ReplyDeleteEven though the ethnonym "Joseon" was not Eskaleut, it was also not Uralic. And it did not come as part of the Uralic / Altaic foundation, so it was treated as local and traditional from the Manchurian-ish region.
And hence, subjected to the same inversion of feature redundancy that Eskaleut words were -- voiced stops and voiceless fricatives, not a voiceless stop like "c".
There's another ethnonym of Yeniseian origin using most of these morphemes, although it's an exonym for the Evenki, used by the Ket -- "Tɨsdɛˀŋ", meaning "stone people", referring to the stony bank of the Yenisei River where the Evenki live, not a figurative use of "stone" to mean "strong".
ReplyDeleteHere, "Tɨs" reflects "cewç" from P-Y, where "ç" is a singulative suffix, not the nominalizer suffix "še". But still includes "cew" = "stone".
And "people" is "ɟenVŋ" in P-Y, including the plural / collective suffix "Vŋ", reflected as "dɛˀŋ" here.
They're not entirely the same group of morphemes, but both end in the collective suffix, and the main root is based on "cew" = "stone". So there's a precedent outside of the immediate Manchuria / Korea region, for a Yeniseian ethnonym with those two morphemes.
As it turns out, there was a similar language spoken in North Korea once upon a time, though more of a sister to Yeniseian -- Wa. Japonic does not mark singular vs. plural number, although there are collective suffixes like "re", but they are not cognate with P-Y "Vŋ". So, the name "Joseon" must derive from Wa, not Japonic.
Three Korean suffixes from Wa-Dene-Yeniseian origin. I know, what's more fascinating than suffixes? Well, they're necessary to show that the relationship is shared ancestry rather than borrowing, since nobody borrows suffixes or other bound morphemes.
ReplyDeleteThe most productive Korean nominalizer is "-m", which sometimes has a preceding epenthetic vowel that is back and unround like "o" or "u" (not "wo" or "wu"). It is used to derive nouns that are a bit more abstract, like "the act doing some verb" (action nominalizer), or attaching to a noun or adjective and meaning "-ness" -- the quality of being like that noun or adjective.
https://en.wiktionary.org/wiki/-%EC%9D%8C
This is cognate with the suffix "mi" in Japonic, which also means "-ness" and attaches to nouns and adjectives. We'll get to why the Japonic form has a vowel after "m", and why it's that vowel in a second. (I've already covered this before, when showing the Dene-Yeniseian cognates of Japonic words and morphemes).
https://en.wiktionary.org/wiki/%E3%81%BF#Etymology_4
Both of these are cognate with the Proto-Yeniseian suffix "Vŋʷ", which is an adjectival and action nominalizer. We're just focusing on its role in forming nouns.
https://en.wiktionary.org/wiki/Reconstruction:Proto-Yeniseian/-V%C5%8B%CA%B7
Exactly like in Korean, this suffix is used to express "the act of doing some verb", to make a verb into a noun, but an abstract kind of noun, not a concrete noun. E.g., attaching to the verb meaning "to fish with a rod", to express "the act of fishing with a rod". Why do this? If you want to say something like "The act of you fishing with a rod, is preventing us from doing our chores". A verb does not "prevent", a noun "prevents". So, "the act of doing some verb" becomes the subject of a sentence.
In the Japonic and Korean usage, about "-ness" attaching to adjectives or nouns, that is also a kind of abstract noun, although not the act of doing some verb. Still, an abstracting nominalizer.
Neither Korean nor P-J has "ŋʷ" in their phoneme inventory -- only "m" and "n", for nasals. Since "ŋʷ" has secondary rounding, it can be distinguished from "ŋ", which lacks rounding. Therefore, when resolving these problematic nasals, using only "m" and "n", assign "ŋʷ" to "m" since both have rounding ("m" has primary labial feature, its place of articulation), and "ŋ" to "n" since neither has rounding.
That means both Korean and P-J will change "ŋʷ" -> "m".
Korean sometimes uses the preceding vowel "o" or "u", and that is also the unpredictable vowel V from P-Y "Vŋʷ" -- typically back, non-low, and unround. So that establishes the cognate status of the Korean suffix with the Yeniseian suffix.
The P-J form requires a following dummy vowel, since it has obligatory CV syllable shape. It will choose based on frequency. In OJ, the most common "m" syllable is "mo", but that is already taken by the non-exclusive topic particle. To avoid homophony, the 2nd-most common "m" syllable was chosen -- "mi", QED.
But there's a wrinkle, which is that Korean only began using "m" for a nominalizer suffix fairly recently, beginning in the 14th C. Before then, the old nominalizers were "n" for a realis mood and "l" for an irrealis mood.
ReplyDeleteWe'll start with "n". See etymology 3 under Old Korean:
https://en.wiktionary.org/wiki/%E9%9A%B1#Korean
Recall that in the word for "sun; day", the Yeniseian and Na-Dene forms begin with "xʷ", the P-J form begins with "p", and the P-K form begins with "h" (velar fricative). This says that at the earliest stage of Korean, they did not preserve the secondary labialization, as P-J did, they simply deleted that feature, merging "xʷ" -> "x" ("h" in Korean studies).
Since we're looking at the oldest stage of Korean nominalizers, then they must have done the same sound change -- simply deleting the secondary labialization, changing "ŋʷ" -> "ŋ", and since that does not exist, applying the other nasal sound change "ŋ" -> "n". In other words, unless it was originally "m", all other Wa-Dene-Yeniseian nasals would have been merged into "n", at the earliest stage of Korean.
Well, that yields "n" as the nominalizer suffix, QED.
Why did Korean change from using "n" to "m", as a reflex of the same Wa-Dene-Yeniseian suffix "ŋʷ"? I can only speculate that this reflects the incorporation of literal Japonic-speaking cultures into the Korean speech community. Not just Wa, but specifically Japonic, which was spoken in South Korea, among the former Samhan states.
Japonic, from its earliest reconstructed stage P-J, preserved Wa's secondary labialization as a primary labial place of articulation, e.g. "ŋʷ" -> "m" and "xʷ" -> "p". Since that sound preference was so strong, they carried it over when they began adopting Korean, and the non-Japonic-speaking residents of Korea compromised with them, to incorporate them easily.
Why did it take until the 14th C for this to happen? Well, making Korean a standard national language requires national unification, and the first polity to unify all of Korea was Goryeo, which began around 900 AD and lasted until around 1400 AD. For the first time, the northern regions where a non-Japonic form of Wa was spoken before Korean emerged, and the southern regions where a Japonic form of Wa was spoken before Korean emerged, became part of a single national polity and culture.
It would take some time for the Japonic influences on Korean to show up, but it would be during this era, and toward the end. And so, by the end of the Goryeo era of national unification and cultural centralization / homogenization, Japonic influences entered Korean. One of them was treating Wa-Dene-Yeniseian secondary labialization as a primary labial consonant, rather than deleting it altogether, as they did during Proto-Korean and Old Korean.
I don't know a whole lot about the internal dynamics of Korean linguistic changes, so maybe there's something just about Korean, not an outside influence from Japonic. But it makes sense, so I'll go with this explanation, unless there's a purely Korean historical trajectory that led them to abandon "n" in favor of "m" for their nominalizer.
Last, the "l" nominalizer, for the irrealis mood. Phonetically, we can see that this is not cognate with the previous suffixes. The Old Korean form is transcribed as "lq", where "q" is a glottal stop. See etymology 2 under Old Korean:
ReplyDeletehttps://en.wiktionary.org/wiki/%E5%B0%B8#Korean
The form "lq" is a compound of 2 distinct suffixes, "l" for the irrealis mood or tense, and "q" for the nominalizer. Eventually, the "q" element was dropped, and in Modern Korean there's only "l", which conveys irrealis mood.
I think both of these have Yeniseian cognates, so make that 4 total suffixes from Korean that are Wa-Dene-Yeniseian. I forgot that "lq" was a fusion of 2 separate suffixes, for a single usage.
P-Y has a verb nominalizer "ɢa", which I assume is the cognate of the glottal stop "q" in Korean. It is clearly eroded in some way, since the glottal stop was not a common phoneme in Korean, and must have been some other phoneme earlier. Well, if it's eroded, it might have also lost a final unstressed vowel.
I think the "a" was lost from the Yeniseian form, and the voiced uvular stop "ɢ" was eroded to a voiced glottal stop, which is still easy to pronounce, and very far back in location.
During my demonstrations of the Dene-Yeniseian cognates of Japonic words, I noted that the uvular stops "q" and "ɢ" from Dene-Yeniseian corresponded to the null phoneme in P-J, like a word beginning with a vowel, when P-J strongly wants to begin words with a consonant. I said it would be natural for the change "q" or "ɢ" -> 0 to go through an intermediate stage where it was severely eroded, but still some kind of sound, like a glottal stop.
Korean seems to have fronted uvular stop "q" to velar stop "k", but it can't do that with the voiced counterpart "ɢ", since P-K and OK don't have voiced obstruents like "g". So perhaps the voiced uvular stop from Wa was mapped to the glottal stop in Korean.
That accounts for Old Korean nominalizer "q" being cognate with P-Y "ɢa".
As for Old Korean "l", the irrealis mood suffix that combined with "q" to form a nominalizer, I think the cognate is the P-Y instrument noun suffix "ɬ". It attaches to a verb or other descriptive root, and forms a tool that is related to that root. For example, the descriptive root "hook-shaped" receives this suffix, and forms the noun meaning "hook", the tool. It has a cognate "l" / "ɬ" in Na-Dene as well.
ReplyDeleteSo, "ɬ" is also a nominalizer, but used to form tool nouns. It's similar to the use of the gerund / present participle "-ing" in English tool words in a general sense, like "striking weapon" or "piercing tool", although we have overt words like "weapon" and "tool" as well.
In Old Korean, the "lq" suffix was used for general qualities, such as "loving" in the phrase "loving mother". That's what makes it irrealis -- prospective, potential, general, rather than some specific, definite, already happened or already-is-happening act of the verb "love".
It's not hypothetical as in "she can be loving on some occasions, and on other occasions she can be not-loving, it all depends." It just means in a general way that is not only confined to past or present actions -- and so, belonging to the unhappened future tense or prospective / potential mood.
Well, the same is true for tools. When you call a hammer a "striking" tool, you don't mean it has done the verb "strike" in the past, and in the present -- you mean it may also do so in the unhappened future. It has the potential to strike, in a general sense, it is capable of striking, and it probably will do so in the future as well as the past or present.
Using "striking" to refer to the kind of tool that a hammer is, doesn't mean "hypothetical" as in "maybe you could use it to strike, if you really wanted to use it that way". It just means it has the general potential to be used that way in the unhappened future, not only in past or present instances of striking that have definitely happened.
In Old Korean, "l" was not a pure mood suffix, it was 1 of 2 suffixes that fused together in "lq". So the fact that the P-Y form is a nominalizer suffix, nothing in particular to do with mood or tense, is irrelevant. It was not a mood or tense suffix in Korean either, more like a nominalizer that also carried meaning about mood or tense as well. It did not appear as part of a regular verbal paradigm to indicate irrealis mood. It only showed up in this nominalizer "lq".
Phonetically, we've already seen P-Y "ɬ" correspond to "l" in Korean, at least in coda position, which this suffix appears in, QED.
The P-J cognate would begin with "s", since it does not have "l". But that picture is muddied by the fact that P-Y also has the verb nominalizer "še", which also derives tool nouns. The dummy vowel for "s" is overwhelmingly "si". But that is also the expected reflex of "še". So, it's hard to tell which of these Yeniseian suffixes is cognate with the "si" in Japonic "kusi" = "comb; skewer", but it's one of them.
In Korean, it's easier to tell which Yeniseian and Na-Dene cognates they have, since Korean does have "l" in addition to "s".
Korean numeral suffix of Wa-Dene-Yeniseian origin. For ordinal numbers, i.e. listing items in a sequence like "first, second, third...", Old Korean and probably P-K has a suffix "-(ə)p" that attaches to the base or cardinal form of the number (like "one, two, three...").
ReplyDeletehttps://en.wiktionary.org/wiki/Appendix:Historical_Koreanic_numerals
Proto-Yeniseian has a multiplicative numeral suffix, "pa". This is for how many "times" ("once, twice, thrice...") or how many "-fold" or "-way" something is.
OK has lots of suffixes that are clusters of consonants, some of which must have had vowels afterward, but lost them due to being unstressed. E.g. "lq" that we saw before, "ms" (necessity or "it is fitting"), and others.
So it's not unreasonable to assume that earlier "pa" lost its unstressed final vowel and became simply "p" by the OK stage. Very straightforward phonetic derivation.
Semantically, ordinals and multiplicatives are similar, and different from cardinals, in assuming a subsets within a whole set.
ReplyDeleteTo use the notion of sets, there's a great big whole set that includes all the members, for some frame of reference in the conversation. Say, the entire army that is mobilized. At the highest level, there is 1 set -- the whole army.
At the lowest level, below which no further sub-division can happen, is individual soldiers -- there's no way to sub-divide a single soldier, and still consider them part of the mobilized army. Counting the number of soldiers in the entire army involves dividing the whole set into a bunch of non-overlapping subsets, which collectively exhaust the whole set (i.e., they account for all the members at the lowest level). Each of these "subsets" is of size 1, namely an individual soldier. I don't like using "subset" to refer to a set of size 1, when the term "member" is better, to show that no further sub-division can occur.
Cardinal numbers only involve the highest and lowest levels of the set in the frame of reference -- the whole set, and the individual members in it. How many soldiers are there in the army? That answer is a cardinal number, like 1000.
Counting the size of a whole set is the most pressing need for numerals in human language, so every language has cardinals.
But sometimes more fancy things are needed, that involve a level of structure between the whole set and the individual members -- which is what I'm calling "subset" rather than "whole set" or "members".
Multiplicative numerals are one natural way of dividing the whole set into subsets, above the level of members. For example, the whole army is mobilized -- but how many "prongs" is their attack? A one-prong attack? A two-prong attack? A three-prong attack? Etc. The answer is a multiplicative numeral.
ReplyDeleteIf there are 2 prongs, the numeral is multiplicative -- "two-fold" or "two-way" or "two-prong" attack, not just "2 attacks" or using the numeral to modify the counter, like "two folds" or "two ways". They combine into a complex numeral, "two-fold" or "two-way" or "two-prong".
Or you may speak of how many ways you will attack the enemy -- militarily, economically, and culturally, for a "three-fold" / "three-way" / "three-prong" attack. The whole set is the entire multi-faceted attack, the individual members are specific policies or actions, but these lowest-level actions are grouped into subsets like "military attack", "economic attack", and "cultural attack". These subsets are non-overlapping, and collectively exhaust the whole set.
Aren't these like division, rather than multiplication? Well, the idea is that you're taking a single whole set and multiplying it somehow -- instead of just 1 attack from the whole army, you divide it into 2 prongs, and each of those prongs is strong enough on its own, so that you double the strength of a single attack from an undivided army. Likewise with attacking on 3 fronts like military, economy, and culture.
Ordinals are somewhat different, but still involve subsets or sub-structure within the whole set. You aren't just tallying them up as members of an undifferentiated whole set. You're ordering them with respect to one another -- from least to greatest in a sequence.
ReplyDeleteThis is the only difference from multiplicative -- ordinals involve a sequence, whereas multiplicative involves a simultaneous or parallel happening. A "three-prong" attack happens at the same time, not one prong first, another prong second, and a final prong third. A combined military, economic, and cultural attack happens simultaneously -- there's no way to order those attacks as first, second, and third.
And yet, ordering a sequence still involves relating the members of the whole set to each other, instead of considering them members of an undifferentiated blob of the whole set. Here, you are dividing the whole set into "subsets" of size 1, which usually I'd just call "members", but comparisons and relations involve sets rather than members, so grudgingly I'll call these subsets of size 1. After dividing the whole set into non-overlapping subsets of size 1, that collectively exhaust the whole set, they are ordered in some kind of sequence.
When you tally up the individual members to count the whole set, you're not comparing them, relating them to each other, etc., so you aren't cognitively grouping them into subsets of size 1. There's no sub-structure to the whole set in simple enumeration. But for ordering a sequence, there is sub-structure.
In fact, ordering a sequence can apply to subsets of size greater than 1, like "the first 10 days of a month", then "the second 10 days of a month", and "the third 10 days of a month". The whole set, the entire frame of reference, is the month, whose lowest-level is individual days (members). But these members can be grouped into subsets like "week" or "10 days", which are non-overlapping subsets that collectively exhaust the whole month, account for all the lowest-level members, but are each of size greater than 1.
In the simple enumeration of a whole set, there's no semantic reason to sub-divide the whole into subsets of size greater than 1. So the notion of subset or sub-structure does not apply. Relations and comparisons happen between sets, not members of a single set. If you're comparing individuals, you're really treating them as subsets of size 1, and that is not what enumeration of a whole set is about. There's no relations or comparisons there.
And so, that's why the same phonetic form may be used for ordinal numeral suffixes in one branch of a language family, and multiplicative numeral suffixes in another branch, making them cognates despite the slight differences about sequential / least-to-greatest vs. simultaneous / all-are-equal. They're both different from cardinals, which are only used for the basic task of enumerating a whole set, with no relations or comparisons, and no sub-divisions.
Multiplicative numerals can shade into a sequential meaning, when you say how many times some event has happened, like "the army has attacked once / twice / eight times". Doesn't that imply a sequential ordering of the attacks -- first attack, second attack, ..., eighth attack? Literally, of course, there is no world where the army has attacked eight times, without having already attacked 7 times, 6 times, ... twice, and once.
ReplyDeleteBut that's not how the human mind is treating the attacks, cognitively or semantically. They're using a single interval of time, like "within the past week", and there is no flow of time or sequence possible. The question is, within this undivided, unsequenced block of time, how many attacks have there been? In that sense, the attacks have been simultaneous and parallel, not sequential. "There has been a 7-fold attack within the past week", even if strictly speaking, there was 1 attack per day for 7 days in sequence.
Technically, there are no discrete chunks of time -- it's all continuous and divisible into smaller and smaller intervals. But the human mind, and social converations, do not agree with that nerd approach to time. Sometimes, the building-blocks of time are days, sometimes weeks, sometimes months, years, decades, centuries, millennia, eons, and so on and so forth.
Within a single block of time, events are considered parallel and simultaneous -- same block of time, not necessarily a block that equals an "instant" or "moment". Would anyone really disagree that "there was a simultaneous 2-pronged attack" if one prong attacked 90 seconds before the other prong? It depends on the context -- maybe in some video game, those 90 seconds are a lifetime, and give the target enough time to wiggle their way out of the 2-prong attack. But in other contexts, like real-life battlefields, 90 seconds is tantamount to simultaneous.
Same thing with "there has been a 3-fold increase in the price of beef since Trump took office", or "the price of beef has tripled", or "beef costs thrice / three times as much as it used to". Those numerals are multiplicative, not ordinal. And yet they seem to be involved in a timed sequence -- how can there be a 3-fold increase, without an earlier 2-fold increase, 1.5-fold increase, etc.?
Well, technically yes, but human beings are not so technical. It depends on the context of the conversation. In this context, there is a single interval of time -- the arrival of the Trump administration. And within that single unsequenced block of time, there have been a parallel and simultaneous set of forces, which have driven up the price of beef, to the magnitude or intensity of "3 times". You're still not enumerating a set, and there's no sequence or ordering. More like the magnitude of a simultaneous / parallel group of happenings.
Another Korean ordinal numeral suffix possibly of Wa-Dene-Yeniseian origin. Old Korean used another series of letters for an ordinal numeral suffix:
ReplyDeletehttps://en.wiktionary.org/wiki/%E9%9F%B3%E7%AC%AC%E5%8F%B1
This has 2 phonograms, "m" at the start and "s" at the end, and a logogram in the middle, which is the Chinese character for "sequence, number". But, there's no indication how that middle letter was pronounced. It's reconstructed for convenience as "CAHI", since that's what the Middle Korean reflex is, and MK did use it for ordinal numerals. But this is just a convenience with no solid evidence to back it up.
There's no way that the ordinal suffix was "m + CHAI + s". It's way too many sounds, but more importantly, too many morphemes -- 3 in a chain. This is *in addition to* the cardinal numeral base that it's stuck onto! Parsimony and cross-linguistic evidence says this suffix should be only 1 morpheme, if possible, and only model it as 2 if necessary, and as 3 only if necessary after that.
So we turn to Proto-Yeniseian and find the ordinal numeral suffix is "wasVŋʷ" (we saw the multiplicative suffix before, this is the ordinal one). As it happens, this is morphologically complex, with the "Vŋʷ" being a generic adjectival suffix, and "was" being the specific ordinal numeral morpheme.
Recall that Korean lacks "w" (part of their Eskaleut accent, while speaking Wa), so the most natural and faithful way to render it in Korean is "m" -- voiced, bilabial, resonant, only difference is oral vs. nasal airflow. Better than the alternatives "p" or "y". That's the 1st C, and the 2nd C is identical, "s".
The only assumption, then, is that there was a single vowel "a" between them, which was represented by the Chinese character for "sequence, number". Well, the Middle Chinese pronunciation of that character is "dejH", with nuclear vowel "e", but the Koreans heard it with an "a", since the MK form of the Sino-Korean borrowing is "cahi". So assuming this character was read as simply "a", in this particular context, is supported (though it's not the only possibility compatible with the evidence).
Hard to call this "case closed", since little is known about how that middle character was pronounced, but it still works phonetically, and semantically.
And more importantly, it reduces the morphological complexity of this suffix from 3 morphemes all the way down to 1.
Any superior alternative to my theory here must provide a better fit to the phonetic evidence, without proliferating the number of morphemes. And must be somewhat geographically plausible, or shared by some plausible distant relative.
I doubt any such superior alternatives will arise, so I'll consider this "case closed enough". ^_^
Important addendum: perhaps "mas", my proposed Proto-Korean form of this morpheme, is the original form in Wa-Dene-Yeniseian, and Yeniseian altered "m" -> "w", since Yeniseian bans nasals in initial position.
ReplyDeleteThis is a suffix, not a prefix, so why worry about word-initial rules?
Well, it could be a fossilized standalone word, meaning "number, sequence, etc.", just like how we say "fold" or "way" or "times" for multiplicative numerals. It may have become used solely as an ordinal suffix, but originally it could have been a standalone word.
It sure looks like one -- P-Y has standard word-shape CVC, including others with "wVs" like "wes" = "wide, spread; wild; open space; sky, sky deity".
And "was" is followed by the adjectival suffix "Vŋʷ" in full suffix "wasVŋʷ", and adjectival affixes usually attach to nouns, which are standalone words.
If, as seems likely, the "was" in "wasVŋʷ" was an earlier standalone word, it cannot begin with a nasal in P-Y, so it was altered in the most faithful way, "m" -> "w".
Worth keeping this possibility in mind. But they're still cognates and from a shared proto-ancestor.
The final answer, from Japonic, says maybe it was "was" originally, not "mas". Yeniseian allows "w" initially, but not "m". Korean allows "m", but totally lacks "w". Well, glorious Proto-Japonic has both consonants, and allows both word-initially. What does P-J say?
ReplyDeleteWell, there are absolutely no words beginning with "mas" in P-J (listed at Wiktionary, anyway). Not good for the "m" theory.
But "was" has 3 words, one of which refers to time and sequences, namely "wasai" = "early ripening rice". This is not a compound, about "early" and "rice", only about "early". In its combining form "wasa", it means "early ripening" and can apply to any crop, not just rice, and there's no morpheme referring to "crop" in "wasa".
https://en.wiktionary.org/wiki/%E6%97%A9%E7%A8%B2#Etymology_2
Notice that this is not just about the flow of time, but ordering things in a sequence -- this type of rice ripens *before* that type of rice. This is a comparison or relation between two elements of a set, based on a sequence in time. So this directly relates to ordinal numerals, not just cardinals.
Another word, "wasi" = "eagle", nothing to do with numbers or sequences.
A third, "wasuru" = "to forget", with what appears to be the dummy verb suffix "ru", and so whose semantic root is "wasu", which also has what appears to be the obligatory verb ending "u".
How does "forget" relate semantically to "sequence" or "early"? IDK, but maybe about "moving forward in the sequence of time" so that the past is out of sight, out of mind. Maybe like the English expression "to fast-forward", based on audio/video tech, but meaning "to skip over the parts you want to forget".
That settles it -- the original form was "was". In P-J, obligatory word-shape CVCV requires a final dummy vowel, and based on frequency, the most common "s" syllable in OJ is "si", followed by "sa", both of which are used in the P-J reconstruction "wasai", QED!
You use your verbal lobe to play games like Wordle and crosswords.
ReplyDeleteI use my verbal lobe to unlock the deepest linguistic mysteries of prehistory.
We are not the same.
(I play and win Wordle every night, too -- along with Spanish Wordle. ^_^)
I don't mean that in some inane status contest, "I have a higher verbal IQ" (which I probably do).
ReplyDeleteI mean, you are a slave to base temporal forces, like greed, status-striving, and so on. Everything has to be some zero-sum game, some pointless contest like comparing Wordle stats. And so, nothing of your life will outlast you, and you contributed nothing -- just boosting the engagement metrics of some media website like nytimes.com.
I am using my gifts to heed a higher eternal supernatural calling, whether I want to or not. And I don't mind it, actually -- it gives my life meaning. Wandering in search of secret passageways, following their course, discovering hidden treasure, and hauling it back to normie-dom, to share with people today, and even future alien civilizations from the far side of the galaxy.
Life is too short to waste on pointless games. They're fine as a little amusement, but for most game-players, everything in their lives is a game, not like they compartmentalize a little game-playing time, separate from heeding their higher calling and fulfilling their duties to the supernatural.
Some people don't have gifts to use in a higher calling, that's fine. But too many of those who do have them, waste them. Especially in this collapsing shithole stage of our imperial lifespan, where everyone's trying to maximize greed, and higher callings rarely get you rich.
But they are punished for wasting their natural talent on pointless games -- they do not get to slide the dial on some fake trade-off scale, like "talent used to perfect something mundane, which nets more wealth and status, but less eternal supernatural contribution". No, you don't even perfect the mundane bullshit you pursue instead of a higher calling -- all those dumbass "quants" in finance rapidly blew up the entire global financial system.
They're not applying their talents to perfect something mundane or profane -- they're using it just to scoop up as much wealth and status as they can, even though they end up destroying the mundane domain they have hijacked control over.
But I rest comfortably knowing they will burn in Hell for their choices, and I will be rewarded for mine.
I don't even think of myself as an ubermensch or anything -- more like they have sunken themselves so low, that doing the right and proper thing only seems super-duper in comparison.
If, like me, you're an American who follows "no institutional religion", cuz we've severed our roots to the Old World, and few of us live in Mormon-land, where the new American religion thrives, then just think of those future alien civilizations from the far side of the galaxy, as the gods who will judge your life once it's over.
ReplyDeleteThey're there, and they'll find out about us eventually. And they will pass some kind of judgement on us, collectively and individually.
How are you going to be remembered in their records? Make choices like that's your ultimate guide -- making the list of "one of the noble Earthlings" in the judgement of advanced alien civilizations from the far side of the galaxy in the distant future.
It's as close enough to the Dark Age Christian guidelines as we Americans can manage in our de-Christianized New World society, whose myths are instead about the prehistoric (cavemen and dinosaurs) and the futuristic, specifically about civilizations in distant regions of outer space.
The Japonic word for "eagle" ("wasi") may relate to ordering, sequence, and ranks, after all. My main shower thought was that word not fitting in with the rest. Maybe it does.
ReplyDeleteIf "wasai" can refer to rice or other crops that are "early ripening", that is that come before others in a sequence, maybe the Japonic speakers thought the same about the eagle -- as coming before other bird species, or any other animal species, in a sequence.
Not a sequence based on maturation, of course, but orders and sequences and ranks can use any dimension they want to impose their ranking.
So, the eagle may be "first" or "foremost" among the animals, or among the birds, or even among the birds of prey. Earliest in a sequence is tantamount to "first" for ordinal numerals.
It could also refer to the eagle as a mythological progenitor, totem animal, or other kind of spirit ancestor. That uses a temporal sequence -- the earliest member in a lineage whose latest members are those who claim descent from the eagle today.
That is paralleled by the mythological founder of Thebes, Cadmus, having the Semitic tri-consonantal root for "to precede, come before" in his name -- "q - d - m". He came before anyone else in Thebes, since he founded it. That's another sequence word for a mythological figure's name.
Some clans in Japan do claim descent from a mythological eagle, or the eagle as a whole species, such as the Awa-Inbe clan. But the Yamato culture as a whole did not claim descent from an eagle, nor was an eagle involved in the Wa / Yamato creation myth, so I doubt "wasi" was chosen to mean "progenitor, ancestor".
More likely, it meant "first, foremost" in importance or rank. There is a god in the Nihon Shoki, one of the earliest writings in Japanese, called Ame-no-Pi-Wasi-no-Kami, which means "Heavenly Sun-Eagle God". He is part of the effort to coax the sun goddess, Amaterasu, out from hiding in a cave, so sunlight can be restored to the world. He plays a guitar to lure her out with song, and an eagle lands on his guitar while playing -- an auspicious omen.
Japan also has plenty of "Otori" or "great bird" shrines, where the eagle is the central species.
So, the eagle is not their central god or creator animal or even totem animal or spirit guide animal (which is the wolf, as in other Eastern Steppe cultures, like the early Turks and Mongols). And therefore, the sense of "ancestor, progenitor" is not the right origin of "wasi" as the name of the eagle. Rather, it is the eagle's "foremost" rank among the birds of prey, birds as a whole, and among most animals as a whole.
And so, it *is* cognate with the other "was" words in P-J, for "early ripening" and "to forget", as well as the Korean and Yeniseian ordinal numeral suffixes.
Ahhh.... glad I wrapped up that loose end, it was really bugging me! And apologies to the eagle spirit, for so blithely dismissing any possible relevance of eagles to sequences, orders, and ranks...
Korean imperative suffix of Wa-Dene-Yeniseian origin. In OK, it's "sye" (pronounced "se"):
ReplyDeletehttps://en.wiktionary.org/wiki/-%E5%8F%A4%E5%8F%AA%E8%B3%9C%E7%AB%8B#Old_Korean
In Proto-Yeniseian, the imperative prefix is "ɟe". As a reminder, here's the long list of P-Y roots I'm consulting:
https://en.wiktionary.org/w/index.php?title=User:Ama%C3%A7s%C4%B1zBirKi%C5%9Fi/qfa-yen-pro-vocab&oldid=86584623
"ɟ" is the voiced version of "c", but the following palatalizing vowel "e" can soften it further into a sibilant, as in the Kott reflex "č" / "š". That is also what happened in P-K, where instead of simply de-voicing to "c", it also assibilated to "š", and since there's only 1 sibilant in P-K, shifted slightly forward to alveolar "s".
Unlike the Yeniseian reflexes, which deleted the vowel, P-K preserved it exactly, yielding "se" ("sye" in Korean studies), QED.
This is a very crucial link in the Dene-Yeniseian model, since the P-Y prefix is cognate with the P-Na-Dene valency-reducing classifier "də", one of the hallmarks of the Na-Dene languages.
https://en.wiktionary.org/wiki/Reconstruction:Proto-Na-Dene/d%C9%99-
In my work on the Dene-Yeniseian origins of Japonic, I argued that the passive verb suffix "re" is another cognate of the Dene-Yeniseian forms. Like the ND form, it is valency-reducing. Phonetically, it's a voiced alveolar consonant followed by "e".
Now I've found a Korean cognate as well, with the same function as in Yeniseian (imperative).
Korean directional and instrumental suffix of Wa-Dene-Yeniseian origin. It's unusual, but not unprecedented, to have the same case marker for a directional (to, toward) case and an instrumental (using, by means of) case. Just within the region, Ainu does this as well, with the single marker "ari".
ReplyDeleteIn OK, it's "lwu", but earlier in P-K it's reconstructed, e.g. by Vovin, as "lVpV". There are signs it used to have a labial stop consonant that weakened to fricative "b" between vowels, then to "w", which is where it ended up.
https://en.wiktionary.org/wiki/%E7%95%99#Old_Korean
It looks complex, with "lV" and "pV" being separate suffixes. If it were a single word, the standard shape would be CVC. And since the vowels weakened / deleted, they may have been dummy / epenthetic vowels to begin with, to prevent an illegal coda cluster.
So I'll only look at the "l" and "p", ignoring the vowels.
Semantically, it is used for 2 cases that are usually distinct, which suggests it is a fusion of two separate case markers from an earlier stage.
Proto-Yeniseian has an instrument-forming nominalizer suffix "ɬ", which we've already seen, and is cognate with instrument affixes in Na-Dene, either "ɬ" or "l". And possibly the "si" in Japonic "kusi" = "comb; skewer". And we've already seen it in Korean as "l".
Well, why not also use it for the instrumental case? It already has to do with instruments, tools, ways of producing outcomes or achieving goals, the means to an end. It's semantically identical, just being deployed for a slightly different grammatical purpose, marking a case for nouns that are the means by which some action happened.
As with the instrument noun-forming suffix, the only sound change is "ɬ" -> "l", very straightforward.
Proto-Yeniseian also has a morpheme "p", a thematic verbal prefix, that says the verb is spatial, involving a shape, or an area. So, of a spatial nature. That is precisely the same semantic field that directional terms like "to, toward" belong to -- spatial relations.
On the Na-Dene side, the Dene-Yeniseian model proposes Athabaskan "qu", a prefix relating to "flat broad area", as the cognate. Na-Dene lacks labial place of articulation, so "p" would have to be realized as a consonant elsewhere, with secondary labialization. Well, Na-Dene *does* have labialized velars and uvulars, so perhaps "qu" used to be "qʷ", whose labialization vocalized into "u". Then that "qʷ" would correspond to "p" in Yeniseian -- and Korean, as it turns out!
Phonetically, they're the same, "p". Only difference is that Korean is primarily suffixing, so it's in a suffix rather than prefix, as in the highly prefixing Yeniseian languages.
For whatever reason, these two separate case markers fused into a single marker, but retained their 2 distinct meanings or usages.
Semantically, the idea is that an instrument is equated with its product or outcome, or a means with its end, or a path with its destination. Did Korean speakers find this equation simply irresistible, or did these two case markers collide for some other reason? IDK, but they did.
The important point is that both of them are cognate with Dene-Yeniseian case markers, more evidence for Korean having a Wa origin, just like Japonic.
Japonic also has a cognate of that spatial term -- "pe" in P-J, later weakened ultimately to "e" in MJ. As a directional term, it means "to, toward", just like in (Old) Korean. And it has a separate usage as a noun meaning "place, general area, vicinity", similar to the Yeniseian and Na-Dene cognates -- a generic spatial term.
ReplyDeletehttps://en.wiktionary.org/wiki/%E3%81%B8#Etymology_2
https://en.wiktionary.org/wiki/%E8%BE%BA#Etymology_2_2
The consonant is the same, "p", but the vowel is likely not a dummy, since "pe" is one of the *least* common "p" syllables in OJ. The most common is "pa", but that's already taken by the exclusive topic particle. And "pu" has a typical verb ending, which this is not. However, "pi" is available and common, yet it was not chosen.
Possibly that means the Korean vowel in "pV" was close to "e", and then became rounded due to the rounding of the preceding "p" / fricative "b" / "w".
There's another directional term in Japonic that can mean "to, toward" -- "ni" -- but it emphasizes the destination, while "pe" emphasizes the path or journey. That makes "pe" exactly like the Korean "pV", which is linked to another marker, "lV", for the instrumental case -- the means used to reach a destination, not the destination itself.
One great big happy Wa-Dene-Yeniseian family. ^_^
Korean and Japonic accusative particles of Wa-Dene-Yeniseian origin. Proto-Yeniseian did not mark cases on the direct objects, rather they expressed this by means of agreement prefixes to the verb. This is a difference between languages that are head-marking (Yeniseian) vs. dependent-marking (Korean and Japonic). We don't need to go further into that difference, just to note that originally, direct objects were "marked" using verbal agreement prefixes.
ReplyDeleteIn general, the object of a verb is not a 1st or 2nd person, but a 3rd person ("it"). Only in special cases are "I" or "you' the object of a verb. So, when Korean and Japonic switched from head-marking to dependent-marking, they adapted the 3rd-person verbal agreement prefixes from their older Wa language, which was sister to Yeniseian and Na-Dene.
In P-Y, there were 2 different noun classes, animate and inanimate, and the animate came in 2 genders (for verb agreement), male and female.
The 3rd-person inanimate agreement prefix is "w", and this is the one that P-J adapted for its accusative marker, which is "wo". It needs a dummy vowel to fit obligatory CV syllable shape, and the most common "w" syllable in OJ is "wo", so that's what it took.
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/wo#Etymology_3
The 3rd-person animate agreement prefixes are "d" for male and "r" for female, which are fairly similar sounds. Korean used "r", probably after merging "d" into it, as a genderless animate marker.
https://en.wiktionary.org/wiki/%E4%B9%99#Old_Korean
As I said earlier, I'm guessing Korean "r" came from the voiced interdental fricative from Eskaleut, since they based their character for it on a Chinese word that ended in something like "t" or a fricative "t", that to their ears shaded into "r".
Regardless of that, Korean allows word-final C's, so they didn't need a dummy vowel like Japonic, and it remained "r" (then in Middle Korean, merged into "l", where it is in Modern Korean as well).
So, while on the surface the Korean and Japonic accusative particles appear totally unrelated -- "wo" and "r" -- they are in fact derived from a common ancestor, namely the 3rd-person verb agreement prefixes from a Wa-Dene-Yeniseian language. They only look different cuz Korean used the animate form and Japonic used the inanimate form.
Dene-Yeniseian unlocks yet another mystery of Korean and Japonic!
Wa-Dene-Yeniseian origin of Korean and Japonic words relating to subdivided space & time, sequences, "again", and "each, every". LOTS to go through here.
ReplyDeleteWe'll start with some compound forms, with 2 morphemes, since that's what led me down the rabbit-hole, and then we'll see about the single morphemes afterward.
Old Korean "mata" means "each, every", a distributive particle that divides a whole into parts and says some predicate applies to every part, while also viewing these parts as belonging to a larger whole.
https://en.wiktionary.org/wiki/%E4%BA%87%E5%A6%82
This is cognate with P-J "mata" = "again", "and, also", "another (one)".
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/mata
https://en.wiktionary.org/wiki/%E3%81%BE%E3%81%9F#Japanese
This pair of words has already been proposed as a cognate pair between Korean and Japonic, and the semantic link is that to express "again, also, another", you first need to divide time or a whole noun into intervals, subdivisions, or sections. Then you say some predicate applies to more than one of those subsets.
The only subtle difference is that the distributive particle implies completeness -- the predicate applies to every single one of the subsets -- while the "again, also, another" particle is indefinite, maybe all of them, maybe only some of them.
However, we'll see some Japonic forms that do imply completeness or "ever / every". So, they originally meant an exhaustive list of parts of a whole, and the predicate applied to "each, every" subset, not just "some". In Japonic, there was a later semantic shift that weakened this condition, where it could be used for indefinite examples like "again" and "another", which does not imply "always" or "every".
Could these be a borrowing from one language into the other? I doubt it, since this class of words is hard to borrow. They're abstract, and impossible to define with a single synonym, requiring you to explain it at length using them in a full sentence. Plus, there's no need to borrow something for the easily understandable usage like "and" -- you already have a word for that in your own language, why borrow it?
And when you do borrow it, it's always in a larger word that does make sense, and you may have a reason to borrow it, since you may lack such a word -- or you want a more exotic, elevated tone, not your ordinary dull word for it. For example, English has never borrowed any form of Latin "semper" = "always", but it did borrow the adjective containing it -- "sempiternal" (among SAT-word users, anyway), meaning "forever-lasting".
If it were a borrowing, the direction would be Japonic -> Korean, since it's in P-J, which is much older than P-K. Also, phonologically it looks more Japonic, of shape CVCV, whereas a distinctly Korean suffix would look like CVC.
I doubt it's a borrowing, given the difficulty of borrowing such a word. However, its precise phonetic form in Korean may have been influenced by the Peninsular Japonic speakers they incorporated during the 1st millennium AD, by the time the OK word is recorded. Ultimately, though both are inherited from Wa, sister to Na-Dene and Yeniseian.
That semantic shift is the obvious direction, BTW, from "all" -> "most, some". Just like in English, we use "all" to mean "some or most, but not literally 100%". Same with "every". It's easier to weaken "100%" down to "90%" or so, treating it as a case of poetic license.
ReplyDeleteIt's hard to raise "some" or "most" to mean "100%".
Why? Cuz there's a razor-thin barrier to cross from 100% to less than 100%, but there's a yawning chasm between 25% and 100%, or even 60% and 100%. The former is much easier to cross.
And that's why we do say "everyone here is a dude..." when we literally mean "a majority or sizeable minority are dudes", when complaining about a party's demographics.
But we never say, "some of the people here are dudes..." when it's literally 100% male, or close to 100% male.
Before looking at the 2 separate parts of "mata", I'll note some other cognates within Japonic, since Wiktionary does not note these connections.
ReplyDeleteFirst is "matu" = "to wait". This is semantically related since it's about an interval between two points in time -- now and the point when some event happens in the future. Like "to wait for the bus to arrive".
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/mat-
It prefers being used in situations where there's a break in an event, and a later resumption of the event -- which is related to "again".
Like the common imperative saying "mate" -- usually that means the listener is already doing X, and the speaker is asking them to briefly pause X, and only resume doing X "again" at a later time. Same usage as in English imperative "wait", meaning "hit pause" or "time-out", indicating some action is already unfolding, and you're asking for a break, before the action resumes "again".
Second is OJ "matasi" = "complete, perfect". This is simply "mata" with the adjective suffix "si". This shows that "mata" originally meant "100%", universal, always, etc., and only later weakened to "again; also, another". In this compound form, with suffix "si", "mata" survives in its sense of "100%".
https://en.wiktionary.org/wiki/%E5%85%A8%E3%81%84#Japanese
Third is "matu" = "pine tree".
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/matu
Two of the proposed etymologies at least use the cognates, "matu" = "to wait" and "matasi" = "complete", but they don't understand the relevance to "pine tree". It's not that a pine tree has to wait for anything, and it's not that a pine tree spends the complete winter in snow. Rather, the root means "ever, always" in the sense of "100%" -- it is, as we say in English, an "evergreen" kind of tree. It always has green growths on its branches, rather than the kind of tree that loses its leaves in the fall and has to regrow them in the spring.
https://en.wiktionary.org/wiki/%E6%9D%BE#Japanese
Since the final V is "u", it looks like a verb, like "to always be" or "to always have", and that the predicate "green" was deleted from a longer phrase since it was tacitly understood was quality was "ever" or "always" about a pine tree -- its foliage. And that's what distinguished pine trees from maple trees -- not the presence of foliage or the color green, but whether they always had foliage, or periodically lost and regrew their foliage.
Two of these 3 are in P-J, and "matasi" is in OJ, which testifies to the very old depth of the compound particle "mata". It was present in P-J, and it was so productive that several derived forms were already present in P-J as well. It must be very old indeed.
About "wait" vs. "stop". I already said "wait" implies some action is already happening, and you're asking for a pause in that action, before it resumes unfolding "again".
ReplyDeleteIf the action were already unfolding, and you don't want it to resume "again" later, you just say "stop". You can always tell when your teacher is compromising vs. putting her foot down, when she escalates from "wait...!" or "hold on...!" (allowing resumption), to "stop!" or "enough already!"
There's no implied "again" when you say "stop".
Is there something "again" about an action that is not already happening, but will or may happen in the future? Like "to wait for the bus to arrive". That doesn't mean that it's here already, and you're just asking it to pause being-here, but allowing it to be-here "again" later.
You use "wait" relating to the bus and "again", referring to its being in motion, like when it drives past you, or it doesn't see you cuz you're a block away, and you're running after it screaming "wait!" It's in motion, you ask for it to pause being in motion so you can board, while allowing for it to resume being in motion, once you're on board.
But that's an imperative usage. What about "I'm waiting for the bus to arrive" -- it hasn't already arrived. You're not even implying that it has arrived before, and it will "again", according to its repeating schedule. You can also say you're waiting for something that has never happened before, like "I'm waiting for them to break up before I tell her how I feel about her". They got together, they haven't broken up yet, so there's no repeating schedule here. You're waiting for something to happen for the first time.
Maybe there's something "again" about it, though -- in your imagination, it has already happened. You can see, feel, and sense it happening, and you're already acting accordingly, in your mind. BTW, this can be true for desired or undesired events -- "I'm just waiting for my boss to notice I skipped work yesterday and chew me out".
The "again" or second time this event occurs, is when it happens in real life, not in your quasi-real imagination-world.
You've already had thoughts, feelings, and at least made plans for behavior, in your imagination. So, when it happens in reality, it'll be happening "again", and those thoughts and feelings will rush through your mind all over again, and you'll be able to implement your behavioral plan, having already experienced the event before, in a sense. It's like your imagination is a rehearsal for reality.
And so, "waiting" for something to happen does, in a very keen way, have a feeling of "again". It already has happened, 80%, in your mind -- the other 20% can be completed when it happens in reality, for the second time.
BTW, girls think dreams are real, which supports this model of "waiting" implying "again", based on the common belief that "it happened in the imagination" qualifies as "it already happened for the first time".
ReplyDeleteIf a girl finds out you dreamed about another girl, even in a totally non-sexual or non-romantic way -- that's cheating, in her view.
And when she is with some guy in her dreams, she carves another notch in her bedpost, as though it were a bona fide sexual conquest of hers -- which she will absolutely brag about to her girl friends in the exact same way as she would about a real-life conquest.
Boys, when they're young, think dreams are real, too. Girls simply stay more childlike in their brain development -- and it's what makes women so much cuter and more endearing than men. ^_^
So, there is a widespread treatment of the imagination as real, or real-enough, for its happenings to qualify as "the first time", and when they happen in reality, they happen "again", for the second time.
After all, dreams are not the opposite of reality, a kind of anti-reality or non-reality. They seem very very real, so real you believe it's real while you're dreaming. Or even imagining, while you're awake during the day -- it can be VERY real, real-enough, sometimes even MORE real than reality, like when it happens in reality and you're disappointed cuz it was so much better in your imagination.
That's not about a non-happening vs. an actual-happening -- it's about one kind of happening vs. another kind of happening. It was so much better the first time, and the second time was a let-down, like the spark is already sputtering out.
Conversely, when it happens for the first time in the imagination and it's only so-so, then when it happens for the second time in reality and it's amazing, it feels like an "again", and like a growth or improvement. An improvement implies it's happened twice, once before and again now. Improvement is a cause for joy and optimism! That's why it feels so much better than just being a great happening -- it's an improvement over the first time it happened, in your mind.
So, "expectation vs. reality" should not be seen as "non-reality vs. reality", but about "mental happening vs. IRL happening".
Also, you don't say you're "waiting" for something to happen, that you have not already thought about and imagined happening in your mind, for the first time.
ReplyDeleteSome events have not happened yet, possibly have never happened before (the first break-up of a new couple). And they will / may happen in the future. But you aren't "waiting" for them to happen, unless you've already experienced it happening in your mind, for the first time.
When it does happen, without you having already experienced it in your imagination, it will feel like an impromptu or guerrilla happening -- for the first time, with no prior rehearsal.
"I wasn't waiting for the bus to arrive, but now that it's here, I might as well use it instead of continue waiting for my brother, according to our plans, since he still hasn't arrived..."
No prior thought or imagination = no first time, so the actual-happening is the first time, nothing "again" about it, and you explicitly say you were NOT waiting for this to happen.
If only English were as cool as Nihongo, this semantic link between "wait" and "again" would be morphologically transparent. But I'm sure Japanese people say the same thing about some cool facet of English that they lack in their language.
ReplyDeleteThe morphology is always more transparent in the other language...
Men can liken dreams to reality, if it's part of a boyish charm or innocence or inability / refusal to grow up. But there's certainly a niche for that, judging from all sorts of mega-popular songs by male performers about "dreams are real (enough)".
ReplyDelete"All I Have to Do Is Dream" by the Everly Brothers
- Youthful, teenager-in-love tone, and yet still has a mature bittersweet tone to show the awareness that dreams aren't actual happenings
"Dream Lover" by Bobby Darin
- Longing, pained tone of voice shows the awareness of dreams not being real, and emphasized by the lyrics -- he doesn't even have a girl in his dreams! He's wishing for a lover in his dreams, not a lover in reality -- and he can't even get that! The way he painfully croons, "Dream lover, where are youuuuuu...?" Even the realm of dreams is no guaranteed refuge from the harshness of reality...
"In Dreams" by Roy Orbison
- Tinged by the mature awareness that dreams are not 100% real, and that substituting dreams for reality is desperation or escapism
"I Like Dreamin'" by Kenny Nolan
- Also has a bittersweet tone to signal awareness that dreams aren't fully real
Really, the only song about "dreams being real", that is not pained by the realization that they are not the same as actual-happenings, and instead that celebrates dreams as an alternate reality in their own right, is from female-led Heart, "These Dreams". The music video uses a surrealist style, which also treats dreams as an alter-reality, not a non-reality or anti-reality or mock-reality.
https://www.youtube.com/watch?v=41P8UxneDJE
It even uses the English translation of Japanese "mata" -- "another"! In dreams, she "lives another life". It's more like a parallel life, separate from the waking life. Not exactly tying into "again", but related to it, by "another". Leave it to women to celebrate dreams as equally as real as waking reality, just unfolding in a parallel dimension. Girls may not dig sci-fi, but they love surrealism and dreams-shading-into-reality -- also why they love horror movies like A Nightmare on Elm Street, which also equates the real-ness of dream-world and awake-world, just unfolding in a parallel dimension.
These dreams go on when I close my eyes
Every second of the night I live another life
These dreams that sleep when it's cold outside
Every moment I'm awake the further I'm away
So men view dreams and reality as zero-sum, dreams as substitutes for reality, and the more you indulge in dream-world, the more you are retreating from awake-world.
ReplyDeleteWomen view dreams and reality as orthogonal, happening in independent dimensions, maybe linked somehow, but not as substitutes or other zero-sum way. They can indulge in dream-world without it requiring them to retreat from awake-world.
They live "another" life, or a second life, in dreams. So when they go bragging to their girl friends about who they bedded while dreaming, it's more like confessing their sordid tryst while on vacation in some exotic country.
Girls could not be cuter if they tried... ^_^
Gays probably think dreams are real too, while lesbians are probably more like men.
ReplyDeleteWhole lot of Korean & Japonic particles of Wa-Dene-Yeniseian origin. This will account for all the Old Korean particles listed at Wiktionary -- every single one with a Japonic and Dene-Yeniseian cognate. I'll get back to "mata" afterward, since that is a compound, and we need to see what its building blocks are first.
ReplyDeleteFirst is OK conjunctive particle "two" (pronounced "to"), used for "too, as well" and to mark the nouns in "either / or" and "neither / nor" constructions.
https://en.wiktionary.org/wiki/%E5%88%80#Old_Korean
This is cognate with the P-J conjunctive particle "tə". The vowel in OK is simply raised and rounded, much as the OJ reflex of the P-J form will be, "to".
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/t%C9%99
Japonic not only uses this for marking items in a sequence or list or group, like "and", it also uses it for "with" when there is a companion that the action is done "with" (the comitative marker).
Crucially for "mata", it has an archaic usage meaning "every" (distributive marker), which now only survives in certain fixed expressions.
Japonic has a wider range of "tV" particles that all come from the same root "t". Aside from "tə", there is "te", which marks verbs and adjectives rather than nouns like "tə". Aside from "and", it can also mean the method through which some action is done, and a reason or cause for which some action is done. These senses are closer to "with" than "and", suggesting a link or joining between 2 things, not just a temporal succession.
https://en.wiktionary.org/wiki/%E3%81%A6#Etymology_2
The P-J counter "tu", which attached to cardinal numbers, when a numeral is used to quantify a noun. Counting things is like marking each item in a list, like "tə". For the numerals "20" and "30", the counter is "ti".
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/-tu
There's no "to" in this group of "tV" words, perhaps cuz it sounds too similar to "tə". The "to" words in P-J have to do with "door, portal", "place", and "bird".
The remaining vowel "a" is used for "ta" in P-J, to mark a "sake, cause, reason", which is a kind of link, and therefore related to "with" and the comitative usage of "tə".
https://en.wiktionary.org/wiki/%E3%81%9F#Etymology_2_2
The Korean particle and the many Japonic particles beginning with "t" here, are cognate with the Proto-Yeniseian and Proto-Na-Dene "tɬ", the comitative or "with" marker.
ReplyDeleteYeniseian and Na-Dene do not have a fixed conjunctive particle that separates items in a list. Instead of saying "X and Y went fishing", it says "X with-Y went fishing". At the oldest stage, it probably marked every noun in the list.
It also uses this marker for instruments or tools or other means to an end, and that ties into the "sake, reason" usages above.
Phonetically, this is a suffix in Korean and Japonic, both of which ban coda clusters like "tɬ". Also, neither language has "ɬ" -- Korean changes it to "l" word-finally, Japonic changes it to "s" unconditionally. So it would be "tl" in Korean and "ts" in Japonic. Korean would have to delete one of the C's, and Japonic would have to as well, or use 2 dummy vowels instead of only 1 to break up "ts" into "tVsV".
That strategy involves too many changes, so a simpler strategy was taken -- just delete the "ɬ", which is only 1 change. (This strategy also took place in some Yeniseian daughters.) That avoids the extra step of altering "ɬ" to "l" / "s". It simplifies the coda cluster. And it leaves a phoneme in both inventories, "t".
Japonic needs a dummy vowel after, and it used every vowel except for "o", which is the least common "tV" syllable in OJ, for the various usages. The most common syllable is "tə", which received the most common usages, "and" and "with".
OK "two" is modeled as having an earlier V between "t" and the labial consonant, and that the labial weakened over time, e.g. from "p" -> fricative "b" -> "w", and finally the 1st V deleted, leaving "two".
Since only "t" or "tV" from this suffix is cognate with the Wa-Dene-Yeniseian forms, it's hard to tell whether it really used to have a vowel, since P-K allows word-final C's. Maybe the ancestral form was "t + WV", where "W" is some labial consonant, and Korean simply inherited "t" by itself from Wa. If it was followed by a vowel, it was likely epenthetic, only to break up the C sequence "tW", and therefore not necessary to account for in the phonological analysis of its similarity to the Japonic, Na-Dene, and Yeniseian cognates.
I'm leaving aside OK "kwa", the later comitative particle, which Vovin modeled as an earlier noun or verbal phrase, something like "kolp-a" meaning "comparing them", which became fused and grammaticalized. No insight into broader connections outside Korean for this one.
ReplyDeletehttps://en.wiktionary.org/wiki/%E6%9E%9C#Old_Korean
Likewise the conjunctive "i", whose early usage is obscure, but mostly involved an intimate connection between the two nouns it joined. Likely a Korean innovation, both phonologically and semantically / structurally.
https://en.wiktionary.org/wiki/%E5%88%A9#Old_Korean
OK suffix "k" has an emphatic usage after verbs, and another usage that is unknown. So, not much to go on here.
ReplyDeletehttps://en.wiktionary.org/wiki/%E5%8F%AA#Old_Korean
Korean & Japonic genitive particle "s". It was highly common in OK, though later replaced. Its Japonic cognate likewise only shows up in a handful of fixed expressions in OJ that survive today.
ReplyDeletehttps://en.wiktionary.org/wiki/%E5%8F%B1#Old_Korean
Some examples of the Japonic example, like "harusame" = "spring rain", where "haru" = "spring" and "ame" = "rain", leaving an "s" interfix to connect them, with the meaning that the rain "belongs to" or "pertains to" spring.
https://en.wiktionary.org/wiki/%E6%98%A5%E9%9B%A8#Japanese
These are cognate with the Proto-Yeniseian singulative suffix "ç", some of whose Yeniseian reflexes are indeed "s". Korean and Japonic made the same phonetic change, shifting it slightly forward since it's a mid-mouth voiceless fricative, and only "s" matches that in their phoneme inventories.
A singulative marker begins with a whole and carves out a single element from it. For example, in Ket, it attaches to the word for "eye" (construed as a pair of eyes), and returns a word meaning "a single eye" (i.e., one part of a pair). Similar logic as English expressions like "a grain of sand", although we employ full words rather than bound affixes.
Semantically, both the genitive and the singulative usages share the sense of "part belonging to the whole", in an inalienable way, like a body part belonging to a person, or a kin member belonging to other kin members / the whole family.
This inalienable sense is the one used for Yeniseian singulative -- one eye cannot be separated from the larger pair of eyes, nor can a grain of sand be separated conceptually from the mass or pile of sand that it belongs to.
The genitive has a part-from-the-whole sense, about inalienable possession, like the eyes cannot be separated from the person or whole body, nor can "one's mother" be separated from "one" or any other family member, conceptually they form an indivisible kin network.
This shows up especially in the Japonic phrase "harusame", which "rain" is like an inalienable part or possession of the "spring" season. Rain belongs to spring, as much as the eyes belong to the whole person, or a mother belongs to a whole family.
There is no Na-Dene cognate in this group.
Semantically, the likely path was that this marker was originally genitive, which it remained in Wa (and in both of Wa's offshoots, Korean and Japonic). Yeniseian innovated the usage of singulative. It likely did this via an intermediate stage where it shifted the genitive (about possession, belonging) to partitive ("some of" or "a piece of"), then further to singulative ("one part of a whole").
BTW, OK genitive particle "uy", which later replaced "s", is modeled as coming from the locational suffix "kuy", which we already discussed, and whose P-Y cognate is "qej", a locational nominal suffix.
ReplyDeletePhonetically, the model is that "k" weakened to "h" and then to null.
Semantically, this is like saying a person is "in" a household or "at" a certain home, rather than "belongs to" or "is possessed by".
Korean nominative suffix "i" in OK, which it remains in Modern Korean. This marks the subject of a verb.
ReplyDeletehttps://en.wiktionary.org/wiki/%E6%98%AF#Old_Korean
Korean and Japonic have nominative-accusative morphosyntactic alignment. So the subject of an intransitive verb (verb with only 1 noun argument), is given the same case marker as the agent of a transitive verb (do-er noun, not the done-to noun, for a verb with 2 noun arguments).
Yeniseian has a split alignment. In the relevant case here, it has a 3rd-person animate agreement prefix for a verb, "j". This marks the subject of an intransitive verb, but the patient (done-to) rather than the agent (do-er) of a transitive verb. So the language is ergative-absolutive here, not nominative-accusative.
Nevertheless, this is a subject marker of sorts. As we saw before, Yeniseian is head-marking, not dependent-marking, so it bundles the info about subjects and objects into verb markers, not noun markers. Verb markers, like these agreement prefixes.
So, when Korean and Japonic become dependent-marking, they will need markers to go on nouns, not verbs. But they simply adapted their older verbal agreement markers, for use as case markers on nouns. Which agreement marker? Well, generally a subject of a verb is not "I" or "you", but some 3rd-person noun. And they chose animate over inanimate (whose P-Y marker is "w"), since animate nouns are naturally more agentive or do-ers or subjects of verbs.
It's true that for transitive verbs, Yeniseian used "j" for the patient, not agent, of a transitive verb. But that's just due to its split alignment. Korean will change that to nominative-accusative, and consider only the intransitive verb, which has just 1 noun argument -- the subject -- who, in their new alignment, is like the agent of a transitive verb, not the patient.
Phonetically, "j" vocalized into "i", and that's all. Korean allows coda consonants, but perhaps they thought this is such a common particle -- every subject of a verb needs it -- that it wanted to avoid a coda cluster, if the subject ended in a consonant. So make it attachable to any noun, and make it a vowel instead.
This was not where Japonic got its nominative case marker from, but I've already covered that, and will get back to it later. I want to get through with Korean first.
Korean infinitive marker "a", from OK onward, used on nouns and adjectives.
ReplyDeletehttps://en.wiktionary.org/wiki/%E8%89%AF#Old_Korean
This is cognate with Proto-Yeniseian "ɢa", a verb conjugation prefix that gives the verb an atelic sense, which is contrasted with the prefix "çi" in the same slot, which gives it a telic sense.
The jargon about telicity is from Greek "telos" = "end, goal". Telic means it successfully reached the goal, atelic means it's uncertain whether it did or did not reach the goal.
The semantic overlap is that a verb in its infinitive form has a lack of certainty about reaching a goal -- it hasn't even been deployed in a sentence. Even when it is used in a sentence, it has a generic usage, e.g. "I like to run" in English. There's no particular act of running, where you either did or did not reach some goal, or other completion event. It has an indefinite quality, where it goes on and on, with no particular sense of completion.
In a finite form, a verb has some definite end-point, like "I ran yesterday" -- and that's over and done with, completed.
The open-endedness, indefinite-ness, lack of closure or completion, links the infinitive form of a verb and the atelic (rather than telic) aspect of a verb.
Phonetically, as we saw before, even Korean simply deleted the voiced uvular stop from Dene-Yeniseian, "ɢ", just like Japonic. P-K seems to have fronted "q" to velar "k", but since it doesn't have a voicing contrast, it can't do that for "ɢ" as well, without a merger into "k". So it simply deleted "ɢ", probably first treating it as a glottal stop, then deleting altogether.
This allows the old Dene-Yeniseian contrasts to be preserved -- where "q" has an overt consonant, while "ɢ" appears as the zero phoneme.
That just leaves "a" in Korean, QED.
Korean and Japonic topic particles from Wa-Dene-Yeniseian origin. The "topic" is background info, already known info, and then the predicate introduces some new bit of info about it.
ReplyDeleteFrom OK onward the topic particle is "n", which is also used to mark the realis mood (creating a gerund).
https://en.wiktionary.org/wiki/%E9%9A%B1#Old_Korean
In P-J, the exclusive topic particle is "pa" (modern "wa"), while the non-exclusive form is "mə" (modern "mo").
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/pa#Etymology_1
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/m%C9%99
Ready....? These are both cognate with the Proto-Yeniseian "ŋʷ", the perfective verb prefix. Again, Yeniseian was head-marking, not dependent-marking, so it didn't have noun markers for "old info noun". When Korean and Japonic switch to dependent-marking, they need morphemes to go on nouns, such as "the topic". Well, they just recruited ancestral morphemes that were semantically linked to their contemporary need.
How is "topic = old info, background info" related to "perfective" aspect of a verb? Perfective is an action that has definite beginning and ending points. "I ran across the room", not "I was running away from the fire". It could be brief or long in duration -- "I ran yesterday for 3 hours" -- but it has clearly defined boundaries that make it into a whole or gestalt.
This is opposed to imperfective, where either or both boundary is porous, allowing for a hazy ongoing indefinite aspect, and giving it more of an amorphous blob aspect, rather than a clearly defined whole or gestalt. "I was running around yesterday".
In a topic-based language like Korean and Japonic, the topic is the old, background, already known, presupposed noun. That is semantically similar to a verb whose action has crisp clear boundaries, rather than hazy boundaries and ongoing-ness. The shared sense is "compartmentalizing" something with definite boundaries, whether an "already understood" noun for the topic, or an action that has certain beginning and ending points.
The former is compartmentalizing a noun within the "space" of information in the discourse. The latter is compartmentalizing a verb within the "space" of time-flow, regardless of whether it's compartmentalized in the past, present, or future.
Phonetically, Korean shows the same sound change that we saw before with Dene-Yeniseian "xʷaj" = "sun; day", where it simply de-labialized the labialized C (to "h"), as opposed to Japonic making the labialization primary place of articulation (as "p"). That leaves "ŋ", but Korean lacks that nasal until recently, so it shifted it to the nearest place where there's a nasal, which is "n", QED.
ReplyDeleteP-J preserved the labial feature, but as primary labial place, just like with "xʷaj", which it converted to "pi". Since the primary consonant is nasal, it now has the option to keep it nasal, or alter its airflow to oral. It did both. By preserving nasal manner, and making "ʷ" into labial place, that led to "m". By allowing oral airflow, "ʷ" became primary labial "p".
However, both need a dummy vowel, to fulfill CV syllable shape. The most common "m" syllable in OJ is "mə", QED. And the most common "p" syllable is "pa", QED.
Why did "p" get the exclusive meaning, and "m" the non-exclusive meaning? IDK, maybe they flipped a coin. There is no such split usage in Korean, nor in Yeniseian, since it's just a single verb morpheme there.
Perhaps the speakers couldn't decide whether to preserve the nasal manner or not, since most of these conversions of secondary labialization into primary labial place, involve oral consonants like "k" and "x", not nasals. So they decided, "why not both?" and developed distinct usages for each one afterwards.
And that does it for the Korean particles! Wa-Dene-Yeniseian, all the way down!
IDK off the top of my head, if I ever discovered the Dene-Yeniseian cognates of the Japonic particles "pa" and "mo", but if I did not before, I just did now! ^_^
Why didn't Japonic allow "ʷ" to become the other labial-place consonant in its inventory, "w"? It allowed "m" and "p", what gives?
ReplyDeleteWell, "w" is already taken by the accusative marker, "wo", and that semantically conflicts with being a topic. First, it's mixing up the "topic vs. news" dimension, and the "subject vs. object" dimension. And if anything, a topic behaves more like a subject than an object -- it's followed by a predicate, usually in an active rather than passive voice.
So it could not allow "w", to avoid homophony, only "m" and "p".
Wa-Dene-Yeniseian origin of Japonic verb "to be; to have" and irrealis verb stem vowel. This relates directly to the Korean infinitive suffix, since it's from the exact same W-D-Y root, and has the same phonetic form.
ReplyDeleteNamely, the P-Y atelic verb prefix "ɢa", whose voiced uvular stop is simply deleted, as elsewhere in Japonic, and evidently in Korean as well. That leaves "a". There's no labialization for Japonic to hang onto, by turning it into a primary labial consonant. So it just goes bye-bye, probably after an intermediate stage of being the glottal stop.
In P-J, the verb for "to be, exist; to have" is "aru", whose "ru" I analyze as the default verb suffix, i.e. "r" is not part of the root. Rather, "ru" is added to a vowel-stem root, namely "a".
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/ar-
What is the semantic link between atelic aspect of a verb, and the verbs "to be", "to exist", and "to have"? Well, those are THE three canonical atelic verbs, which lack an inherent end-point. They're even more atelic than other atelic verbs, such as "run away from the fire", since they are stative rather than dynamic.
At any single sub-moment of time, "being" is still "being", and "having" is still "having". But if you freeze someone running away from the fire, you might catch them mid-stride, which doesn't feel as truly "running" since "run" is a dynamic verb as well as atelic. So, "to be, exist; to have" is the most prototypically atelic verb in existence. That's why it adapted Wa's ancestral atelic verb prefix, as the semantic verb root, "a", before sticking on the default verb ending "ru", yielding "aru", QED.
But wait, there's more! Where else does this "a" appear? As the defining vowel for the irrealis mood in P-J verb conjugation, for verbs with consonant-final roots. After the final consonant of the semantic verb root, "a" appears to indicate the irrealis mood. This mood is used for negation, but other actions that have not yet happened, like potential actions.
https://en.wiktionary.org/wiki/%E6%9C%AA%E7%84%B6%E5%BD%A2#Japanese
What is the semantic link between atelic aspect of a verb, and the irrealis mood of a verb? Again, its lack of an end-point, boundary, completion, etc. Just as an atelic verb lacks a defined end-point for concluding, the irrealis mood says the verb hasn't even taken place in reality yet, and therefore not only lacks a defined end-point, but lacks a defined start-point as well. Not being anchored in a real, particular instance of the verb, it floats in a hazy drifting indefinite state, only vaguely potential, not bounded by the real world.
So it's only natural for Japonic to adapt the same atelic aspect marker, for its irrealis mood marker.
Korean adapted W-D-Y atelic marker "ɢa" for the infinitive marker on verbs and adjectives, but Japonic adapted it as well -- not for the infinitive marker, but for other equally unbounded, uncompleted, indefinite usages, like the root of the ultimate atelic verb "to be; to have", and to mark the irrealis mood in verbs.
BTW, I forgot about the negation particle in Korean and Japonic, and their Wa-Dene-Yeniseian origins, and will get to that later.
Korean & Japonic negation is of Wa-Dene-Yeniseian origin. We'll start with Korean. Old Korean has several related morphemes that combine to form negation words. There's "ANti", which negates nouns, and "ANtol", which negates verbs. There's "ANTOk", to express "cannot" or inability to do a verb. Finally, "mwotol", also for "cannot", but also "poorly".
ReplyDeletehttps://en.wiktionary.org/wiki/%E4%B8%8D%E7%9F%A5#Old_Korean
https://en.wiktionary.org/wiki/%E4%B8%8D%E5%86%AC#Old_Korean
https://en.wiktionary.org/wiki/%E4%B8%8D%E5%8F%AA#Old_Korean
https://en.wiktionary.org/wiki/%E6%AF%9B%E5%86%AC#Old_Korean
There are 2 distinct negation roots here, "an" and "mwo". The "to" is modeled as a semantic light noun like "objective fact". Those two combined mean "is not a fact". The final suffixes ("i", "l / r", and "k") are unknown in meaning, but are probably used to agree with the part of speech they're negating, since "ANtol" and "mwotol" both negate verbs, "ANti" negates a noun, and "ANTOk" has a unique suffix for verbs.
We'll focus on the main negation roots, "an" and "mwo".
Yeniseian languages have a case marker that means "without, lacking", which is "an", whose semantic overlap with "no, not" and negation in general, is obvious. "Without, lacking" negates the basic verb "to have" rather than "to be", but still a very basic atelic indefinitive stative kind of verb, to use as a prototypical example.
It's thought that P-Y did not use this case marker on nouns, back when it was highly polysynthetic and mainly altered verbs. Then only later on, under influence of the suffixing languages surrounding the family, like Uralic, it was adapted as a case marker on nouns. IDK about its use as a "without" case marker, but they didn't get the phonetic form out of thin air. They must have had "an" in use somewhere in their language, with a meaning related to negation, and then adapted it later for the "without" case marker.
That means there was an ancestral P-Y morpheme "an" relating to negation, and that is cognate with the P-K negation root "an".
As for OK "mwo" (pronounced "mo"), this is cognate with the basic negation particle in P-Y, which is "wǝ ~ wǝn". From internal analysis of Yeniseian, there is disagreement over whether the labial C is "w" or "b", but my external comparisons with Korean and Japonic show it must be "w" at the oldest stage. Maybe that means the P-Y stage has "w", or that it hardened a Pre-PY "w" into "b" by the P-Y stage of its history.
https://en.wiktionary.org/wiki/Reconstruction:Proto-Yeniseian/b%CA%8C
Recall that P-K lacks "w", and the most faithful rendering of it is "m", with the phonemes in the P-K inventory. It's either that or "p", which also loses the continuant manner (which "m" preserves), or "y" (which also loses the labial place, which "m" preserves).
The vowel rounding to "o" in Korean is paralleled in Yeniseian as well, such as Kott "mon" and Pumpokol "amut", which breaks down into "a + mu + t", where "mu" is the negation particle. These Yeniseian forms also both hardened "w" to "m", as did Korean.
Addendum: the external comparison to Korean and Japonic shows that the oldest form of the negation particle beginning with a labial, had "m", not "w" or "b". This was retained in Wa, leading to Korean and Japonic.
ReplyDeleteHowever, in P-Y, the ban on initial nasals took effect, which did not affect Na-Dene, Korean, or Japonic. So P-Y had to alter "m" to "w" or "b", as this C is word-initial, since this negation particle is a freestanding word, not a bound morpheme within the verb template.
Now onto Japonic negation, starting with another cognate of the "m"-word in Korean and the labial word in Yeniseian. There's a false cognate, the prefix "mu" = "not, un-", which Japanese borrowed from Chinese.
ReplyDeleteHowever, there's a P-J word "muna" = "empty", which is another common semantic anchor for expressing negation. It's another way of negating the prototypical verb "to have", in the sense of "to contain within it". Negate that, and you get "empty".
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/muna
Over the centuries, the OJ adjective form "munasiki" (modern "munashii"), acquired several other senses relating to negation of basic adjectives -- "false", "uncertain", "lifeless", and "unthinking".
https://en.wiktionary.org/wiki/%E7%A9%BA%E3%81%97%E3%81%84#Japanese
Phonetically, this is cognate with the variants in Yeniseian that end in "n", since there is variation in P-Y between those with and without final-n. It also has the back-round vowel the several others do in Yeniseian, and in Korean as well.
However, P-J needs CV syllable shape, so a dummy vowel was added. It could not choose the 1st and 2nd most common "n" syllables in OJ, "nə" and "ni", since they were already used for suffix particles (genitive and dative). In the word "munV", these would sound like genitive or dative suffix particles attaching to "mu", rather than including "n" in the lexical root.
So to avoid that, the 3rd-most common syllable was chosen, "na", yielding "muna", QED.
The kanji used to write "muna" in "munashii" has a rare alternate reading of "ana", although usually "ana" is written with a different kanji, for "hole" -- but still, related to "empty space". And just as with "muna", "ana" acquired several other senses relating to negation -- "loss, deficit" and "flaw, defect".
ReplyDeletehttps://en.wiktionary.org/wiki/%E7%A9%BA#Japanese
https://en.wiktionary.org/wiki/%E7%A9%B4#Japanese
I think "ana" here is from "an" with the same dummy vowel "a", also chosen to avoid confusion with the genitive or dative particles attaching to "a", and so to include "n" in the lexical root "an".
These Japonic uses of "ana" are therefore cognate with the Korean negation particle "an", and the Yeniseian "without, lacking" case marker "an".
To emphasize, Japonic shows that the original labial C in the W-D-Y negation particle was "m", not "w". Korean lacks "w", so Korean "m" could have been earlier "m" or "w". Yeniseian lacks initial "m", so it's initial "b" and "w" could have been those earlier, or "m". Na-Dene lost its labial consonants altogether!
ReplyDeleteBut the language of Glorious Nippon allows initial "p", "m", and "w". The fact that the Japonic form has "m" says it always did, not that it hardened earlier "w" to "m". P-J not only has "w", it loves "w". It generally keeps "w".
There is one exception, when it's followed by "u". The syllable "wu" is banned all the way back to P-J, just like "yi", to dissimilate glides from their vocalic counterparts.
What about "muna", then? Couldn't that have originally been "wuna", and because "wu" is banned, "w" -> "m"?
Hypothetically, that's possible, but in reality that's not how Japonic resolved "wu" -- they typically just deleted the "w" and left "u". That is also how they resolved "yi" in the palatal location, deleting "y" and leaving "i".
As other "w"-syllables and "y"-syllables have become banned in Japanese, the same solution occurred -- delete the C, leave the V. Although "wo" was popular at first, it later became banned, and "w" was dropped, leaving "o". Likewise for "ye" being legal at first, then later becoming illegal, leading to deletion of "y" and leaving "e".
This is such a long-standing solution in Japonic, it must have been their solution to a Pre-PJ "wu" as well. Well, you could also alter the vowel to avoid the illegal combo "wu". But assuming you left the vowel as "u", then delete the "w" rather than alter it to something else like "m".
Japonic negation is mainly handled by "n" particles, though, not "mun" or "an". P-J "no, not; don't!" is "na", and the negating suffix on verbs is "nu".
ReplyDeletehttps://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/na
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/-nu
The vowels are likely dummy, since "nu" attaches to verbs, and verbs obligatorily end in "u", so it looks more verb-like.
To avoid homophony with the already taken "nV" particles (genitive and dative) with the 1st and 2nd most common "n" syllables in OJ, the 3rd-most common was chosen for negation, "na".
These P-J particles don't have cognates in Korean or Yeniseian, but they do in Na-Dene. This is not surprising, since there are several strategies to pursue for negation, and not every branch of a family will choose the same ones.
In Na-Dene, negation involves 3 morphemes: a freestanding negation particle that lies outside the verb template, a prefix on the verb, and a suffix on the verb.
The first one, the freestanding negation particle, is reconstructed as "de-n" or "ne-d", containing a component "n". Originally it was a verb meaning "is absent, is lacking, has disappeared".
This "n" component is not only the phonetically closest match to the Japonic "n" negation particle, but morphosyntactically as well -- since Japonic allows it to appear unattached to a verb. For instance, "nai" is a freestanding word, not an affix or particle.
https://en.wiktionary.org/wiki/%E3%81%AA%E3%81%84#Etymology_2
And originally it must have been a verb that meant "to be absent, to be lacking, to be non-existent". To express "I have no money," Japanese says, "Okane ga nai", where "okane" = "money", "ga" is the subject marker (not the topic particle), which means there must be a verb somewhere, whose subject is the noun marked by "ga". The only remaining word is "nai", so this must be a verb, meaning "is absent / is lacking / is non-existent". Translated literally, the Japanese sentence means "money is absent / lacking / non-existent".
Later, "nai" grammaticalized further into a suffix that attached to verbs, still for negation, and replaced the older "nu" suffix. This is what Modern Japanese uses, but it only started in the Early Modern era, in the Eastern dialects.
https://en.wiktionary.org/wiki/ない#Etymology_3
I don't think the "n" words in Japonic are cognate with "an" in Korean and Yeniseian. Crucially, Japonic allows a freestanding word like "nai", which has no vowel whatsoever before "n". The Korean and Yeniseian morphemes are always have "n" preceded by "a".
ReplyDeleteThe only vague link is that in Japanese verbs, the negative suffix like "nu" or later "nai" follows the irrealis stem vowel "a". So that gives a sequence of "an" in the larger verb. But as I showed earlier, this irrealis "a" is a morpheme unto itself, just "a", and it's cognate with the atelic marker in P-Y verbs, "ɢa".
But Japonic "n" words relating to negation *are* cognate with the freestanding negation particle in Na-Dene.
Probably this split in the family relates to which particular semantic anchor they chose for expressing negation -- "an" seems to negate the verb "to have", "nai" and "de-n / ne-d" negate "to be" or at least "to be here / to be present", and "mun" (or whatever that vowel is) negates "to contain within it".
It's possible that Japonic shows a cognate with Na-Dene negation in one other way. The Na-Dene prefix used in negation is "zə / sə". P-J has no voiced obstruents, so this will be "sə".
ReplyDeleteAs languages change from polysynthetic to not, they may adapt earlier morphemes for semantically similar uses, but in a different morphosyntactic structure. Some may even be adapted as the root of a new basic word. Polysynthetic languages like Wa have so many morphemes that belong to the verb template, why not adapt some of them, even when you no longer build sentences that way?
I think that's what happened for the P-J word "sora" = "sky", but also "a place that is distant", which involves negation ("is not here"), as does its use as a prefix, meaning "fake" or "incorrectly perceived".
https://en.wiktionary.org/wiki/%E7%A9%BA#Japanese
It must have originally been "səra", since it appears as "sora" in OJ, and usually that meant "ə" was raised to "o". It's hard to tell, though, since "sora" is the only PJ word beginning with "so", and "səri" = "razor" is the only one beginning with "sə". For whatever reason, P-J did not like "s"-initial words to have the 1st V be either of those vowels.
OJ has a pluralizer suffix "ra", which is what I think "sora" has -- in this case, more of a collective suffix, since it's a mass noun, or rather a vacuum noun, rather than a count noun. So, I think the root "sə" meant something like "vacuum, void, emptiness, etc." Note that the kanji for "sora" is the same one as for "muna" = "empty", confirming their semantic likeness in Japanese.
Adding the plural / collective suffix "ra" to "empty" or "emptiness" not only ensures that it's a noun, but a great big collective of smaller elements of emptiness. That is, a great big void like the sky, between the earth and the stars. In English, too, we use a plural marker for "fly the friendly skies", or "up in the heavens", to make the "sky" / "heaven" sound more collective.
The Japanese word "sora", then, means something like "the voids" or some other collective sense of a negation word, with a tangible semantic anchor like "to (not) contain within it".
The last Na-Dene morpheme, the suffix "he", I don't see a cognate for in Japonic, Korean, or Yeniseian.
But there does seem to be one crucial trace of the Na-Dene negation prefix in Japonic -- "the sky", conceived of as "the voids". ^_^
Phonetically different pair of words about mouth organs is shared among Yeniseian, Japonic, and Korean. This shows that these languages share a common ancestor, in exactly the same way that sharing a pair of homophones does.
ReplyDeleteIn this case, it's not 1 phonetic form with 2 unrelated meanings (homophones), it's 1 meaning that comes in 2 phonetically unrelated forms (synonyms). If languages share just 1 of those arbitrary sound-meaning links, maybe that's unlikely but allowed by chance. However, when they share the other arbitrary sound-meaning link, which overlaps with the first pair in either element (sound for homophones, meaning for synonyms), then that's unlikely squared = impossible.
Suppose I told you that language A and B have a similar-sounding word meaning "ground". OK, maybe it's a loan. But now suppose I told you that in both A and B, this word has a synonym -- phonetically unrelated, but also meaning "ground", or something very near to "ground". OK, big deal, languages have synonyms. But now suppose I told you those synonyms were also phonetically similar across A and B -- that's unlikely squared. It's not just sharing a (meaning, sound) link, but (meaning, sound1, sound2) link.
"Well, maybe one language borrowed both synonyms from the other language" -- not for everyday language. Maybe for learned jargon, but everyday speech just borrows one phonetic form for the meaning. The whole point is utilitarian -- we don't have a word with that meaning, so we'll borrow it, in some foreign phonetic form. You're not going to borrow the same, or nearly the same, meaning in a totally unrelated phonetic form -- that's confusing, and serves no purpose. Only for learned jargon, like how linguistics in English has borrowed "oral" and "buccal" from Latin, both relating to "mouth" semantically, but phonetically unrelated.
With that general argument made, let's look at the particular case here. The synonyms are words relating to the organs of the mouth, or what they do (a simple semantic shift). One of them is a purely anatomical word, or whose function is eating. The other is more abstract, and whose function is vocalizing, speech, and language.
ReplyDeleteStarting with Proto-Yeniseian, there's "ɬVb" = "tongue", whose other connotations are eating, e.g. "morsel". The vowel is back, either "a" or "u" in the reflexes.
https://en.wiktionary.org/wiki/Reconstruction:Proto-Yeniseian/%C9%ACVb
There's also "ej ~ aj" = "tongue", whose other connotations are about vocalizing and speech.
https://en.wiktionary.org/wiki/Reconstruction:Proto-Yeniseian/ej#Etymology_1
Early on in my revelation about Japonic being related to Yeniseian, I pointed to P-J "supa" (listed as "sunpa", but no evidence of a nasal; just a convenience to say that later in OJ "p" -> "b"). This means "lips, tongue", with the only other connotation being "saliva", so a pretty low-level anatomical term, not related to language.
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/sunpa
Since P-J lacks "l" to preserve the lateral feature, "ɬ" -> "s" (its only voiceless fricative in mid-mouth location). P-J lacks voiced obstruents, so "b" -> "p". The P-J vowel is "u", as in one of the Yeniseian reflexes. It gets a final dummy vowel to fit CVCV shape, and by far the most common "p" syllable in OJ is "pa", yielding "supa", QED.
Sadly, no trace of this "tongue" word in Korean, but this does build half the case of Yeniseian and Japonic sharing a pair of synonyms.
The other word does appear in both Japonic and Korean. In P-J, "ipu" = "to say", leaving behind the body part sense altogether, but preserving the connotation of speech and language.
ReplyDeletehttps://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/ip-
In Old Korean, "ip" = "mouth; lips", with the connotation of "word"
https://en.wiktionary.org/wiki/%EC%9E%85#Korean
The external comparison to Korean suggests that the P-J root is "ip-", with "u" being the obligatory verb ending. That is confirmed by internal analysis of P-J, where verbs whose final stem consonant is "p" are rare, suggesting the "p" is not a dummy consonant after a root of "i-", but rather part of the root itself, "ip-".
There is no use of the default super-popular verb ending "ru" attached to "i", so that should have been the first choice, if the root were only "i". The fact that there's no "iru", but there is "ipu", means the "p" in "ipu" is part of the root, not a dummy sound. Only the final "u" is outside the root, being obligatory for verbs.
These Korean and Japonic words are clearly cognate with each other. But what about Yeniseian? Compare them with "ej" = "tongue", with connotations of speech rather than eating. Semantically, they fit together.
Phonetically, the P-Y word lacks "p" or other labial C, or any obstruent C. I think the final "p" in the Korean and Japonic roots is contamination from the other synonym for "tongue", which ends in "b" or "p", which will be "p" in P-K and P-J due to lack of voiced obstruents. The semantic overlap is heavy, since they're synonyms.
Phonetically, I think Korean or Japonic, or maybe their parent Wa, wanted more consonantal meat somewhere in that word. And "ej" has no consonant onset at all, a nuclear vowel, and a semi-vowel coda -- way too vocalic. Where's the consonantal beef? So perhaps in Wa itself, not its offshoots Korean or Japonic, a "b" or "p" was appended to the end, for the phonological reason of giving it more of a canonical CVC shape (and "p" being more of a prototypical C than is "j"), and by contamination from the synonym which ends in "b" / "p".
Something like "ejp" in Wa.
For P-J, "j" is vocalized to "i", which replaces the preceding vowel, as always, yielding "ip", which gets the obligatory vowel ending "u" in P-J to yield "ipu", QED.
For P-K, it's not a verb, and Korean is fine with coda C's, but it doesn't want the coda cluster "jp". So just like its half-sibling Japonic, it vocalized "j" to "i", which replaced the preceding vowel, yielding "ip", QED.
It's not a coincidence that if you ran P-Y "ej" through the standard filter to convert to P-J, it gives "i", which is half of the root for "to say". And it's no coincidence that the Korean word for "mouth; word" has the same vowel as half of its root.
ReplyDeleteSince the other half of the Korean and Japonic root is coda "p", there must have been a coda "p" earlier on -- and since it's not there in Yeniseian, it was likely there in Wa, contaminated by the Wa cognate for P-Y "ɬVb", which we know must've been there since P-J shows a reflex of it, as "supa".
Therefore, after accounting for this "p" contamination, P-Y and P-J and P-K share this root about "tongue" or "mouth", with the connotation of speech.
And that means Yeniseian and Japonic share a pair of synonyms, each member of which has a phonetically corresponding form across the languages. This means the languages share a common ancestor.
Strictly speaking, Korean only has a cognate with P-Y "ej", but since this comes from Wa, and we know from P-J that Wa had a cognate of P-Y "ɬVb", P-K probably had a reflex of the Wa form as well, but just lost it. And so, Korean also shares a common ancestor with Yeniseian, the same as Japonic -- namely, Proto-Wa-Dene-Yeniseian. ^_^
Final reminder that borrowing is impossible anyway, between Yeniseian and either Japonic or Korean. Japonic came into existence in the Korean peninsula and Japanese islands, and Korean came into existence in the Korean peninsula.
ReplyDeleteKorea, let alone Japan, has never been in any contact with speakers of Yeniseian languages, whose homeland is the Yenisei River and Altai-Sayan region, although some of whom joined the Eastern Steppe confederations in the late Classical era.
But those Eastern Steppe confederations did not cross the mountains into Korea, let alone into the Japanese islands.
Since they have never lived near each other, or spoken to each other, Yeniseian and either Korean or Japonic must owe their similarities to a distant shared ancestor.
Korean and Japonic share a parent, Wa, while as creoles having minor second parents (Eskaleut for Korean, Emishi / Ainu for Japonic). And Wa is sister to Yeniseian, and to Na-Dene, whose similarities also cannot be explained by borrowing, since they came into existence in the New World!
Untangling correspondences among Yeniseian, Japonic, and Korean, involving labials and nasals. P-Y bans nasals in initial position, while Japonic and Korean allow them.
ReplyDeleteSo, it should be possible to find cases where Japonic and Korean have "m", while Yeniseian does not, since it's banned, but it has something similar to "m" instead, like "p / b" or "w". That's what I found with the negation particle, which begins with "m" in Korean and Japonic, but "b" (or possibly "w") in Yeniseian.
However, sometimes "m" in Japonic comes from earlier "p / b", and the labial stop has been converted to nasal airflow. This is well documented from OJ onward, e.g. the noun "smoke" is now "kemuri", and has been pronounced with "m" since at least the 1300s, but in OJ it was "keburi" with "b".
There are several other famous examples (like "to sleep" used to be "neburu" but is now "nemuru"), but this is not a systematic sound shift, it's only sporadic. Most words with "p / b" in OJ remain "p / b" in Modern Japanese, not shifted to nasal "m".
Still, the fact that there was even a sporadic tendency to shift a labial stop to a labial nasal, means that could have been going on much earlier, as part of a Japonic family sporadic preference. It is well documented from OJ onward, but it could have happened earlier than that -- and I've found a perfect example of it having occurred before the P-J stage.
The P-J word for "fire" is "poi".
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/poi
This is one of those words where the "oi" was not pronounced in succession, since P-J banned vowel sequences. Rather, it alternates between "po" and "pi" (and in some dialects, "pu"), depending on how it appears within a larger word or sentence. Somehow, though, both forms are listed in the mental dictionary of Japonic speakers, something like "p(o/i)".
In my work on the Dene-Yeniseian origins of Japonic, I showed that these vowel-alternating pairs seem to come from Wa forms where there was an original consonant between 2 vowels, almost certainly a uvular consonant, and probably a voiceless uvular fricative ("X"), which disappeared. Japonic's general strategy for dealing with uvulars is to simply delete them, unless they have labialization, in which case they could convert that into primary labial place, as "p" or "m". But when the C has been deleted, it was probably "X", perhaps going through an intermediate stage of "x", since deletion of velar fricatives between vowels is common cross-linguistically.
For example, P-J "mai" = "eye" shows final vowel alternation, suggesting a lost C between 2 vowels. Its cognate in Na-Dene is P-Athabaskan "naX" (voiceless uvular fricative), also meaning "eye". The P-A form begins with "n" instead of original "m" (preserved in Wa), since Na-Dene lost their labial consonants early on, so "m" became "n".
https://en.wiktionary.org/wiki/Reconstruction:Proto-Athabaskan/n%CE%B1x%CC%A3
By the time Wa / Japonic required CVCV shape, it was given dummy vowel "i" (all of the vowel alternating words have "i" as their final V), yielding "maXi". Then later, "X" deleted between vowels, leaving "mai", whose vowel sequence was later banned, and so it was codefied in the mental dictionary as "m(a/i)".
We'll get to the Dene-Yeniseian cognates of P-J "poi" in a bit, I just want to set this up now, about the vowel-alternating words.
Right now, the main point is that "poi" shows a derived form that begins with "m", also in P-J, and is therefore proof that the nasalization of labial stops, including word-initially, was already occurring -- sporadically -- before P-J.
ReplyDeleteNamely, the P-J word for "to burn" is "moyu", later expanded to "moyaru" for intransitive and "moyasu" for transitive.
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/moy-a-
The semantic overlap between "fire" and "to burn" is self-evident.
The final "u" is the obligatory verb ending for Japonic. But the "y" is part of the root. It's not part of a dummy verb ending "yu", since "y" is rare as the final C of a verb stem. If the root were only "mo", then it would receive the highly common dummy CV endings "ru" or "ku", and yet neither of those verbs exist in P-J. So, the root is "moy-".
The "o" vowel is the same for "poi" and "moy-", the higher rounded one, not "ə". They both begin with labial C's, seemingly a case of "p" -> "m" (the noun "fire" being more semantically basic, and therefore temporally earlier, than the verb "to burn").
To derive "moy" from "poi", all that remains is to convert "i" to its consonantal counterpart, the semi-vowel "y" in the same palatal location. This is done to give the verb root the default shape CVC, before the obligatory "u" ending.
The reverse process happened commonly, where coda "j / y" from Wa-Dene-Yeniseian became the vowel "i" in Japonic, since P-J banned coda C's. But by converting "i" to "y", for a verb stem, that does not violate the rule against coda C's, since it is the onset C of a 2nd syllable in the word, before the obligatory 2nd V "u". It also resolves the ban on vowel sequences, by turning one of them into a C. Everybody's happy...
I'm not the first to propose that "moyu" was derived from "poi". An obscure master's thesis by Bentley (1997), "Mo and Po in Old Japanese", also derives the verb from the noun. But the summary sounds like something different from what I'm saying, regarding the derivation. Well, ignoring *why* we think "moyu" comes from "poi", the point is, it did.
It's not part of a systematic change of "p / b" -> "m", it's just one example. Perhaps there are a few others I'll find, but again that will just be sporadic, not systematic -- just as in the well documented phenomenon from OJ onward. This little investigation shows that this sporadic tendency was already at work before P-J, since "moyu" is a P-J word, with Ryukyuan reflexes as well, not just an OJ word (when it was expanded into separate transitive and intransitive forms).
Bringing it back to the main theme of the Wa-Dene-Yeniseian family, it means we can't always rely on initial "m" in P-J to reflect "m" in Wa or Proto-Wa. It could come from earlier "p / b", sporadically and unpredictably. And so, if P-Y shows initial "p / b", that is not an instance of avoiding their ban on initial nasals, since it never was a nasal, it's always been a stop. We were only fooled into thinking it was a nasal due to the P-J form, which sporadically shifted "p / b" to "m".
This is where Korean can help, to triangulate with Yeniseian and Japonic. If Yeniseian and Korean say it was a stop, and only Japonic says it was a nasal, perhaps it's a case of sporadic "p/b" -> "m" in Japonic. They're still all cognates, but figuring out the original shared ancestral form, can be tricky.
Na-Dene can weigh in as well, although since it's further away, cognates may not be as easy to spot.
For the Old World members of Wa-Dene-Yeniseian, though, having 3 mini-families is helpful, since triangulation is now possible.
Now onto the Wa-Dene-Yeniseian origin of Japonic "poi" = "fire". Well, first is the obvious one, Korean, to establish its Wa origin. Today, "fire" is "bul", but before voiced obstruents, it was "pul", and before "r" was merged into "l", this was probably "pur", based on the Chinese character that Old Korean used, whose Middle Chinese pronunciation was "bwojH" or "bwot", where "t" is closer to "r" than to "l". But the fact that one reading of it ended in semi-vowel "j" puts it even closer to P-J "poi" and the final consonant of the stem in P-J "moyu". However, the P-K vowel was an unrounded back vowel, not "o" or "u".
ReplyDeletehttps://en.wiktionary.org/wiki/%EB%B6%88
In Proto-Yeniseian, "fire" is "beg", "begja", or "boʔk". Something unusual about the back consonant, but it was back.
https://en.wiktionary.org/wiki/Reconstruction:Proto-Yeniseian/beg
In Proto-Na-Dene, "to boil" is "ʔwiːgw".
https://en.wiktionary.org/wiki/Reconstruction:Proto-Na-Dene/%CA%94wi%CB%90gw
The Dene-Yeniseian words have unrounded vowels, front, mid-high. IDK why Wiktionary lists the P-ND form with "i", when the Tlingit reflex has "oo" and the P-A one has "e". The Na-Dene words reflect the loss of labial stops in the family, altering them to labial approximant "w" instead (or deleting the C altogether in Tlingit).
In Yeniseian, the reflexes have front unround "i" in Kott, but back round "o" in Kett, "ɔ" in Yugh, and "u" in Pumpokol. Tlingit also has back round "oo".
Vajda doesn't build in the rounding with a coda "w", a la "bewg", but rather says the round vowels are a sporadic case of labial "b" spreading its rounding onto the following unround vowel, and perhaps the following C being velar, pulled the vowel back as well.
I propose the same happened in Wa, also sporadic. Korean did not round the vowel, but did pull it back. Japonic pulled the vowel back, and rounded it as well. Both Korean and Japonic forms begin with "p", which can spread rounding just like "b" in Yeniseian.
And although the following C has been deleted in Wa, it was likely a velar or uvular at the point of deletion, as in the other cases of P-J final vowel alternation. So that back C could have pulled the preceding vowel back, as in Kett, Yugh, and Pumpokol on the Yeniseian side.
To reconstruct the Wa form, we'll start simple, mimicking P-Y "beg". Voiced obstruents were lost, leaving "pek". The back location of "k" pulled the vowel back to "ə", "pək". Now comes the first wrinkle -- why does "k" delete in both Korean and Japonic?
ReplyDeleteJaponic is easy, it has obligatory CVCV word shape, which forces the addition of final "i", then inter-vocalic velar deletion occurs, perhaps by first softening "k" to "x" in Pre-PJ. At any rate, leaving "pəi", then the rounding of "p" spreads forward, yielding "poi", QED.
However, P-K does not have obligatory CVCV shape, it's fine with word-final C. There's no reason to insert a final "i". So how does the "k" delete, and how does a "y" or "r" show up in its place? This calls for a revision to the simple form "beg".
Vajda also reconstructs the P-Y form as "beg-ja", with the suffix "ja" being a generic nominalizer that attaches to nouns. Suppose it had a cognate in Wa -- I've already established the Korean and Japonic cognates of all sorts of bound morphemes from P-Y and sometimes P-ND, so why not here as well? It was originally "ja", but whose "a" became reduced due to being unstressed, and then ultimately deleted. From "begja", that leads to "pək" as before, but now with final "j", that is "pəkj", with an illegal coda cluster.
There's no sign that Wa or Korean or Japonic ever had palatalized consonants ancestrally, so merging "kj" into a single palatalized "k" is out.
Korean seems to have simply deleted the inner member of the cluster, and kept the outer member, to preserve the word boundaries. That is, "pəj", which is compatible with the Middle Chinese reading of the character used for OK "fire". Perhaps the deleted "k" preserved some of its features by semi-hardening "y" to "r", to make it less vocalic and more consonantal, since "k" is a stop, the most consonantal you can get. Although "r" is not a stop, it's less resonant than semi-vowel "y".
In Japonic, "pəj" would have resulted in "pi", not a vowel-alternating pair. That is, "j" vocalized into "i", vowel sequencess are resolved by keeping 2nd and replacing 1st, yielding "pi". So what happened to the Wa form "pəkj", the cluster was first resolved by vocalizing "j" to "i", yielding "pəki". Then, perhaps by first weakening "k" to "x" in Pre-PJ, the velar C was deleted between vowels, leaving the vowel-alternating form "pəi". Lastly, the rounding from "p" spread forward, leaving "poi", QED.
So, now we see that not all cases of Japonic vowel-alternating pairs owe their final "i" to an epenthetic vowel, but perhaps to an earlier semi-vowel "j" as the onset of the 2nd syllable, which was vocalized into "i" to avoid coda C's, when the vowel of that 2nd syllable deleted due to being unstressed.
Another natural element, "wind", shows Korean and Japonic cognates with Na-Dene and Yeniseian.
ReplyDeleteIn P-Y, "wind" is "bej". The reflexes have a vowel that's either "e" or "a" -- unround, non-high, not too far back. If P-Y had "ə", it would probably use it in this word. But since it doesn't, "e" or "a" it is.
https://en.wiktionary.org/wiki/Reconstruction:Proto-Yeniseian/bej
The cognate in Na-Dene is P-Athabaskan "tšʼʷəy" = "wind blows", which has a labialized stop / affricate as 1st C, then "ə", then semi-vowel "y". Since Na-Dene lost labial stops at the earliest stage, a labialized C may have been their way of preserving a labial stop from a Pre-PND stage. Na-Dene does have "ə", so it's used here. So, the Dene and Yeniseian forms are semantically and phonetically very close.
https://en.wiktionary.org/wiki/Reconstruction:Proto-Athabaskan/t%C5%A1%CA%BC%CA%B7%C9%99y
In P-J, "wind" is expressed by a compound "kasai" (later "kaza / kaze"), which Wiktionary notes may be "ka" = "smell" and "sai" = "rapids", meaning "flow of smell".
Putting that aside as not phonetically related to "bej", we find the real cognate, the verbal root "poi-" = "to dry". In pre-modern societies, and sometimes even in modern ones, drying is done by "airing" it out, that is exposing it to the wind, like a primitive blow-dryer.
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/poi-
The OJ form is simply "pu", with obligatory verb ending "u" making it impossible to tell what other vowel used to be after "p". But it evolved into an intransitive form, "piru", and a causative form, "posu" (to cause to dry out or dessicate, like making food dry for preserving). That is a vowel alternation pointing to "poi-" as the verb root. The "ru" in "piru" is the default verb suffix, so "r" is not part of the root. And the "su" in "posu" is the causative suffix, so "s" is not part of the root either.
P-J lost voiced obstruents, so the only labial stop is "p".
Given that the Na-Dene form has "ə", and the P-Y reflexes can't decide between "e" and "a", and P-J did have "ə" in its inventory, it was probably "ə" in Wa. Then as in the P-J word for "fire", this is a sporadic case of "p" spreading its rounding forward, to turn "ə" into "o".
Codas are banned, so "j" was vocalized into "i".
Technically, there is no vowel sequence "əi" or "oi" in the P-J forms, so preserving both vowels in the mental entry for this word, does not violate the ban on vowel sequences. The root is listed as "p(o/i)-", or perhaps the full words were listed as a pair "piru, posu". In either case, no vowel sequence.
So unlike other cases where a vowel sequence is resolved by the 2nd replacing the 1st, here both vowels were preserved in the lexicon, though in a split entry or in 2 linked entries, but not as a single word with a vowel sequence.
The Korean cognate means "to blow" (said of a person, or the wind as well), which semantically overlaps "wind" and "to dry". In Modern Korean it's "bulda", where "da" is the verb suffix, and whose root is "bul". Before voiced obstruents, it was "pul", and before "r" merged into "l", the final C was "r", given the Chinese character used to write a derived form (see below). In Middle Korean, the vowel was "ǔ".
ReplyDeletehttps://en.wiktionary.org/wiki/불다
As for the noun "wind", the OK form is "polom", with the same Chinese character as OK used for "fire", meaning the 1st syllable ended in "y" or "r", not "l". The final "m" is the common OK nominalizer that attaches to verbs, in this case the verb meaning "to blow", producing "the thing that blows" -- i.e., "wind, draft, or other air movement".
In the complex form meaning "wind", the "o" is a regular ablaut variant of "u" from the basic verb "pulta". Later, the "u" rounded to "pwulta", as the "p" spread its rounding forward. This forward-spreading of a labial C's rounding is sporadic in W-D-Y, not systematic.
https://en.wiktionary.org/wiki/%EB%B0%94%EB%9E%8C
https://en.wiktionary.org/wiki/%E1%84%87%E1%86%9E%E1%84%85%E1%86%9E%E1%86%B7#Middle_Korean
Assuming, as with P-J, that the Wa form had "ə" as the vowel, the phonetic derivation is simple. No voiced obstruents means "b" became "p", and "ə" remained as near as it could (in P-K, more like the slightly further-back form "ʌ", still unround and mid-height).
Since there's no lost velar C, as there was in the case of "fire", the OK word must have ended in "y" originally, as in the reading "bwojH" of the relevant Middle Chinese character. IDK why it semi-hardened into "r" in both "fire" and "to blow", but the P-K form must have ended in "y", and developed to "r" in OK or MK. Word-final "y" matches the original Wa form, so no change needed.
So, both the Korean and Japonic forms trace back to the Wa form "bəj", almost exactly the same as in P-Y, but with "ə" as in the Na-Dene cognate. P-Y was unique in eliminating the most indistinct of vowels -- unround, mid-height, neither front nor back, "ə". It forced a choice between vowels that were extreme and clear to hear in at least 1 dimension, like "e" (maximally front) or "a" (maximally low).
Another great big etymological love-fest in Wa-Dene-Yeniseian land... ^_^
One final natural element word, "earth, land", which in P-Y is "baŋ". What's with all these natural element words beginning with "b" and an unround / non-back vowel? Possibly like Funnel Beaker sound symbolism, which survied mainly in Germanic. E.g., onset + nuclear vowel "dri-" relates to falling liquid ("drip", "drizzle", "drink", devoiced variant "trickle", etc.). Anyway...
ReplyDeletehttps://en.wiktionary.org/wiki/Reconstruction:Proto-Yeniseian/ba%C5%8B
Its proposed cognate in Na-Dene is P-Athabaskan "ŋənʔ", meaning the same thing.
https://en.wiktionary.org/wiki/Reconstruction:Proto-Athabaskan/%C5%8B%C9%99n%CA%94
The cognate in P-J is easy to spot -- "pama", meaning "beach".
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/pama
Japonic speakers were surrounded by beaches in the southern Korean peninsula, but more importantly, this was the main sign of "land" as they became seafaring and traveled among, and settled into, the Japanese and Ryukyuan islands. It's like in English, when sailors exclaim "land, ho!" when they spot land as opposed to the water they've been traveling on for so long.
The main form of "land" that Japonic-speaking seafarers saw, was semantically narrower, a "beach". But the narrowing of the W-D-Y word, from "land" to "beach", is totally expected for seafarers.
The final "a" is a dummy. In OJ, the most common "m" syllable, "mo", is already taken by the non-exclusive topic particle. If that had been allowed, "pamo" would sound like that particle attaching to the noun "pa", which has 3 noun meanings in P-J ("tooth", "leaf", and "blade"). Talk about confusing! They avoided that, and chose between the next 2, which are tied, "ma" and "mi", and probably favored "ma" in order to make both vowels harmonize / be identical.
Starting from the P-Y form, "b" -> "p", and "a" is no problem. Final "ŋ" in P-Y should correspond to "n" in P-J. The fact that P-J has labial "m", means the ancestor to all the families must have been labialized as well, "ŋʷ", which de-labialized in P-Y, and became primary labial nasal "m" in P-J. The P-A form ends in "nʔ", also hinting at some exotic back nasal. Presumably the P-ND form began with a labialized nasal, which became delabialized to "ŋ" in P-A.
In P-Y, "ŋʷ" is rare / non-existent outside the affixes. In lexical roots, there's only "m", "n", and "ŋ". So in lexical roots, "ŋʷ" must have been de-labialized -- and Japonic can point out which cases those are, when P-J has "m". If P-J has "n", then P-Y "ŋ" was always "ŋ", not a de-labialization of "ŋʷ".
ReplyDeleteThe Korean cognate is easy to spot as well, "pan", meaning "place for events" and somewhat more abstractly, "situation, state of affairs".
ReplyDeletehttps://en.wiktionary.org/wiki/%ED%8C%90#Etymology_2
Phonetically, "b" is devoiced as always to "p", and "a" is not problematic.
As we saw with P-WDY "xʷaj" = "sun", which became "hoy" in P-K, Korean does not convert secondary labialization into primary labial place, it simply de-labializes the phoneme. In the present case, that yields "paŋ", but Korean did not get "ŋ" as a phoneme until recently, so it was shifted to the nearest nasal, which is "n", yielding "pan", QED.
Semantically, it was extended from a natural element term like "earth, land, soil" to a generic "place" term -- but one involving terra firma, not events unfolding at sea or way up in the air or the heavens or in the watery depths of the Underworld.
In English as well, we extend the natural element word "ground" to mean a place where events unfold -- "fairground", "battleground", "stomping ground", and so on.
Abstract words like "situation" and "state" come from Latin words that also imply unfolding on land, since "situs" comes from the verb "sino", meaning "to place, lay down", and "sto" means "to stand". So that abstract sense in Korean also relates back to "ground". Its a figure of speech, using "ground" to refer to the events unfolding upon that particular area of ground.
This semantic shift is not the same as the one in Japonic, which involves seafaring and the beach as the place where land contrasts with water. What both Korean and Japonic share in common is a basic natural element sense, "land, earth, ground" -- the same sense as in Yeniseian and Na-Dene. So that must have been the meaning of the Proto-Wa form, when it was still sister to Yeniseian and Na-Dene.
Neat. ^_^
To be exact, the ancestral form of this word in Proto-Wa-Yeniseian (leading to P-Y, P-K, and P-J), is "baŋʷ". Both P-Y and P-K de-labialize "ŋʷ" -> "ŋ", while P-J converts it to primary labial nasal "m". That's it for the P-Y derivation, and I've already spelled out the derivations from there to the P-K and P-J forms. Just want to be explicit about the ancestral form ending in "ŋʷ".
ReplyDeleteWhere the Na-Dene cognate ties in exactly, IDK, there's only a P-A form, and the closest match in Tlingit is "aan" = "inhabited land, settlement". That looks like the same de-labialization of "ŋʷ" and shifting forward to "n" after losing "ŋ". The initial labial has been simply deleted, rather than converted into another labial like "w". Perhaps the sound laws to deal with nasal stops were different at different periods of Tlingit or broader Na-Dene history, and when this word was dealt with, the solution was "just delete the labial".
In any case, the P-ND form was likely a resolution of the very same word in Proto-Wa-Yeniseian. That means "baŋʷ" = "land, ground, earth" allllll the way back to Proto-Wa-Dene-Yeniseian. Well, it's one of the most basic words in any language -- it *should* go that far back, fairly unchanged!
But maybe a better understanding of the Na-Dene side will show that the P-WY form is not sufficient to be the ancestor of all 3 families. But for now, it does seem to be the most parsimonious way to reconstruct it back to P-WDY, and the P-ND form is mainly going to be coping with all those labials, which it doesn't like.
To commemorate the debut of Hololive's next generation of vtubers tomorrow, let's look at the Wa-Dene-Yeniseian origin of one of the words in their name, "Asobi Mawari", meaning "playing and hanging around". Specifically, the verb "mawaru" (P-J "maparu"), which means "to spin, revolve, go around" and in modern times, "to roll" when said of movie cameras.
ReplyDeletehttps://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/mapar-
Related is the verb "mapu", meaning "to move in a graceful circular motion, to whirl".
https://en.wiktionary.org/wiki/%E8%88%9E%E3%81%86#Japanese
The root for both verbs is "map-", and the latter takes the simple verb ending "u", while the former is a somewhat extended form with the default verb suffix "ru" after a dummy vowel "a" ("pa" being the most common "p" syllable in OJ).
This P-J root is cognate with the P-Y word "paxʷ" = "a flat and round object". The proposed cognate in Na-Dene is P-Athabaskan "ʔaˑxʸ", meaning "showshoe".
https://en.wiktionary.org/wiki/Reconstruction:Proto-Yeniseian/pax%CA%B7
https://en.wiktionary.org/wiki/Reconstruction:Proto-Athabaskan/%CA%94a%CB%91x%CA%B8
Note that roundness here means a circular perimeter, which goes around an object that is flat. So, like a disc. Some languages have shape words or classifiers or counters, where roundness means something spherical or cubical or having 3D solid bulk, while flatness means it's simply a 2D sheet, and another word referring to length means it's a 1D line. In Yeniseian and Na-Dene, though, both flat and round are combined in this shape word, meaning disc-like.
When a person or other object "turns", "revolves", "spins", or "whirls", it sweeps out a cylinder. That gives it a circular perimeter, and unless the object is quite large, it also has minimal height, and so is flat as well. At least, that's what your attention is focusing on -- the roundness applying to the perimeter only, without a curve in the height dimension as well. It's not spherical or bulbous or bulky, but sweeping out a thin graceful disc shape.
In fact, most objects whose defining motion is to "turn, revolve, spin, whirl, etc.", are disc-shaped to begin with -- like a wheel, a gear, a roll of film in a movie camera, the arms of a windmill, the blades of a mechanical fan, and so on and so forth.
That establishes the tight semantic fit between the Yeniseian and Japonic roots.
Phonetically, "paxʷ" wants to convert secondary labialization into primary labial place, and since the original base consonant is an obstruent, it wants to make "xʷ" into "p", yielding the root "pap-", with dummy vowel "a" after, yielding "papa".
ReplyDeleteHowever, Japonic has a rule against the 1st and 2nd C's of a basic lexical root being the same. Martin first studied this systematically in the 1980s, and is now a standard observation about Japonic phonotactics going back to P-J.
The only exceptions are usages where reduplication is deliberately chosen for sound-symbolic or similar effect -- nursery words like "papa" and "titi", critter names like "memesu" ("earthworm"), and so on.
Even the later explosion of reduplication for verbs of motion, adjectives and adverbs of fine-grained aspect, etc., involve reduplicating an entire CVCV base, where the 2 C's are different, like "pachi pachi". Not reduplicating a CV base to yield a standard word-shape of CVCV, whose 2 C's are the same.
So, 1 of the 2 "p"s must be altered, and the natural alternative is "m" ("w" would do as well, but it is more constained phonotactically by what vowel can vowel it, unlike "m" which takes any vowel, so "m" is a better choice).
If the 2nd C were changed to "m", that would yield root "pam-", which runs into the same overcrowded homophone problem that we saw for "pama" = "beach", and this time the dummy vowel "a" has already been taken by "pama".
So instead, it chose to alter the 1st C to "m", since there are no competing P-J words beginning with "map-" ("mape" = "before, front" is a compound of "ma" = "eye" and "pe" = "toward", and in any case it ends in "e", which will never be a dummy vowel since it's so rare, following any consonant).
That leaves the root as "map", QED. ^_^
Notice that initial "m" in Japonic came from earlier "p" in Wa. I don't think this is the sporadic "p" -> "m" shift, since it's phonotactically motivated by the ban on identical C's in a typical root. Still, it emphasizes the need to figure out if "m" in Japonic was always "m", or came from something else. There are tons of "m"s in Japonic, suggesting that quite a few of them came from some other sound originally.
And therefore, the Yeniseian form beginning with "p" is not avoiding initial nasals -- it was never nasal to begin with, and the P-J form with "m" is misleading, since it came from earlier "p".
But wait, we still need to invite Korean to this word-origin love-fest! And they did not disappoint... ^_^
ReplyDeleteThere's one phonetic form that is used for 2 similar disc-shaped nouns. In Modern Korean it's "bakwi" = "wheel", and also serves as a counter word for the number of turns, cycles, rotations, etc.
https://en.wiktionary.org/wiki/%EB%B0%94%ED%80%B4
Earlier, in Middle and Old Korean, it meant "rock", reconstructed in OK as either "PAhwo" or "PAKwo".
https://en.wiktionary.org/wiki/岩乎
If you think of the typical kind of rock you see when you're out and about IRL, they have rounded edges or perimeters, and tend to be flat on the top and/or bottom. Not necessarily perfectly smooth tops and bottoms, like ideal cylinders -- but still not much curvature going on in the height dimension, as opposed to the footprint or base area dimensions.
Rocks' relative flatness is what makes them easy to skip across a pond, or able to stack on top of each other, which would be impossible if their shape were spherical, bulbous, or bulky.
Most rocks that you see in nature are sedimentary, which are formed in horizontal layers. When those layers break up, and individual rocks are produced as fragments, they tend to inherit the flatness on top and bottom from the layered structure that they came from.
Also, when they're resting in a moving water body like a river or stream, most of the water rushes over and under them, eroding their tops and bottoms into flat shapes. Around their edges, the water is diverted around either side, giving it a more circular perimeter.
Although not ideal discs, the typical rock is more of a disc than a sphere or a "long" line or a flat shape with rectilinear edges.
We know that Korean thought of "rock" as a flat and round object, since it later applied the same root to "wheel", which is one of the most prototypical "flat and round objects, whose defining motion is turning and revolving", in human history.
That establishes the semantic tight fit between Korean and the Japonic, Na-Dene, and Yeniseian cognates.
Phonetically, Korean has no problem with "p" and "a", only the final "xʷ", which as we've seen many times before, it will simply de-labialize into "x". That velar fricative is written "h" in Korean studies, and that is one of the two possible readings of the 2nd C in the OK word, i.e. "PAhwo".
ReplyDeleteIn the history of Korean, there was a broader sound change of softening "k" to "h", but that is not necessary here, since the Wa word originally had a velar fricative at the end.
The only wrinkle is why the Korean form has a final vowel -- it does not have obligatory CVCV shape like Japonic.
Perhaps it wanted to preserve the labial feature from "xʷ" as a separate vowel appended to the end, and "wo" is a perfectly good back-round vowel. It didn't do this when "xʷ" was word-initial, since there's already a vowel following it. But word-finally, it has the option to split the labial feature off into a back-round vowel, and preserve the nature of the consonant base it was attached to, namely "x / h". The point is, it didn't pursue the distinctly Japonic strategy of converting the "ʷ" into primary labial place, i.e. "p".
Linguists studying Korean do not assume the "wo" is a suffix of any kind, and yet Korean words strongly prefer word-shape CVC. So it makes perfect sense to think that OK words with final vowels like this represent the spinning-off of a labial feature attached to a consonant, from an earlier stage, such as "xʷ".
Later, the Korean word for "rock" evolved to "bawi", slightly contrasting with "bakwi" = "wheel", but they are doublets that come from the same P-K root, which itself comes from a Wa root, and ultimately Wa-Dene-Yeniseian root, that refers to objects that are flat and round.
To every word, turn turn turn
There is an origin, turn turn turn
And a sound to every meaning, in the head-room
Following up on Japonic words relating to "round", though not flat. There are 2 words in P-J that seem to be related to "map-" = "to spin, whirl, revolve, etc." First is "mari" = "ball", which *is* spherical.
ReplyDeletehttps://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/mari
And the related adjective / noun in OJ, "maro" = "round; round thing", with the connotation of a bulky 3D sphere, since another sense is "thick, plump" -- not thin and flat. This later shifted to "maru".
https://en.wiktionary.org/wiki/%E4%B8%B8#Japanese
Second is "maku" = "to wrap around", and related "makuru" = "to roll up", which is attested through its resultative noun form "makura" = "pillow". Both come from "mak-".
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/mak-
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/makura
These 3 roots, "map-", "mar-", and "mak-", all share a sense of "round" and even a verb involving motion typical of round objects. Yet phonetically they only share "ma", with 3 different C's afterward. So what does "ma" mean?
It is a spatial term, but nothing to do with roundness or any shape for that matter, and nothing to do with motion. It means "room, space" (and a homophone meaning "truth, reality").
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/ma
Since "ma" has nothing to do with shapes, then the "p", "r", and "k" additions would have to bear that meaning -- yet they cannot, since those C's are all different. They clearly must provide subtle differences of meaning, upon the shared sense of "round". But "ma" does not refer to "round" or any shape. So, this derivation from "ma" = "space" cannot work.
Well, how *are* all 3 of those roots related, then?
Here's where the Wa-Dene-Yeniseian framework settles the issue. Only 1 of them has cognates in the other branches of its family, which refer to "round", namely "map-". This one is primitive, and the other 2 derive from it. How?
ReplyDeleteI think Japonic speakers reanalyzed the verb "mapu" = "to whirl" as having root "ma" and verb suffix "pu", which is not the default verb suffix, but was a verb suffix for iterative / repetitive actions, which attached to the irrealis stem, which is "a".
https://en.wiktionary.org/wiki/%E3%81%B5#Etymology_2
Turning, whirling, revolving, etc., are all iterative actions, and they are used for cycles -- things that repeat. This semantic contamination made them think the root was really "ma" instead of "map", and that the suffix was "pu" instead of "u".
After reanalyzing the root as "ma", they assigned it the meaning relating to "round" or "to turn, revolve". But it was still a verb root, not a standalone word. So they attached the most common default verb suffixes "ru" and "ku", yielding "mar-" and "mak-".
It's striking that the 2 forms that do *not* have cognates in W-D-Y, have the default verb suffixes, while the one that *does* have cognates, has a more unusual final consonant "p". That argues for "map-" being primitive, and "mar-" and "mak-" as being derived from it (after reanalyzing the root to be simply "ma").
So not only does this explain how these 3 are all related, it explains which one is primitive, and why the derived forms have the exact additional C that they do -- "r" and "k" belong to the most common / default verb suffixes, which attach after a vowel stem (such as "ma").
This suggests that the alteration of "p" -> "m", from the original Wa root, happened very early. It had to have happened before P-J, since it led to the reanalysis of its structure to derive 2 separate "round" words in P-J, one of which is itself a derived form ("makura", from "makuru + -a"). That takes a lot of time.
Well, so be it -- there's already 2 subtly different forms from "map-" -- "mapu" and "maparu". So clearly this shift had begun early enough for that -- that's early enough for the reanalysis and spin-offs "mar-" and "mak-".
To be exact, I think "mari" = "ball" derives from the verb "maru" = "to roll", with the "i" ending for its continuative form, similar to the gerund suffix "ing" in English. Both can derive nouns from verbs, just like "shaving" is a noun derived from "shave" (a piece that is shaved off of the object being shaved).
ReplyDeleteThe verb "maku" is already listed in P-J, though "maru" is not as a verb, only through its continuative form "mari" = "ball". But it must derive from a verb, "maru", with the default verb suffix "ru".
Korean has 말다 mal-da for 'to roll up', including things like kimbap/sushi rolls, do you think this could be somehow traced to the common Koreo-Japanese-Dene-Yeniseian root, or a later borrowing/substratum from Japonic based on the m- initial?
ReplyDeleteIf you end up deciding that Koreanic and Japonic are directly derivable from a common lang family with Dene-Yeniseian without having to explain them as creole outcomes, then I think we are due for some cladistic terminology coinings because otherwise the language family name is getting a bit too unwieldy with more than one hyphenation.
Koreanic+Japonic can be Koreo-Japonic which has been used before, but Koreo-Japonic-Dene-Yeniseian is a bit too long, and Wa/和 by itself is a bit short to encompass both K and J, and since the Ryukyus are not usually considered Yamato in sensu stricto directly, it is also a bit short for Japonic. Perhaps Boreo-Pacific or something of a geographical nature, in case you discover any other small linguistic phyla to lump into this family (maybe ancient Buyeo languages or something).
With your recent spree of identifying phylogenetically-fused origins of various linguistic phyla, I would be curious if you have thought about whether the entire spectrum between heavy substratum to creoloid to simplified creole to radically grammatically-restructured creole to full-blown mixed language creole needs to be described and delineated with more academic rigor so that we have more precise terminology to describe linguistic contact phenomena. For example, there is an immediately noticeable difference between a creole like "Hawaiian Creole English" (which is generally intelligible for many native English speakers), vs. a very restructured English-based creole like Tok Pisin or Bislama, vs. calling Japonic a "creole" of Dene-Yeniseian+Ainuic or Turkic a "creole" of Samoyedic/Yeniseian, vs. what you see in a hybridized language like Wutun in China. Do you have a set of criteria in mind for determining whether for example Turkic should be considered a creole of Samoyedic and not directly classified under Uralic, while Japonic should be directly classified as part of the same lang family as Dene-Yeniseian?
Wa-Dene-Yeniseian word for "to take" and "to take away" (by a simple semantic shift, as in English "take" evolving to mean "to take away"). The original sense, as in "to hold in the hand", is in Proto-Yeniseian "ɢaš" = "to take", and a Proto-Na-Dene form something like "ɢats'", exemplified by Eyak "ɢats'" = "to handle" and Proto-Athabaskan "ʔaˑ" = "to handle". This pair of cognates has already been identified by Vajda in the Dene-Yeniseian framework.
ReplyDeletehttps://en.wiktionary.org/wiki/Reconstruction:Proto-Athabaskan/%CA%94a%CB%91
In the Northern branches of Yeniseian, this reflects as "as", with the uvular stop deleted and palatal sibilant brought to alveolar place. I showed before that Korean and Japonic both seem to delete the uvular voiced stop (although perhaps Korean preserves the voiceless one, "q", and shifts it forward to "k"). So that treatment of "ɢ" is paralleled within the Yeniseian family as well.
Old Korean has a verb "As" = "to take away, remove, subtract", which is phonetically identical to the Northern Yeniseian reflexes. Not cuz Korean is an offshoot of N.Y., just that these are common solutions to the problem of an exotic back consonant like "ɢ", and merging a palatal sibilant into the common alveolar one. This evolved into Middle Korean "as-ta" and Modern Korean "at-da", with the same meaning.
https://en.wiktionary.org/wiki/%E5%A5%AA%E5%8F%B1
The semantic shift is highly common, "to take" -> "to take away", without any further morphology, as in English "to take", which also means "to take away, remove, steal, etc." The simple sense of "to hold in the hand" shifts to "to possess" in some indefinite state, not literally holding it in the hand 24/7. And how did you come to possess the thing? Unless you found it lying around, or made it yourself, or exchanged something of yours for it, it's likely you caused a change in ownership -- you took it away from its former possessor. Very natural semantic shift.
Tying this back into the new Hololive generation, the first word of their name, "Asobi", comes from P-J "asopu" = "to play" (no evidence of a medial nasal, and the later form "asobu" is due to reanalysis of morpheme boundaries and rendaku).
ReplyDeletehttps://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/asump-
This has already been analyzed as a compound verb by others, but they propose the "as-" element is related to Modern Japanese "asai" = "shallow, superficial". That is, "to play" is "to wander in a shallow, superficial manner", leisure vs. purposive productive labor.
https://en.wiktionary.org/wiki/%E6%B5%85%E3%81%84#Japanese
However, I think "asopu" combines two distinct form of play activities, both of which are prototypical games that people play for fun. First, the otherwise unattested "asu" = "to take away possession of" (like soccer, football, capture the flag, etc.), and P-J "opu" = "to chase" (like racing, tag, and half of the games like soccer and football, where taking possession of the object requires you to chase after its current possessor).
https://en.wiktionary.org/wiki/Reconstruction:Proto-Japonic/op-_~_%C9%99p-
The combined sense originally meant "to chase and take possession of", then lexicalized into a simple sense of "to play, amuse oneself, have fun".
When forming compound verbs, the first one takes the "i" ending, and the second remains in the "u" ending. In this case, "asi + opu", but since P-J prohibits vowel sequences, and outside the 1st syllable the 2nd V replaces the 1st V, yielding "asopu", QED. ^_^
Again, the "asu" verb in this compound is not otherwise attested, but is cognate with the Korean verb "As" = "to take away", and the Dene-Yeniseian verbs "to take". The Japonic form simply adds the obligatory verb ending "u" to the root "as".
I think this sense of "to take away possession" fits better into a compound verb meaning "to play", since one of the most prototypical games that all cultures play to amuse themselves, is competing over possession of an object. It's also better anchored in the ancient language family that Japonic belongs to, whereas Modern J "asai" = "shallow" does not even go back to P-J, as far as we know.
I wasn't even planning on figuring out the other half of the new Hololive generation name -- I literally just picked an Old Korean word at random from Wiktionary, and looked for its Dene-Yeniseian cognates, and then its Japonic cognate... which led me to "asopu".
Randomly chosen -- or guided by the hand of Fate? I don't believe in accidents! ^_^
Korean "mal-da" = "to roll" looks like a borrowing from (Peninsular) Japonic. It's only attested in Middle Korean, although it used to have a different vowel, "o" (unround back mid), rather than "a" (unround back low).
ReplyDeleteThere's an OK auxiliary verb "MAl", but it means "to not do X", nothing to do with roundness. And that one ends in "l", back when Korean distinguished "r" from "l", so it's not phonetically related to P-J "maru" either.
But by Middle Korean, they'd merged "r" and "l", so if they borrowed "mar-" from Japonic, it would show up as "l".
Given that "mar-" is in P-J, that places it in the Korean peninsula, early on, and so with plenty of time for Korean to adopt it from the locals.
I like "Wa" for the ancestor shared by Korean and Japonic since it's ancient, refers to Japanese, and although the Ryukyuan islands may be culturally distinct in other ways, linguistically they speak a sister to Japanese.
And as I've emphasized with religion, mythology, and other cultural shibboleths, Yamato / Japanese is part-Wa, but also part-Emishi / Ainu. Like the Sanzu no Kawa -- that comes from Ainu, and is part of their Uralic heritage of the Emishi / Ainu. The same view of the Underworld shows up in Finnic and other Uralic cultures, and it's unique to them and the Ainu, much like the bear cult. I'll get around to the Uralic origins of Ainu sometime soon, when I'm not distracted by these current investigations...
So, "Wa" covers the non-Emishi/Ainu parts of Japanese culture, and so that would apply to Korea as well, where Emishi/Ainu influence is far less.
And "Wa" and "Dene" are pretty short, so Wa-Dene-Yeniseian isn't that long. If there were a clearly bounded geographic region, or way of life, I'd use that. But they're very spread out, even just the Old World ones, and don't seem to have a simple unifying lifestyle like "horse archer" or something, or a unifying material culture -- unless we go with "Mirror Breakers". ^_^
As for creole terminology, I'm using it broadly. When I need to be more specific, I'll say Turkic is "Samoyedic spoken with a Yeniseian accent", or Japonic is "Wa spoken with an Emishi accent", and Korean is "Wa spoken with an Eskaleut accent".
Whereas Tungusic is so lexically Eskaleut, and it's not just some foreign accent applied to it, it's like they took something of an Altaic base accent, but then also inverted the feature redundancy of Eskaleut consonants... not a single accent filtering the target language.
These particulars depend on the case at hand, and some may belong to a sui generic pattern like Tungusic. I don't really care what the jargon is, as long as the substance gets across.