Prakritic "Kroraina" and Old Sinitic reconstructions of "Loulan", part 2
« previous post | next post »
What follows is Doug Adams' draft of an excursus that is not trying to be complete in itself (i.e., it's not a free-standing article), but rather something that will provide a certain amount of orientation to readers of the review of Schmidt's Nachlass (for which see the first item in the "Readings" below).
[Excursus: The Name of Lóulán/Kroraina: It is universally assumed (1) that Lóulán (the contemporary Chinese pronunciation of the relevant Chinese characters) and Niya-Prākrit Kroraina (Sogdian krwr’n) refer to the same place[1] and, further, (2) that they are, at bottom, the same word. In discussions of Lóulán/Kroraina, Lóulán is confidently given the earlier (Old/Middle?—the age is not usually noted) Chinese pronunciation of *γləulan or the like (Schmidt gives *γlaulan). Since Middle Chinese (ca. 600 AD) /l/ is known to reflect Old Chinese (ca. 1000-200 BC) /r/, it would seem to be a short hop to a reconstruction of *γrəuran in, say, 500 BC.
But, in discussions focused on Chinese alone, Schuessler (2009) reconstructs Old Chinese *ro and *ran for the two characters and Baxter (1992) reconstructs *ro and *g-ran. There were Old Chinese syllables of the shape *γrô, *grô, and *krô, but in Schuessler’s (and Baxter’s) opinion they would have given modern Mandarin hóu, góu, and kóu (via Later Han Chinese *go, *ko, and *kho, and Middle Chinese γəu, kəu, khəu (Schuessler, 2009:146-147). What you do not get is *Kro > **Klo > ***lóu. Indeed, there is widespread (not universal) agreement that lán goes back to an Old Chinese *g-ran, but that lóu goes back to simple Old Chinese *ro, with no velar “prefix.” (Some researchers would take the equation of lóu with the first syllable of Kroraina as by itself sufficient to reconstruct Old Chinese *g-ro, but most would not.) Not knowing exactly when the “prefix” was lost, we could most reasonably reconstruct (a late) Old Chinese *ro-g-ran or *ro-ran.[2] A reconstruction *g-ro-g-ran might be possible, but NOT **g-ro-ran, which Prakrit Kroraina would seem to demand. The answer may lie in the supposition that we do not see the result of direct borrowing by Old Chinese of a Tocharian C place-name, but rather the complicating effect of the transmission of that name via one or more (unknown) intermediate languages. In any case there appears to be no warrant to reconstruct an Old Chinese *γləulan as the antecedent for Lo-lan of the Han era.[3]]
[1] So already in the latter part of the nineteenth century by Karl Himly (1836-1904), as reported in Conrady (1920:162).
[2] If a Chinese *rəuran, or the like, had existed at some time in the second half of the first millennium BC, it would have been an obvious source for the Khotanese equivalent, raurana-, given by Schmidt. However, one might note that this Khotanese raurana-, given confidently by Schmidt, does not appear in Bailey (1979), nor in Bailey (1985) which is substantially devoted to Khotanese geographical terms from Kashgar to Lop Nor. Until it is better supported, we should perhaps treat it as a vox nihili. [VHM: The Khotanese is actually raurata (Staël-Holstein scroll apud Kumamoto in the o.p.)] Ptolemy’s Chaurana, which is sometimes brought into this discussion, is best left aside as well. If it refers to anything real at all (and Ptolemy’s information on Inner Asia is most sketchy), its position suggests Khotan rather than anything further east (see Stevenson, 1932:145 and Asia, map 8).
[3] The discussion in this subsection has been much aided by the material gathered on-line in Victor Mair’s Prakritic “Kroraina” and Old Sinitic reconstructions of “Loulan”, to which the reader is referred for much fascinating information.
=======================
Readings
"Prakritic 'Kroraina' and Old Sinitic reconstructions of 'Loulan'" (5/14/19)
"Tocharian C: its discovery and implications" (4/2/19)
"Tocharian, Turkic, and Old Sinitic 'ten thousand'" (4/23/19)
Jonathan Smith said,
May 21, 2019 @ 8:53 am
Old Chinese in the context of a discussion like this is like me in the locker room with fit people :D brutal
A couple things are not quite right:
re: "There were Old Chinese syllables of the shape *γrô, *grô, and *krô, but in Schuessler’s (and Baxter’s) opinion they would have given modern Mandarin hóu, góu, and kóu (via Later Han Chinese *go, *ko, and *kho, and Middle Chinese γəu, kəu, khəu (Schuessler, 2009:146-147). What you do not get is *Kro > **Klo > ***lóu."
Not really, because Schuessler and Baxter happen to use notational "g-r" to mean "like 'gr' or 'kr' but gives Middle Chinese l-", which is a thing that no-one has a great explanation for.
re: "Schuessler (2009) reconstructs Old Chinese *ro and *ran for the two characters"
that is true but Schuessler (2007: 80, 343) has *g-ran (note "g-r"!) for 'orchid' due to the doublet noted in the other thread. Given the extensive discussion in 2007, maybe this makes the better source.
re: "A reconstruction *g-ro-g-ran might be possible, but NOT **g-ro-ran, which Prakrit Kroraina would seem to demand."
Not sure if anyone would go out on that branch :C The above could in fact make sense if there was really a body of native "CrV-rV"-type reduplicative vocabulary spelled in similar ways, as some have suggested.
What is actually "fit" about early Chinese is, among other things, that medieval onset l- came from earlier *r-, so using Schuessler-2009-style "Late Han" with two r- onset syllables would be fairly unassailable (= at least leaving our pants on.) Beyond that, too many wheels are in spin… which would be true even if we were actually talking about the two independent lexical items 'orchid' and 'building' which unfortunately we aren't…
David Marjanović said,
May 21, 2019 @ 6:24 pm
Not that it matters here, but I'd expect hóu, gōu and kōu: absent complications that don't apply here, the first tone of Mandarin corresponds to (Old and) Middle Chinese voiceless onset consonants, the second tone to voiced ones.
Chris Button said,
May 21, 2019 @ 11:16 pm
But he also entertains the notion of it being a "copying error" (one of the possibilities proposed by Baxter) since the supposed doublet 蕳 is also associated with 蓮. With little else at our disposal, I see no convincing reason to overturn the basic reconstruction of 蘭 as *rán in accordance with its later development.
It just reflects the possibility of a dropped prefix, which happened from time to time and is not typologically bizarre. Theoretically 樓 and 蘭 could both have had earlier prefixes since they have liquid onsets at their roots (in front of which such cases can be found), but where's the compelling evidence?
Jonathan Smith said,
May 22, 2019 @ 7:23 am
@ Chris Button My post including reference to "*gr" vs. "*g-r" was about lack of regularity or our failure to find it, not an argument for some "prefix" or other… I think we are in agreement that *r- + *r- would be a pretty objectionable way to represent this item.
Victor Mair said,
May 22, 2019 @ 8:28 am
From Hiroshi Kumamoto:
De la Vaissière's remarks (Exegisti Monumenta Fs. Sims-Williams 2009, 530-531) with references to the new edition (Humbach / Ziegler 1998, 2002) should not be overlooked.
To be added there is:
Maria Gabriela Schmidt, Die Nebenüberlieferung des 6. Buchs der Geographie des Ptolemaios. Griechische, lateinische, syrische, armenische und arabische Texte. Wiesbaden 1999
Jonathan Smith said,
May 22, 2019 @ 8:52 am
*pretty UNobjectionable :/
Victor Mair said,
May 22, 2019 @ 1:07 pm
From Axel Schuessler:
Perhaps I may add a comment to the OC proposals for 蘭. Those who try to understand and reconstruct OC are faced with a number of choices. E.g. what weight to put on phonetic series, vs. on actual language, i.e. on the Qieyun System (MC) or modern Sinitic languages? If one privileges phonetic series and graphic loans, then ‘orchid’ would perhaps be OC *g-rân because of the phonetic and a few other words in the series had initial *kr-. Then one could also argue for *g-rô or *s-rô 樓 because the phonetic writes a few words with initial s- and k-, although the phonetic itself does not, therefore most agree that the OC initial consonant was *r-. If one gives priority to language (which I lately tend to do), then all we can say for “sure” is that lán, MC *lân typically came from OC *rân, assuming that writing developed because some one needed to write *rân and picked the graph jiǎn 柬 that might do to convey the idea of what word was intended. Another curiosity about lán is that the phonetic element had the rhyme *-en (not *-an), which would agree nicely with the syllable -rayn- in the name of the town. But this is just a wild thought, we may never know; but is shows once again that one should not rely too exclusively on graphic choices writers made in the first millennium BC.
Chris Button said,
May 22, 2019 @ 5:08 pm
@ Jonathan Smith
That's why I would go with *ʰráw.rán. The phonetic justification for a shift of *ʰráw to *ráw in the first syllable is very strong and far more convincing in my opinion than any notions of *kr- or the like. It also allows us to differentiate between the onsets of the two syllables, and it accords better with the *ɣ- of 谷 noted in the alternation of 東樓 with 東谷 in the first part of the thread. It also occurs to me now that the effects of some voicing assimilation across the two syllables in 東樓/東谷 would bring *ʰr- (later x-) even closer to *ɣ-.
@ Axel Schuessler
Chris Button said,
May 22, 2019 @ 5:09 pm
@ Jonathan Smith
That's why I would go with *ʰráw.rán. The phonetic justification for a shift of *ʰráw to *ráw in the first syllable is very strong and far more convincing in my opinion than any notions of *kr- or the like. It also allows us to differentiate between the onsets of the two syllables, and it accords better with the *ɣ- of 谷 noted in the alternation of 東樓 with 東谷 in the first part of the thread. It also occurs to me now that the effects of some voicing assimilation across the two syllables in 東樓/東谷 would bring *ʰr- (later x-) even closer to *ɣ-.
@ Axel Schuessler
Jonathan Smith said,
May 23, 2019 @ 5:39 pm
@ Chris Button I was interested in Adams' passage because we see "OC" as it gets presented in other/generalist contexts and think about what is or isn't worth saying in such a context. The whole *C(-)r problem is probably worth exactly one sentence saying it's not worth going into, your *ʰr- included: this is ad hoc for this particular item, is an irregular development, no MC x- in this XS, 樓~谷 can be simply explained via standard r~l, etc… not to say don't think imaginatively, but it's useful to keep in mind the distinction between solid ground and quicksand.
Chris Button said,
May 23, 2019 @ 8:52 pm
@ Jonathan Smith
While your argument is logical, it is important here to think like a linguist rather than a Sinologist (of course, further to my earlier comment, that is a problem with many reconstructions of Old Chinese in general).
The simple fact is that "voiceless" (here partially voiced) sonorants are inherently unstable (they're called sonorants for a reason!). To borrow John Ohala's words: "Devoiced, a nasal is a fricative". On purely comparative typological grounds, I have absolutely no doubt whatsoever that there are other cases of *ʰr- merging with *r- in Old Chinese (along with other sonorant pairs) as we see in other languages. But such cases are very difficult to identify since they would have merged without a trace. Here, through external evidence, we might have located one such case, but how can we ever know for sure? At least we have an account for why the two syllables of the Prakrit word have different onsets.
As for 谷, please don't give me *kl- as an onset and another disappearing k- prefix (which with this character appears to have only been preserved in one syllable type)! I would just make an appeal for parsimony by going with *ɣ- to account for both Middle Chinese k- and j- reflexes. Compare the xiesheng series of 羊 *ɣàŋ in which presumably you would favor *l- again? (https://languagelog.ldc.upenn.edu/nll/?p=42523#comment-1563026). As for 東樓/東谷, an alternation of /ɣ/ [ɰ] with /r/ is perfectly reasonable after the velar nasal of the first syllable and all the more, but not crucially, so with /ʰr-/. It's certainly on a par with *kl- with the claim that the velar k- was perhaps elided under the influence of the preceding velar nasal coda.
Chris Button said,
May 24, 2019 @ 6:05 am
Just to be clear, it's probably worth adding that 樓 will still be going in my dictionary with plain *r-. The proposal above for *ʰr- is just a tentative hypothesis–largely unverifiable–for the specific discussion at hand.
Jonathan Smith said,
May 24, 2019 @ 1:14 pm
To the extent possible we should continue to search for patterns and principles rather than, e.g., using "*ɣ- to account for both Middle Chinese k- and j- reflexes", which is a non-comparative fudge — though of course irregular developments are a thing and none of your proposals are unreasonable. But in a paper of this kind, and generally, I would simply emphasize methodological triumphs like *r- > l- and related points of agreement, and as for the rest, dai kao 待考.
Not sure about "this character appears to have only been preserved in one syllable type". If you mean Type A MC k-, Type B- MC y-, this is the kind of "pattern" I am interested in trying to find and interpret. No not the same situation as 羊, but… dai kao :D
Chris Button said,
May 24, 2019 @ 5:52 pm
@ Jonathan Smith
*ɣ- > k- in type A syllables and *ɣ- > j- in type B syllables. Hence 谷 two reflexes may be reconstructed as *ɣákʷ > kəwk and *ɣàkʷ > juawk. I'm following Pulleyblank's proposals here.
Now, if you're looking for a possible exception then 羌/姜 (in which I believe 羊 *ɣàŋ is phonetic) are type B with Middle Chinese kʰ-/k- reflexes (this is noted by Pulleyblank). However, the variation in aspiration is unexplained and it probably comes down to them being coined proper nouns. I would reconstruct OC *kʰ- and *k- for them (in accordance with Schuessler here) and assume some flexibility between *ɣ- and *kʰ-/k- along the lines of *ʁ- and *q-/*qʰ- appearing in the same series (uvulars and velars unsurprisingly have some overlap). Incidentally, you can't regularly reconstruct *kl- and *kɬ- (*kʰl-) since the aspirated counterpart would give a different reflex following Schuessler's proposals.
Jonathan Smith said,
May 24, 2019 @ 11:40 pm
Sorry, I missed what you meant re: "*ɣ-". Sure, if you like. The point here, or my point here, was precisely that if for OC Word X some say (e.g.) kl- but others say ɣ-, etc., etc., than that area of the would-be field is a hot mess, full stop, and should be flagged/deemphasized in discussion. But because such issues are so numerous and so unrecognizable, people have a devilish time acquiring basic more-or-less reliable information about a simple question like the one above.
I personally have no idea / am indifferent about MC K/y in XS… not sure if you are suggesting I love Kl here? Surely the father of this approach is Pulleyblank himself however? Again though — if the answer were really at all clear we would not have Baxter & Sagart now talking about "uvulars", etc.
Chris Button said,
May 25, 2019 @ 7:39 am
I believe the *kl- idea is from Baxter (1992) for the MC k- (from OC *k) and j- (from OC *l-) alternation. In this case, Baxter and Sagart's more recent uvular proposal (replete with phonologically unlikely *ɢ-) actually brings them somewhat more in line with Pulleyblank:
羊 *ɣàŋ > jɨaŋ (B&S *ɢaŋ, where Sagart (1999) has *laŋ)
However, they don't identify a type-A/B distinction and just assign an unspecified "C-" prefix to uvular *q- to account for the k- reflex, so ironically they end up complicating things further in my opinion. For example:
公 *ɣáŋʷ > kəwŋ (B&S *C.qˤoŋ)
容 *ɣàŋʷ > juawŋ (B&S *[ɢ]roŋ; 公 being phonetic in 容)
Chris Button said,
May 25, 2019 @ 7:42 am
Sorry, Baxter & Sagart 容 *[ɢ](r)oŋ with the -r- being in parentheses as something possible but not required.
Jonathan Smith said,
May 25, 2019 @ 10:41 am
No I don't think so; the lateral clusters should basically be Pulleyblank (1962). I don't think Baxter (1992) used clusters here at all. Don't know the history of your formulation. The uvulars… another of those could-be-right things that are useful above all for typifying all that is corny about "OC".
Chris Button said,
May 25, 2019 @ 2:38 pm
Baxter (1992) reconstructs 羌/姜 as *kh(l)jang/*k(l)jang (the -j- representing a type-B syllable), but yes, it seems pretty ad hoc and I can't find a discussion about it in the text. Incidentally, Beckwith (2009:375-6) in Empires of the Silk Road has seized on a reconstruction of 羌 as *klaŋ to claim that it "may also have an Indo-European etymology: the word klānk- in Tokharian means ‘to ride, go by wagon’, as in ‘to ride off to hunt from a chariot’, so Ch’iang could actually mean ‘charioteer’." Needless to say, I don't buy it at all.
Chris Button said,
May 25, 2019 @ 3:14 pm
As for uvulars, inspired by Pulleyblank's formulations, I go with *q-, *χ- (*qʰ-), *ʁ- (paralleling *-q and *-ʁ in coda position). However, the environments are totally different from Baxter & Sagart (it is associated with certain otherwise nebulous inconsistencies in labialization) and there is naturally no *ɢ- either. For example, I reconstruct 桂 "cinnamon, cassia" as *qájs which could regularly go back to an earlier *qʲáts and is clearly associated with Hebrew qetsia "cassia". Baxter & Sagart don't reconstruct 桂 but would presumably go with something like *[k]ʷˤe-s based on their reconstruction of 圭 *qáj as *[k]ʷˤe.
Jonathan Smith said,
May 25, 2019 @ 4:55 pm
The OC problem is that when some particular new idea accumulates a critical mass of explanatory power, a fact which given time is sure to be recognized by a critical mass of investigators, it becomes a triumph; short of that point, the same idea is useless for all practical purposes (like the above article) — "kong xiang 空想". And so "L series" from Pulleyblank (1962) has turned out to the the former, but much else from there and later the latter… so reflection on the sorts of paths which tend to lead in the first vs. the second direction seems crucial.
Chris Button said,
May 26, 2019 @ 8:30 am
@ Jonathan Smith
See p.25 here for some brief context:
https://www.bulgari-istoria-2010.com/booksBG/E_Pulleyblanck_JI_and_JIANG_2.pdf
Chris Button said,
May 26, 2019 @ 9:21 am
Regarding OC in general, the problem is indeed the reliability of the reconstruction. Regarding 羌 with a cluster, I remember noting how similar such a reconstruction made it look to the name "Chin" (applied on the Burmese side of the border to the Zo people) which goes back to Old Burmese *kʰlɐŋ². Beckwith can then suggest some similarly unrelated Tokharian link, and the list of proposals can go on and on…
Jonathan Smith said,
May 26, 2019 @ 4:24 pm
Thank you for that link. I hadn't checked but was afraid this idea was from the 1991 paper :/ the foundations there are unfortunately (1) stipulation of onset/coda symmetry, and much more seriously (2) the ganzhi = "phonograms" idea. These moved Pulleyblank in the direction of various kinds of rearrangements esp. compressions; in this specific case, he wants to split the difference between obstruent and glide "reflexes". So if we are serious about OC we must ask: would his proposed *ɣ- have emerged as a solution to MC k~j in XS independent of these arbitrary postulates? The same kind of tortured history lies behind B&S uvulars. This does not mean that these ideas are necessarily wrong, but they have dropped from a "poison tree" and thus are to be regarded with considerable suspicion.
Chris Button said,
May 26, 2019 @ 7:41 pm
The working and discussion in Pulleyblank's Ganzhi paper should be taken very seriously; many of the conclusions shouldn't. However, his discussion of *ɣ- is not radical, which is why I accept it. As Pulleyblank notes, Karlgren used *g- in a similar manner; the fact that Baxter & Sagart have recently tried to do something along those lines with *ɢ- is not coincidental either (in spite of its own problems).
Chris Button said,
May 26, 2019 @ 9:54 pm
In terms of phonological plausibility within Old Chinese for *ɣ- becoming k- (type A) and j- (type B) in Middle Chinese, it's also worth noting how *-ɣ as a coda can give -j in Middle Chinese and overlap with -k in phonetic series. For example 來 *rə́ɣ > ləj which is phonetic in 麥 *mrə̀k and 賚 *rə́ks.
While the loss of voicing in the fortition of ɣ > k is not a problem in coda position, it could theoretically be challenged in onset position. However, obstruents don't like to be voiced (the opposite of sonorants, albeit not quite as extreme) and this is all the more so in the case of velars than coronals or bilabials, etc. Remember *g- in Old Chinese may well have been pre-nasalized [ᵑg] for that reason (which incidentally seems to account far better for any external correspondences with nasals on simple phonetic grounds than any supposed "nasal prefix" in Old Chinese)
Chris Button said,
May 26, 2019 @ 9:56 pm
– nasalisation being a common mechanism to maintain voicing on obstruents
Jonathan Smith said,
May 27, 2019 @ 11:19 am
I think what you are calling the 1991 conclusions were actually the premises. But be that as it may — here and elsewhere you are arguing for particular values and changes, in general various of Pulleyblank's proposals, from the standpoint of plausibility, but plausibility is a thoroughly secondary consideration in OC and in fact more generally. The key consideration is economy, and quite consistently, it is newly uncovered economies which wind up finding external/comparative support (voiceless sonorants, unified r, Pulleyblank's L- series, etc.) Karlgren's unaspirated g/d/b, to take an example you mention, have been set aside in fact not at all because the sound changes they entailed were totally unreasonable. So it hardly matters whether (e.g.) your *ɣ- gives typologically more likely changes: the point is that K~j is an unsettled corner of OC waiting to be sucked into an economy. Pulleyblank (1962) wanted to make these words a special case of his L- (a.k.a. ð/θ) series ideas (because economical), but we seem not to find the hoped-for resonance with other material. So we have lots of other proposals, none of which has gained footing. The point in this thread is that none of these proposals deserve to be splattered around in generalist discussions, no matter what might in theory be argued regarding their relative plausibility or lack thereof. One could just cite "LHan"/"MC".
Chris Button said,
May 27, 2019 @ 4:15 pm
Not sure I'm quite following you now.
We have alternations of Middle Chinese-j and -k in coda position that go unexplained. We have alternations of k- and j- in onset position that go unexplained. We have external links with 巫 and 馬 suggesting a final velar of sorts which go unexplained in more recent reconstructions. Pulleyblank's /ɣ/ very economically accounts for them all.
Chris Button said,
May 27, 2019 @ 10:19 pm
A couple more thoughts…
It's also worth noting that since *g- > ɣ- in Type A (the frication seeming to be an attempt to handle the incompatibility of voicing with the stop), OC *ɣ- had to go somewhere in Type A to remain distinct. As an onset (and hence more like fricative [ɣ] than approximant [ɰ] in coda position, so the voicing is still going to be an issue) we see the fortition to k- (leaving the new ɣ- sound emerging in its place to then wrestle again with the voicing).
It's probably worth adding that this *-ɣ coda is reconstructed in more recent OC systems as a "coda-less" open syllable instead (regardless of any associations with *-k in phonetic series). The criticism that reconstructing glides -l, -j, -w, -ɣ [ɰ], -ʁ across the whole "yin" category of rhymes leaves no open syllables in Old Chinese, represents little more than a misunderstanding of the difference between surface phonetics and underlying phonology.
Chris Button said,
May 28, 2019 @ 8:49 am
Sorry – was writing late at night. A better sequencing looking at their relative developments would be *ɣ- fortifies to k- thereby leaving ɣ- as an available slot for *g- to lenite to ɣ-.
Jonathan Smith said,
May 28, 2019 @ 9:15 am
By all means go on advocating for your interpretations. The point is that — to beat a dead horse — your, my, or anyone's idiosyncratic views on an unsettled question or twenty aren't necessarily helpful to someone trying to bring two or three not-flagrantly-wrong remarks about Old Chinese into some more general discussion. I'm sure I've encouraged you before to "break off a piece" (say, your and Pulleyblank's vertical vowel system… or *ɣ, etc., for that matter) and publish something brief trying to show systematically how such a revision can solve real problems, as opposed to wrapping everything up in a rather less digestible magnum opus…
Chris Button said,
May 28, 2019 @ 12:35 pm
I think that would be good advice for an academic looking to advance their career. I have rather different aims for my dictionary. In the meantime, this a good outlet for thoughts, and I do very much appreciate the good discussion.
Jonathan Smith said,
May 28, 2019 @ 1:44 pm
As much as advancing a (future or hypothetical) career, I think the above incremental approach would be the best way to advance the field as a whole… so I suppose my suggestion is ultimately selflish :D Naturally your choices must be designed after your own circumstances and goals. Press on ~~
Chris Button said,
May 29, 2019 @ 9:47 pm
I just took a look at Pulleyblank's (1966) "Chinese and Indo-Europeans" article. Based on his OC formulations back then when *l- was being used for *r-, he suggests that 麟 as in 麒麟 may have had a cluster onset like *ɦl- or *γl-. In his "The Consonantal System of Old Chinese" (1962), he similarly speculates that the 樓 in 樓蘭 might have had a cluster onset like *ɦl-.
This reminded me of Prof. Mair's comment on the "Of Reindeer and Old Sinitic Reconstructions" thread (https://languagelog.ldc.upenn.edu/nll/?p=41164) regarding the mysterious velar component represented by the first syllable 麒 in 麒麟:
I wonder if 麟 like 樓 is perhaps another case of *ʰr- being associated with *kr- before ending up ultimately as *r-?
@ Jonathan Smith
Thanks – I hope you will find it worth the wait!
Chris Button said,
May 29, 2019 @ 10:41 pm
By *γl- I mean *ɣl-
Victor Mair said,
May 29, 2019 @ 10:44 pm
@Chris Button
"I wonder if 麟 like 樓 is perhaps another case of *ʰr- being associated with *kr- before ending up ultimately as *r-?"
Thanks. That was a key part of my "reindeer" hypothesis, which I waited many years before presenting to the world in this post.
Victor Mair said,
May 29, 2019 @ 10:46 pm
@Jonathan Smith
"I think the above incremental approach would be the best way to advance the field as a whole…"
I agree, and am grateful to you, Chris Button, and occasionally others for carrying on the extended conversations in which you often engage on Language Log. People will look back on them with gratitude and will gain much of value from them.
Chris Button said,
May 31, 2019 @ 6:23 am
It occurs to me we also have 鐵 *ɬə́c "iron" (/ʰl/ being [ɬ]) and its possible association with external velar clusters elsewhere, as discussed here: https://languagelog.ldc.upenn.edu/nll/?p=41832