This is not about water, but it is too good to miss.
 |
| [high-res] |
Thanks to this highly unusual place-of-articulation shift the Central East Polynesian adjective *witi ‘quick, lively’ became Hawai‘ian wiki (mind you, it can still be pronounced ['witi] or ['viti], but the shifted pronunciation ['wiki] brings it phonetically closer to English quick and increases the odds of its being picked up by an English-speaker). Thus was born one of the most successful linguistic replicators of today. For centuries the virus was more or less confined to its insular homeland, but in the mid-1990s it infected the mind of an American computer programmer visiting the islands. Before long, all major language communities had their Wikis. There is of course a Hawai‘ian one as well!
I want to thank Lara Prescott for bringing this beautiful infographic presentation to my attention, and I hasten to share it wiki-wiki.
Though rare in the world's languages, /t/ > /k/ is extraordinarily common in Austronesian, having happened at least twenty separate times. Austronesian generally is riddled with bizarre sound shifts: see Blust's paper on non-linguistically motivated sound changes.
ReplyDeleteThanks for that interesting reference.
DeleteIt struck me that an (imperfect) parallel to the Uralic consonant gradation is described in the article under § 3.6 "Gemination of the onset of open final syllables in Berawan". It looks closest to the Sami one, where the "strong" grade undergoes gemination.
Anders
Aren't they beautiful? AN is one of my favourite families. With 1000+ members, it's hard to think of a possible shift that hasn't happened in at least one of them.
DeleteThe Berawan change indeed resembles Samic gradation. It also resembles the secondary gradation in Estonian and Livonian, where consonants were lengthened (possibly from geminate to extra-long) preceding a formerly long unstressed vowel: this process of compensatory lengthening may have been behind the original Samic-Finnic-Samoyedic system of gradation as well, due to a longer vowel allophone existing in open unstressed than closed unstressed syllables.
Delete(I get the impression motivations such as this can indeed be though for many of the changes Blust lists as unmotivated in his paper. *dr → Levei tʃ ~ Drehet kʰ seems likely to derive as *dr → *ɖʐ (rhotic → retroflexion) → *ʈʂ (devoicing due to general absence of phonemic voice) → Levei tʃ; → Drehet *ʈx (retraction) → *kx → kʰ. Sundanese *w, *b → *(n)c might have progressed thru a linguolabial stage such as *d̼ð̼. Etc.)
Yes, they might have. But as Blust also says, there isn't a shred of evidence for any such long chains except our prejudices about "naturalness", and in some cases there is evidence that they happened quite suddenly.
DeleteJihn, looking at charts of cognates, you get the sense that it's the overall shape of the word, and especially the vowels, that gets passed along and is the basis of comparison, with the consonants being rather secondary.
ReplyDeleteYes. The famous wisecrack about consonants counting for little and vowels for nothing at all emphatically does not apply to Austronesian, where the vowels are often unchanged right back to PAN while the consonants do the hula/hura around them.
DeleteI have to wonder if it's Indo-European - and maybe even moreso Germanic - bias that tells us vowels are unstable. As the best-understood family, it's no secret that IE isms get overextended to other families.
DeleteIf I had to wager, the relationship between phonetic and phonemic vowels (and vowel-like consonants) probably would have more to do with their stability in a given language (-> family) than anything peculiar to vowels themselves.
The vocalism of Ancient Greek was quite conservative, but it's true that many IE languages have played havoc with the inherited vowel system.
DeleteOn the other hand, Proto-Germanic vowels, even in suffixal and inflectional syllables have been preserved intact in Finnish: *kuningaz, *druxtinaz, *xrengaz → kuningas, ruhtinas, renkas. This is all the more remarkable because, apart from Old Runic, the thematic *-a- is not visible anywhere in the recorded Germanic languages, not even in Biblical Gothic.
Self-correction: rengas (renka-), of course.
DeleteA misconception that I see all too frequently is that phonemes are the same as phones. They're not. And as languages with small inventories almost universally show, they're actually a range of different phones for a given phoneme surrounded by a range of acceptable substitutes. And the reason is because phonemes aren't as much about having a sign to make a word but rather to have a contrast to improve understanding against other words. It's all about the contrasts, not specifically /t/s or /k/s.
ReplyDeleteAfter all, there aren't exactly keys in your mouth labeled a, b, h etc. that produce exact sounds every time. Sometimes they might be a little fronted, a little backed, a little palatalized, a little velarized, laminal, apical, dental, etc. As long as the sound clearly contrasts against the other phonemes, it is typically valid.
A good point, we tend to take some classificatory categories for granted, e.g. we treat /t/ and /k/ as inherently contrasting by virtue of being [Coronal] vs. [Dorsal]. This prejudice is visible even in the way phonological universals are formulated: "IF there is only one place of articulation for a given type of consonant, THEN it is most likely to be the alveo/dental region." In a language with a small consonant inventory even near-universal distinctions may not really matter. The Hawai‘ian /t ~ k/ obstruent is simply "lingual", contrasting with "labial" and "glottal" ones, but underspecified with regard to the place of articulation.
DeleteA trivial request: Could you interchange the position of the "Older Post" and "Newer Post" links? I am always surprised by having to move to the right to go backward and the left to go forward, as if I were reading something in Hebrew or Arabic. (Which I do not read.)
ReplyDeleteI wish I could, but it seems this aspect of the Blogger template can't be edited (or at least I haven't found out how to switch the positions).
Delete