Linguistic Miscellany Thread

[Travis B.]

I have seen people comment that phonemic awareness is only a thing because of literacy, and that people by default have no such thing. The problem with this position as I see it is then, if it were fictional, how do you explain things such as sandhi, free allophony, and like? How do you explain things like people being able to rhyme and alliterate words even when they vary due to environment, including people who do not use alphabetic or abugida scripts (e.g. speakers of Old Chinese (and yes, I know about phonetics, but they would not have come up with these without having phonemic awareness in the first place))? How do you explain things such as illiterate people coming up with new scripts based solely on the knowledge that writing is possible, such as Sequoyah when he came up with the Cherokee script; if phonemic awareness required being literate in the first place, this should be impossible. As for sandhi and allophony, does one chalk it all up to allomorphy? But allomorphy is normally irregular, not regular and applying evenly to all applicable morphemes (aside from special ones such as onomatopoeia and interjections).

[Travis B.]

This links into a somewhat different matter, which has to do with words like Saturday and twenty where I have no awareness of a synchronic connection between the normal pronunciations [ˈsɛːʁˤde(ː)] and [ˈtʰwʌ̃ːj] and the careful pronunciations [ˈsɛɾʁ̩ˤːˌde(ː)] and [ˈtʰwʌ̃ɾ̃i(ː)], i.e. I am not consciously deriving the former from the latter, rather they are separate parallel pronunciations varying in register. (I subjectively feel like I am adding consonants when speaking carefully, not deleting consonants when speaking normally.) This would argue against phonemic awareness in this case, if the two were still linked at some level, or if the two were truly separate, it would point at some deep restructuring of the underlying phonemic system (e.g. reintroducing phonemic vowel length and introducing phonemic vowel nasality). However, the problem with the latter is that people don't construct words that exploit any possible phonemic vowel length and vowel nasality; rather, words are constructed as if both of these were still allophonic. Another problem is that vowel length in many words is still clearly allophonic in final syllables, based on the fortis/lenis values of initial obstruents of following words. But what does this say about these cases, because that implies that this is allophony. Can one have allophony where one is fundamentally unconscious of its operation even if one is literate (if phonemic awareness is predicated upon literacy)? Or is it that one can have a phonemic system that is not fully consciously accessible, such that new words cannot exploit the fully set of possibilities in the phonemic system but rather have to follow a given set of rules, almost as if new words have a different phonemic system from existing words?

[Travis B.]

One possibility is that there is a connection, but the phonemic rules behind that connection are normally unconscious but not completely inaccessible, in that when speaking carefully one can consciously override some of these rules to produce careful pronunciations based on the same underlying forms. This is why the careful pronunciations of Saturday and twenty feel like I am adding consonants rather than the everyday pronunciations feeling like I am deleting them, because I am specifically overriding normally unconscious rules to produce them.

[zompist]

I have seen people comment that phonemic awareness is only a thing because of literacy, and that people by default have no such thing.

I'm not one of those people, but I don't think the case for phonemes is as slam-dunk as you assume. I could put it the other way: if there are phonemes, why is there half-rhyme and assonance? Why are there phonological controversies? Why do (new) orthographies not match up perfectly with phonemes?

The problem with this position as I see it is then, if it were fictional, how do you explain things such as sandhi, free allophony, and like?

You can certainly use phonemes to explain sandhi. E.g. "I'm gonna" can be represented /ajm gʌnɘ/ wtih a sandhi rule /mg/ > [ŋg]. Cool. But we are postulating an unseen entity, this /m/ that is not actually expressed, and we should be careful about such things. If nothing else, it's not always so clear when we're dealing with an active phonological rule and when we're dealing with a historical relic.

What would a purely phonetic approach look like? Perhaps like this: "I'm" has at least two forms, [ajm] and [ajŋ], and the latter must be used before a following [g]. Surely that rule is no more impossible than saying that the indefinite article has two forms [ɘ] and [ɘn], and the latter must be used before a vowel. Maybe a full explanation of English sandhi rules would be cumbersome in this formulation-- but maybe not, we'd have to see.

How do you explain things like people being able to rhyme and alliterate words even when they vary due to environment, including people who do not use alphabetic or abugida scripts (e.g. speakers of Old Chinese

Chinese is actually a rather hard case for phonemes! The native analysis of syllables, inherent in the rhyme tables as well as poetry and the writing system itself, is in terms of onset and rhyme plus tone. The Mandarin onsets contain a classic phonological puzzle (the palatals are allophones of some other series, but there are 3 to choose from). The rhymes are a mess, not easily divided into phonemes... they can be manhandled into a Western framework, but it's evident that whatever phonemes you come up with are highly interdependent. It really is easier to explain and teach by listing the rhymes.

Can you make a good case that, say, initial and medial t in English are 'the same thing'? I think there's a lot of assumptions needed to do so, and not all of them may hold up to a lot of scrutiny. It's commonplace, after all, for languages to allow a different set of consonants in initial, medial, and final position.

[bradrn]

Chinese is actually a rather hard case for phonemes! The native analysis of syllables, inherent in the rhyme tables as well as poetry and the writing system itself, is in terms of onset and rhyme plus tone. The Mandarin onsets contain a classic phonological puzzle (the palatals are allophones of some other series, but there are 3 to choose from). The rhymes are a mess, not easily divided into phonemes... they can be manhandled into a Western framework, but it's evident that whatever phonemes you come up with are highly interdependent. It really is easier to explain and teach by listing the rhymes.

Actually, now that you’ve mentioned it, do you know of any good list of phonetic transcriptions of Standard Mandarin rhymes? (Preferably with Pinyin transcriptions, though I don’t mind particularly.) I was trying to find one just the other day, but I couldn’t find any good list, and the ones that I found all disagreed with each other — I suspect they were trying to transcribe the rhymes phonemically rather than phonetically.

[Travis B.]

I have seen people comment that phonemic awareness is only a thing because of literacy, and that people by default have no such thing.

I'm not one of those people, but I don't think the case for phonemes is as slam-dunk as you assume. I could put it the other way: if there are phonemes, why is there half-rhyme and assonance?

This could be explained if people are also aware of phones on a featural level. Half-rhymes and assonance clearly indicate that people perceive the sounds they produce as features rather than as just unanalyzable wholes.

Why are there phonological controversies? Why do (new) orthographies not match up perfectly with phonemes?

One thing that might explain it is if people have different underlying forms that other people think they should have, i.e. how they understand their own speech as operating may differ from how some linguist trying to analyze their speech would. They might not internally analyze away as much as the linguist may.

The problem with this position as I see it is then, if it were fictional, how do you explain things such as sandhi, free allophony, and like?

You can certainly use phonemes to explain sandhi. E.g. "I'm gonna" can be represented /ajm gʌnɘ/ wtih a sandhi rule /mg/ > [ŋg]. Cool. But we are postulating an unseen entity, this /m/ that is not actually expressed, and we should be careful about such things. If nothing else, it's not always so clear when we're dealing with an active phonological rule and when we're dealing with a historical relic.

What would a purely phonetic approach look like? Perhaps like this: "I'm" has at least two forms, [ajm] and [ajŋ], and the latter must be used before a following [g]. Surely that rule is no more impossible than saying that the indefinite article has two forms [ɘ] and [ɘn], and the latter must be used before a vowel. Maybe a full explanation of English sandhi rules would be cumbersome in this formulation-- but maybe not, we'd have to see.

The big problem to me is things such as final [n] in standalone words; this for me always assimilates in place to a following consonant, without fail. If this were allomorphy, it would require some explaining, because allomorphy is normally invoked to explain irregular things, not entirely regular things; indeed, if this were allomorphy, there should be examples which violate this.

How do you explain things like people being able to rhyme and alliterate words even when they vary due to environment, including people who do not use alphabetic or abugida scripts (e.g. speakers of Old Chinese

Chinese is actually a rather hard case for phonemes! The native analysis of syllables, inherent in the rhyme tables as well as poetry and the writing system itself, is in terms of onset and rhyme plus tone. The Mandarin onsets contain a classic phonological puzzle (the palatals are allophones of some other series, but there are 3 to choose from). The rhymes are a mess, not easily divided into phonemes... they can be manhandled into a Western framework, but it's evident that whatever phonemes you come up with are highly interdependent. It really is easier to explain and teach by listing the rhymes.

It is not an argument against underlying forms, though, just an argument against independent segmental phonemes. One could very well treat rhymes like we treat diphthongs and triphthongs and affricates in more conventional analyses of non-Sinitic languages.

Can you make a good case that, say, initial and medial t in English are 'the same thing'? I think there's a lot of assumptions needed to do so, and not all of them may hold up to a lot of scrutiny. It's commonplace, after all, for languages to allow a different set of consonants in initial, medial, and final position.

Medial and final t have to be the same at least for me, because final t is treated by medial t when it is both preceded and followed by vowels. This is regular, without exception; however, initial t and medial t are different, because there is a very small set of words, namely to, today, and tomorrow, which optionally can start with medial t instead of the usual initial t; this can only be explained by allomorphy, but implies that initial t and medial t are not the same in the underlying forms.

[Rounin Ryuuji]

I have seen people comment that phonemic awareness is only a thing because of literacy, and that people by default have no such thing.

I'm not one of those people, but I don't think the case for phonemes is as slam-dunk as you assume. I could put it the other way: if there are phonemes, why is there half-rhyme and assonance?

This could be explained if people are also aware of phones on a featural level. Half-rhymes and assonance clearly indicate that people perceive the sounds they produce as features rather than as just unanalyzable wholes.

Since our understanding is that poetry (which is frequently based on some sort of phonetic patterning) developed out of earlier oral traditions, I would guess that people are perfectly capable of distinguishing that words are composed of sounds; I don't think humans would've invented phonetic orthographies, much less featural ones like Hangul, otherwise. Even chiefly logographic systems do, in my understanding, tend to have some form of phonetic component to them.

Some languages (French, Italian, Japanese), it might be worth noting, are better at rhyming than others (English); note that both French and Italian also have traditionally rhyming poetry, while English has (or it might be better to say that Old English had) traditionally alliterative poetry.

It might, contrarily, be worth noting that some speakers won't be cognizant of similar sounds not being the same (most English-speakers probably don't realise English has both /θ ð/, both because /ð/ is a rare phoneme, and probably also because of the orthography; Chaucer also frequently wrote "ee" for both [ɛː] and [eː], and "oo" for [ɔː] and [oː], but later orthographers seem to have noticed the difference, unless I misunderstand, so boat and boot ceased to be homographs), so I would guess there's some sort of continuum at work here.

Can you make a good case that, say, initial and medial t in English are 'the same thing'? I think there's a lot of assumptions needed to do so, and not all of them may hold up to a lot of scrutiny. It's commonplace, after all, for languages to allow a different set of consonants in initial, medial, and final position.

Medial and final t have to be the same at least for me, because final t is treated by medial t when it is both preceded and followed by vowels. This is regular, without exception; however, initial t and medial t are different, because there is a very small set of words, namely to, today, and tomorrow, which optionally can start with medial t instead of the usual initial t; this can only be explained by allomorphy, but implies that initial t and medial t are not the same in the underlying forms.

I think I would say that, for me, /t/ has several allophones, morpheme-initial [tʰ], morpheme internally usually [d] ("kitty" and "kiddie" are homophones), but before nasals often [ʔ] ("mountain" [maunʔn], eye dialect "mown'n"; though careful speech can still also have [mauntən]); I don't perceive /t/ and /d/ as distinct intervocalically, but I do perceive there being a /t/ (which is either unreleased or a glottal stop) at the end of "write", probably by analogy with "writing", "written", which show the same allophonic patterning ("writing" and "riding" are homophones, but "written" and "ridden" are not); also unaspirated [t] in the sequence /st/. Of course, these perceptions are probably coloured more than slightly by having had them orthographically identical for three decades or so.

None of these sounds are phonemic elsewhere except [d] (for me, that is), so I would say they're still the same underlying phoneme, but likely won't be for long.

[Nortaneous]

Re: phonemes, Sikaritai has been alleged to have intramorphemic contrastive syllabification.

Some languages (French, Italian, Japanese), it might be worth noting, are better at rhyming than others (English); note that both French and Italian also have traditionally rhyming poetry, while English has (or it might be better to say that Old English had) traditionally alliterative poetry.

Tocharian poetry was a lot like Old English poetry, and Tocharian B was practically Italian at the right edge of the word - words generally end in a vowel or -n, and IIRC it's even been proposed that /n r/ were the only permissible word-final consonants and everything else had a schwa. Couldn't alliteration just be the older IE form?

When rhyming poetry emerged in Europe, was it popularized by languages with final stress? Modern French has final stress, but had the Occitan of the troubadours or whatever undergone apocope yet? Presumably everyone else imitated the troubadours. TB had occasionally mobile but generally penultimate stress - I've also seen second-syllable stress plus a rule preventing final stress claimed, but either way, it didn't have much final stress. Is there any correlation between poetic "type" in this sense and stress pattern?

[bradrn]

Re: phonemes, Sikaritai has been alleged to have intramorphemic contrastive syllabification.

Which page? I can’t find it. (Unless you mean the contrast between complex nuclei and hiatus… but that’s pretty normal, isn’t it?)

[Rounin Ryuuji]

Some languages (French, Italian, Japanese), it might be worth noting, are better at rhyming than others (English); note that both French and Italian also have traditionally rhyming poetry, while English has (or it might be better to say that Old English had) traditionally alliterative poetry.

Tocharian poetry was a lot like Old English poetry, and Tocharian B was practically Italian at the right edge of the word - words generally end in a vowel or -n, and IIRC it's even been proposed that /n r/ were the only permissible word-final consonants and everything else had a schwa. Couldn't alliteration just be the older IE form?

This is an interesting possibility, and perhaps.

When rhyming poetry emerged in Europe, was it popularized by languages with final stress?

I'm not very well-versed (pun not intended, but acknowledged) in the French of the era, but if memory serves, both Chaucer's English and the French of the period are thought to have usually had a pronounced final -e, though Chaucer often seems to ignore it under certain circumstances, as here:

(118) Ther was also a Nonne, a PRIORESSE,
(119) That of hir smylyng was ful symple and coy;
(120) Hire gretteste ooth was but by Seinte Loy;
(121) And she was cleped madame Eglentyne.
(122) Ful weel she soong the service dyvyne,

Terminally, it seems to be mostly ignored, though usual reconstructed pronunciation tends to involve reading it aloud; I suspect, however, Chaucer himself might've pronounced madame Eglantyne without either terminal -e, as it scans better; at least the one in madame was probably eclipsed by the following vowel, as was that in Nonne, owing to the following article; service is certainly trisyllabic, however. Middle English most certainly inherited rhyme from French, but it's been very persistent in English, probably because it isn't all that easy to do in a very creative way, as one might expect from a language with a huge phoneme inventory and fairly loose constraints on where the sounds can go.

The tendency to alliteration is also, if memory serves, preserved in the Chanson de Roland, which I believe was assonal rather than rhyming. I've encountered speculation that Languages with small phoneme inventories (like Classical Latin, and Japanese) tend to eschew rhyming because it happens incidentally often enough that it seems to be considered unremarkable (Japanese poetry also tends to be more grammatically vague as a function of the language's long inflections, it would appear, so more evidence for structure influencing development). My guess is, after all that rambling, a concurrence that the structure of languages during the period did influence the development of poetry, though I don't know Old Occitan at all. The tendency to a stressed final syllable, or a stressed penult with an extremely weak or optionally-deleted terminal vowel would seem to help, at least with Middle English, though I doubt Middle English would've developed rhyme so readily without continental influence.

(This is a later addendum: With some further research, the Wikipedia article on Rhyme seems to suggest it was partly introduced from Arabic, through Spain, which I had not heard before. It seems to have also developed independently in Ireland, and there seems to be some contention over which was the source (I might even guess it was both converging). I wish I could comment further on that, but I don't know either Old Irish or Classical Arabic at all.)

[Kuchigakatai]

Some languages (French, Italian, Japanese), it might be worth noting, are better at rhyming than others (English); note that both French and Italian also have traditionally rhyming poetry, while English has (or it might be better to say that Old English had) traditionally alliterative poetry.

Terminally, it seems to be mostly ignored, though usual reconstructed pronunciation tends to involve reading it aloud; I suspect, however, Chaucer himself might've pronounced madame Eglantyne without either terminal -e, as it scans better; at least the one in madame was probably eclipsed by the following vowel, as was that in Nonne, owing to the following article; service is certainly trisyllabic, however. Middle English most certainly inherited rhyme from French, but it's been very persistent in English, probably because it isn't all that easy to do in a very creative way, as one might expect from a language with a huge phoneme inventory and fairly loose constraints on where the sounds can go.

Rap and fairy tales are often up to the brim full of rhyme though, and used in interesting ways. Also some pre-modern verse, e.g. Alexander Pope's version of the Aeneid. If creative rhyming is hard in English, does that means rappers are better poets than the alleged poets? (Here is a fun article that elaborates on this. The author thinks rhyming is unpopular in English, especially now, because it is "liked by the wrong people"...)

(This is a later addendum: With some further research, the Wikipedia article on Rhyme seems to suggest it was partly introduced from Arabic, through Spain, which I had not heard before. It seems to have also developed independently in Ireland, and there seems to be some contention over which was the source (I might even guess it was both converging). I wish I could comment further on that, but I don't know either Old Irish or Classical Arabic at all.)

Yeah, the Wikipedia article on the less prestigious rajaz poetry says it developed rhyme pretty early on around the 6th century (and has a nice example too!!). I remember once reading an article on this that specified it was assonance, not consonance... It's disturbing to think Ireland and the Arabic world may have developed rhyme independently around the same time(?) and then did a two-front attack into Western Europe...

When rhyming poetry emerged in Europe, was it popularized by languages with final stress? Modern French has final stress, but had the Occitan of the troubadours or whatever undergone apocope yet? Presumably everyone else imitated the troubadours.

Both assonance and consonance predate the Old Occitan trobadors though... It's there pretty early in Irish Latin as Wikipedia says (e.g. the Altus Prosator poem from the 6th century, with some consonance), and out of Ireland, the WP article "Leonine verse" has an 8th-century English example with consonance... In the continent, in Romance, it's there in the Sequence of St. Eulalia (9th century, assonance, early Old French), and I imagine there are somewhat older Latin poems in the West with rhyme...

I find it curious how the Sequence shows assonance by being rather loose with the consonants, yet it's not full-blown assonance either, as the poetry tries to have at least similar consonants. Check out the text, with maksimjɛn - pai̯jɛns (ɛ + nasal), and konseʎɛrs - tsjɛl - prei̯jɛr (=ɛ + liquid), but also mɛrtsiθ - vənir (outright assonance) (all words are stressed on the last syllable).

Actually, now that you’ve mentioned it, do you know of any good list of phonetic transcriptions of Standard Mandarin rhymes? (Preferably with Pinyin transcriptions, though I don’t mind particularly.) I was trying to find one just the other day, but I couldn’t find any good list, and the ones that I found all disagreed with each other — I suspect they were trying to transcribe the rhymes phonemically rather than phonetically.

No, but I can give you one right now—there's only so many rhymes. Do ask any questions you have about disagreements with other sources.

(sh)i [ɨ]~[ʐ̩]
(s)i [ɨ]~[z̩]
(x)i [ i]
ing [iŋ]
in [in]~[iə̯n]

e [ə]~[ɘʌ̯]
eng [əŋ]
en [ən]
un [wən] (e.g. gun [kwən], lun [lwən])
ie [jɛ]~[iɛ̯]
üe [ɥɛ]~[yɛ̯]

Note: these [-ə-] are often written "[ɤ]", intended to be mid, not mid-high.

ai ei ao ou [aɪ eɪ ɑɔ̯ ɔʊ]
uai ui iao iu [waɪ weɪ jɑɔ̯ jɔʊ] (e.g. gui [kweɪ], liu [ljɔʊ])

Yes, there's a good reason why it's -ao and not -au... In fact, the diphthong is so back and low I often seem to, perhaps falsely, hear -ao simply as [ɑ]~[ɒ], and -iao as [jɔ].

a [a]
ia [ja]~[ia̯]
ua [wa]
ang [ɑŋ]
iang [jɑŋ]~[iɑ̯ŋ]
uang [wɑŋ]
an [æn]
wan [wæn]
üan [ɥæn]~[yæ̯n]~[ɥɛn]
ian [jɛn]~[iɛ̯n]

uo (b)o [wɔ]~[uɔ̯] (e.g. po [pʰwɔ], duo [twɔ])
ong [ʊŋ]
iong [jʊŋ]~[iu̯ŋ]

ü (x)u [y]
ün (x)un [yn]~[yə̯n]

Erhua, i.e. the -r [ɹ] suffix generally centralizes or backs front vowels and makes various mergers among these rhymes. Notably, -n tends to simply be dropped, jin+r [tɕiɹ]~[tɕiəɹ], wan+r [wɑɹ], sounding as homophones of ji+r and wa+r, and -ŋ leaves nasalization behind, e.g. xiong+r [ɕjʊ̃ɹ].

[Rounin Ryuuji]

Rap and fairy tales are often up to the brim full of rhyme though, and used in interesting ways. Also some pre-modern verse, e.g. Alexander Pope's version of the Aeneid. If creative rhyming is hard in English, does that means rappers are better poets than the alleged poets? (Here is a fun article that elaborates on this. The author thinks rhyming is unpopular in English, especially now, because it is "liked by the wrong people"...)

Even if I have trouble connecting with most forms of popular music, I've never been bothered by rap and pop and whatever else using some of the same poetic techniques I do (I do sometimes roll my eyes at the rhymes they come up with, then forget them); it feels like a natural way of crafting lyrical expression, no matter how mundane or maudlin, or nostalgic, or grandiose, or comfortably ordinary, the subject matter. I should probably add as a disclaimer that I have an intense dislike of the idea of "free verse". If you want to write prose in a poetic style, that's very nice, but don't call it a poem — that's just being silly, or pretentious if you say others just "don't understand" or whatever. Of course, I'm perhaps old-fashioned, in that I like the idea of good, well-constructed literature (both poetry and prose) that can be consumed by all of society. That's what Shakespeare did, isn't it?

If I were to call out the academy for unfairly dismissing one thing, it would probably be fantasy; no academic wants to admit this, but considering actual cultural relevance, the most influential Twentieth Century author was arguably J.R.R. Tolkien (on this same thread, James Joyce is a near cultural nonentity in most of the English-speaking world), and changes in society have created new modes of fiction and mythopoesis that are both extremely interesting and deeply satisfying in ways what I've heard called "literary fiction" aren't. I do have a bias, though — I write fantasy myself. This is not exactly new, however — Chaucer seemed not to like Arthurian stories (of the Canterbury Tales, only The Wife of Bath's Tale is Arthurian); Gower also de-Arthurianises one tale for the Confessio Amantis. Of course, they've been dead a few centuries, and it might've been less snobbery (they still retold the stories) and more wanting to stand out from something so mind-bendingly popular.

[zompist]

The big problem to me is things such as final [n] in standalone words; this for me always assimilates in place to a following consonant, without fail. If this were allomorphy, it would require some explaining, because allomorphy is normally invoked to explain irregular things, not entirely regular things; indeed, if this were allomorphy, there should be examples which violate this.

Let's approach the original question this way: what are the underlying forms for (say) 'tin'? Not in theory, but in the brain? I can think of several alternatives.

* a minimal phonemic representation, the equivalent of /t ɪ n/.
* a phonetic representation, more like [t^h ɪ n]
* a detailed phonetic memory, like a sound recording of the syllable "tin"
* a motor memory, the nerve impulses needed to get the tongue, vocal cords, and lungs to utter "tin"
* Chinese-style, a representation of the onset [t] and the rhyme [ɪn]
* a combination of these

We don't know. But what I'd point out is that any of these could be used to explain why you have the assimilation. The phonemic theory has no particular advantage here.

For either the phonetic or phonemic versions, we simply add a rule that e.g. [n] before velar > [ŋ]. (Admittedly there's some abstraction in saying that 'tin cup' has an [n] that is realized as [ŋ], but this isn't a 'real' phonemic representation-- it's simply the standalone realization of the word, which you can easily elicit and study directly.)

For the detailed memory versions, the assimilation could be a physiological process that the brain isn't even aware of. If you think of the vocal organs as a machine, it just happens that trying to say "tin cup" at reasonable speeds, you need to anticipate the [k] and the n gets velarized.

[zompist]

If I were to call out the academy for unfairly dismissing one thing, it would probably be fantasy; no academic wants to admit this, but considering actual cultural relevance, the most influential Twentieth Century author was arguably J.R.R. Tolkien (on this same thread, James Joyce is a near cultural nonentity in most of the English-speaking world), and changes in society have created new modes of fiction and mythopoesis that are both extremely interesting and deeply satisfying in ways what I've heard called "literary fiction" aren't.

Oh come now.. Tolkien is most influential with nerds (like most of us on the board). If you're going by mass popularity, Tolkien is far eclipsed by (say) Agatha Christie. Fantasy is pretty popular, but mysteries are far more so. Mysteries and romance both outsell fantasy/sf.

[cedh]

Re: phonemes, Sikaritai has been alleged to have intramorphemic contrastive syllabification.

Which page? I can’t find it. (Unless you mean the contrast between complex nuclei and hiatus… but that’s pretty normal, isn’t it?)

There are at least two things in there:
- a marginal contrast between intervocalic [d] and [ɾ~l], which is analysed as /Vd.V/ :: /V.dV/ because all other instances of the liquid are quite straightforwardly positional allophones of /d/ (p. 99-101)
- a contrast between monosyllabic /au̯ ai̯ ɛi̯/ and disyllabic /a.u a.i ɛ.i/ (p. 111-113)

The supposed [dʒ] allophone of /i/ may also have to do with syllabification somehow, although it's not stated as such in the paper. But otherwise it wouldn't be obvious to me why /brɛiáwa/ (sic, should be /bdɛiáwa/) is pronounced [bɾɛ.já.wa], but /pɛ́íá/ is [pɛ́.dʒá]... (p. 108)

[Rounin Ryuuji]

If I were to call out the academy for unfairly dismissing one thing, it would probably be fantasy; no academic wants to admit this, but considering actual cultural relevance, the most influential Twentieth Century author was arguably J.R.R. Tolkien (on this same thread, James Joyce is a near cultural nonentity in most of the English-speaking world), and changes in society have created new modes of fiction and mythopoesis that are both extremely interesting and deeply satisfying in ways what I've heard called "literary fiction" aren't.

Oh come now.. Tolkien is most influential with nerds (like most of us on the board). If you're going by mass popularity, Tolkien is far eclipsed by (say) Agatha Christie. Fantasy is pretty popular, but mysteries are far more so. Mysteries and romance both outsell fantasy/sf.

Maybe I am biassed, but I think mysteries trace back to Poe, and of course Sherlock Holmes, which are both from the Nineteenth Century.

Also, fantasy is pretty mainstream these days.

[bradrn]

For anyone who needs a laugh, here’s a (sadly accurate) piece of linguistic humour I’d like to share with you: The Linguist Parallel Parking Challenge. Yes, it includes conlangers! (From the same site, I also enjoyed Linguistic Sub-Fields Explained for the Research Novice and Important Idioms in Contemporary Science; I should note the latter is more widely applicable outside linguistics.)

[KathTheDragon]

The Linguist Parallel Parking Challenge

Your link's broken.

[bradrn]

The Linguist Parallel Parking Challenge

Your link's broken.

So it is! Fixed now.

[Ares Land]

For anyone who needs a laugh, here’s a (sadly accurate) piece of linguistic humour I’d like to share with you: The Linguist Parallel Parking Challenge. Yes, it includes conlangers! (From the same site, I also enjoyed Linguistic Sub-Fields Explained for the Research Novice and Important Idioms in Contemporary Science; I should note the latter is more widely applicable outside linguistics.)

Thanks for the link; I liked those!