Linguistic Miscellany Thread

[Ephraim]

In 1709, Jonathan Swift published A Tritical Essay Upon the Faculties of the Mind where tritical comes from trite and critical.
https://www.jstor.org/stable/2914182

[Zju]

I guess grammaticalisation technically counts? Otherwise, all these examples give me some confidence that my hypothesis that portmanteaus are artifacts of written rather than spoken language holds some water.

It's only Chinese that could challenge that notion, but then again it has been written for millenia, and its words used to be all monosyllabic anyway. So are they really portmanteaus or just reshufflings of morphemes?

In 1709, Jonathan Swift published A Tritical Essay Upon the Faculties of the Mind where tritical comes from trite and critical.
https://www.jstor.org/stable/2914182

Now we're cooking

[bradrn]

its words used to be all monosyllabic anyway

I’d be very very careful with that assumption…

(Not that I know much about the history of Chinese myself, mind you. I just know that it used to have a lot more morphology than it does now.)

[Travis B.]

and its words used to be all monosyllabic anyway.

I gather you haven't read Baxter-Sagart.

[Zju]

I’d be very very careful with that assumption…

I've been left with the impression that indeed it had such a period, after the drop-out of presyllables and before mass derivation of bisyllabic lemmas.

I gather you haven't read Baxter-Sagart.

Indeed, not everybody has read Baxter-Sagart.

[Linguoboy]

I’d be very very careful with that assumption…

I've been left with the impression that indeed it had such a period, after the drop-out of presyllables and before mass derivation of bisyllabic lemmas.

The are unanalysable lexemes like 蝴蝶 "butterfly" and 忐忑 "fidgety" which have been reconstructed as bisyllabic for the entirety of their existence, extending back at least as far as Old Chinese.

[Imralu]

I gather you haven't read Baxter-Sagart.

Indeed, not everybody has read Baxter-Sagart.

*sad Baxter-Sagart noises*

[Richard W]

How complex does clusivity get with first person pronouns? I am trying to understand a verbally tetrasyllabic and orthographically monosyllabic Northern Thai phrase meaning ᨲᩪᩦ᩵ᩮᩣ᩠ᨿ /tuːtiːtawta:j/ 'We beat the tortoise to death', where the first syllable, translated as 'we', seems to mean I + you + 3rd person(s).

[bradrn]

verbally tetrasyllabic and orthographically monosyllabic

Wait, how does that work?

[WeepingElf]

I have a phonetics question. I feel as if rounded vowels (especially front rounded ones) sound much like a clarinet, or a rectangle wave from a synthesizer. These are characterized by a spectrum with (almost) only odd harmonics. My question is, do rounded lips change the resonance behaviour of the vocal tract in a way that even harmonics are weakened?

[Creyeditor]

I think this a question about timbre (or Klangfarbe in German). This is about which parts of the acoustic spectrum are louder compared to other parts of the spectrum. This distinguishes vowel qualities as well as different musical instruments.
As for fromt rounded vowels, if I remember correctly, lip rounding affects F2 (rounded vowels sound more back), F3 (rounded vowels sound more 'bunched') as well as higher formants. Since the higher formants are not very relevant for phonological contrast, I don't remember the details.
Unfortunately, I know next to nothing about the timbre of musical instruments.
To me, front rounded vowels always sound like whistling, if that helps at all.

[TomHChappell]

Most of the formants are about odd harmonics anyway. (Like an oboe, or other double-reed woodwind.)
F0 is an exception; it’s just the frequency of the voicing.
F1 is mostly the fundamental frequency that goes with a wavelength four times as long as the path from the glottis to the lips.
F2 is mostly the frequency of a wave that’s a 3rd as long as the F1 wave. Its wavelength is 4/3 as long as the vocal tract. So its frequency is 3 times as fast as the F1.
F3 is mostly the frequency of a wave a 5th as long as the F1 wave (to wit 4/5 the length of the vocal tract); its frequency is 5 times that of F1.

Vowel quality is mostly about F1, F2, and F3. F1 and F2 are related to closeness and openness; F2 and F3 are related to frontness and backness; F3 and F4 are related to roundedness. Semivowels are also determined by these first 4 formants.

I think F4 gets changed by rounding the lips, so that all (or most) of the wavelengths are integral multiples (including even multiples, as well as odd multiples) of twicd the distance from the glottis to the lips. Or that’s my guess.
These go with the even multiples of the frequency of F1.

F4 is mostly the frequency 7 times F1. Its wavelength is usually 4/7 the length of the vocal tract.
But rounding the lips tends to emphasize the even harmonics over the odd ones. If I’m right!

Rhotics, or at least those that are rhotic because they are retroflex or uvular, are determined mostly by F1, F2, F3, and F4.
Retroflexion shortens the vocal tract, more than rounding does. The ratio by which retroflexion shortens the vocal tract, is not an integer, and may not even be a rational number.
Uvularization introduces a constriction further back in the mouth. Once that’s done, the sound passes through two filters; first the one from the glottis to the uvula; then second the one from the uvula to the lips.

Rhotics, or at least those that aren’t trills nor taps nor flaps, involve raising the F1 frequency and lowering the F2 frequency to be near each other; lowering the F2 frequency and raising the F3 frequency to be far from each other; and raising the F3 frequency and lowering the F4 frequency to be near each other. So F4 is more about certain liquids (namely the rhotic ones), rather than vowels or semivowels.

F5’s main frequency defaults to 9 times as fast as F1’s. I can’t exactly remember which consonants are mostly regulated by F5. I think it might be certain fricatives and affricates. I’m pretty sure it’s involved in hushing sounds, like shibolents. I don’t know about fibilants and thibolents.

F6’s main frequency defaults to 11 times as fast as F1’s. It is involved in hissing sounds, like sibilants.

If there’re formants F7 or F8 or F9 etc., they don’t seem to be important in phonology and phonetics.

….

There are other “anti-formants” that suppress (reduce or eliminate) certain frequencies.

The most prominent one in my mind is FN, the nasal antiformant. Its fundamental frequency goes with a wavelength 4 times the distance from the glottis to the nose. So it would suppress frequencies associated with odd multiples of that wavelength. But the nasal tract also involves the sinuses. So frequencies that are integral multiples of the one with a wavelength twice as long as the distance from the glottis to the front-most part of the sinuses, will be suppressed differently (less? or more?) than those suppressed by the tract from the glottis to the nostril.

There is also the lateral antiformant FL. If part of the sound-stream is allowed to “leak” over the sides of the tongue, frequencies resonant with some length either starting or ending where that “leak” is located, will be suppressed.

FL is involved in producing lateral and lateralised sounds. So F4 is involved in rhotic liquids, but FL is involved in lateral (“lambdic”) liquids.

FN and FL are not associated with any necessarily rational multiples of F1.

….

I am not certain I’ve mentioned all the relevant antiformants; I think probably so, but maybe not; there may be one more.

….

Anyway, F1 and F2 and F3 and F4 (and I think F5 and F6) are mostly composed of odd harmonics.

But I think you are correct in thinking rounding does tilt F4 towards the even harmonics.

….

I hope that helps! Some of it I’m sure of, most of it is an educated guess, and a little of it is a SWAG.

[Richard W]

verbally tetrasyllabic and orthographically monosyllabic

Wait, how does that work?

Phonotactics and statistics. Phonetically, every syllable starts with a consonant. If successive syllables share an initial consonant, one can just omit the second occurrence of the syllables and write both vowels attached to the consonant. Similarly, if two successive syllables have the same vowel, one can just stack the consonants and drop the first vowel. So instead of writing /tuːtiːtawta:j/ ᨲᩪ ᨲᩦ ᨲᩮ᩵ᩣ ᨲᩣ᩠ᨿ (with spaces added for exposition), one just omits the common consonant and put the vowels in their usual places, with deduplication, yielding ᨲᩪᩦ᩵ᩮᩣ᩠ᨿ. Sometimes this verges on being a write-only system; and some kind souls sometimes write how many syllables there are. In this case there is already a common contraction ᨲᩦᩣ᩠ᨿ /tiːta:j/ 'to beat to death'.

[jal]

I hope that helps!

Very interesting, thanks! For me, having severe hearing loss above +/- 1500Hz, rounded and unrounded vowels sound alike (at the same position, so /i/ = /y/ etc.). So the difference must indeed be F3, since I can't hear those frequencies.


JAL

[WeepingElf]

Meanwhile, I have realized that I was misguided about lip rounding and odd vs. even harmonics. Clarinets and rectangular waves do not sound like [ø] as like [m], and clarinets are known to behave like a stopped pipe, i.e. a pipe with a lid at the end that reflects the sound waves back into the pipe, causing even harmonics to cancel themselves out. And guess what you have with [m]: a stopped pipe. When you open and close your mouth while humming, it sounds quite much like switching back and forth between sawtooth and rectangular waves on a synthesizer. Nevertheless, thank you all for the answers you gave.

[Travis B.]

I hope that helps!

Very interesting, thanks! For me, having severe hearing loss above +/- 1500Hz, rounded and unrounded vowels sound alike (at the same position, so /i/ = /y/ etc.). So the difference must indeed be F3, since I can't hear those frequencies.

Conversely, for me I have a hard time telling [u] and [y] apart. This is probably because in my native language the two are allophones, and because the [y] in it is more akin to, say, the standard Swedish /ʉː/ than the standard Swedish /yː/ in that it is simply a fronted [u] without any special lip protrusion.

[Raphael]

Conversely, for me I have a hard time telling [u] and [y] apart. This is probably because in my native language the two are allophones,

Hm, isn't it more like the latter doesn't exist in you language?

[Travis B.]

Conversely, for me I have a hard time telling [u] and [y] apart. This is probably because in my native language the two are allophones,

Hm, isn't it more like the latter doesn't exist in you language?

[y(ː)] and [ʏ(ː)] are allophones of /u/ and /ʊ/ after alveolar, palatoalveolar, and palatal consonants in the English I am familiar with. Properly, there are no /y/ or /ʏ/ phonemes in the English here, which is reflected by how many English-speakers replace, say, StG /yː/ and /ʏ/ with /ju/ in names (or in more old-fashioned pronunciations, /i/ or /ɪ/, but this actually reflects dialectal German pronunciations of names brought over from Europe). What I personally found hard about learning German pronunciation of high vowels is being able to deliberately pronounce back vowels after alveolar, palatoalveolar, and palatal consonants, and front vowels after labial and dorsal consonants, because this allophony is so automatic and ingrained for me.

Note that many speakers, such as my daughter, have generally centralized pronunciations of /u/ and /ʊ/ when not after alveolars, palatoalveolars, or palatals as [ʉ] and [ʉ̞] respectively, and my own idiolect is actually rather conservative in this regard in that it preserves historical [u] and [ʊ] after labial and dorsal consonants, when postvocalic, and when initial. I have heard anecdotally that many German-speakers identify English /u/ with StG /yː/ rather than StG /uː/ because of this.

[Travis B.]

I should note that /ju/ is pronounced [jy(ː)], so it is not as far from the StG as one might think.

[Zju]

Why did Latin cavea 'cage, etc' become French cage 'i.d.' and not *cae or something to that effect? Was it w → gʷ? I thought that only occurred in Old Latin.