Linguistic Miscellany Thread

[Richard W]

aw > oː, so <w> comes to be seen as mater lectionis.
aʔ > aː, so <ʔ> can also come to be seen as a mater lectionis.

A curious parallel: in Sumerian, there were glyphs for ʔan, ʔab, ʔin...then the glottal stop was lost, so these became VC glyphs, which had not previously existed.

So how did Sumerian manage without them? I presume the expedients are the argument for the glottal stop existing and then vanishing.

[jal]

All the independently invented writing systems started by representing syllables.

Korean didn't? Or isn't it independently invented?


JAL

[bradrn]

Not sure I understand this?

Loss with compensatory lengthening (and coloring?)

aw > oː, so <w> comes to be seen as mater lectionis.
aʔ > aː, so <ʔ> can also come to be seen as a mater lectionis.

In which language(s) do you mean?

[zompist]

All the independently invented writing systems started by representing syllables.

Korean didn't? Or isn't it independently invented?

It's very likely that the Koreans knew the 'Phags pa script, and possibly Indic scripts as well.

[Zju]

*puts tinfoil hat on*

I assume that refers to the PIE - Japanese relationship? Because the Kloekhorst proposal certainly isn't crackpottery.

The former.

[Richard W]

Loss with compensatory lengthening (and coloring?)

aw > oː, so <w> comes to be seen as mater lectionis.
aʔ > aː, so <ʔ> can also come to be seen as a mater lectionis.

In which language(s) do you mean?

I had thought in Aramaic, but it seems that where the monophthongisation of /aj/ and /aw/, where it occurs, is too late or in the wrong part of the word. They do monophthongise in the Canaanite dialects and many others, but they were slower to adopt matres lectionis, though I've seen it said that Hebrew was much faster than Phoenician.

What can be seen in the record is the writing of final and then, a little later, word-internal /iː/ and /uː/ using the letters <y> and <w>. Now, such may arise by reanalysing /iː/ and /uː/ as /ij/ and /uw/, which one might think was likeliest before vowels, but I'm not sure that the Aramaic of the time (9th Century BC) had any words that would count as beginning with vowels.

[Moose-tache]

I think Phags-pa and especially Hangeul should be considered full alphabets, so "invented only once" should be changed to "probably invented at least twice."

Sometimes my conlanging instincts move me to invent a deliberately terrible script, i.e. one that is so underspecified that it serves more as a mnemonic device than anything else. Very fun, very naturalistic. But every time I try this, I run into the problem that I myself need the detailed information in order to remember things like pronunciation and etymology, especially for a conlang I'm still working on.

Imagine seeing an entry in your own handwriting that says:
m-H-n: onomatopoeia, sound of a cat
...and trying to figure out what the hell you intended by this.

[zompist]

I think Phags-pa and especially Hangeul should be considered full alphabets, so "invented only once" should be changed to "probably invented at least twice."

Fair enough. Still, I think Peter Daniels overemphasizes the alphabet/abjad distinction. And an abugida is pretty much equivalent to an alphabet.

'Phags pa knew a few Indic-derived scripts. But I guess he deserves extra points for making the vowels more separated.

Sometimes my conlanging instincts move me to invent a deliberately terrible script, i.e. one that is so underspecified that it serves more as a mnemonic device than anything else. Very fun, very naturalistic. But every time I try this, I run into the problem that I myself need the detailed information in order to remember things like pronunciation and etymology, especially for a conlang I'm still working on.

Inspired by Linear B? You could use the same approach as modern scholars do-- accompany it by a phonemic reconstruction.

[WeepingElf]

I think Phags-pa and especially Hangeul should be considered full alphabets, so "invented only once" should be changed to "probably invented at least twice."

Fair enough. Still, I think Peter Daniels overemphasizes the alphabet/abjad distinction. And an abugida is pretty much equivalent to an alphabet.

One could say that an abugida is an abjad with mandatory vowel diacritics. And the step has been made at least twice - once in Ethiopia, once (or twice, since we also have Kharoshthi) in India.

'Phags pa knew a few Indic-derived scripts. But I guess he deserves extra points for making the vowels more separated.

You mean, King Sejong (or the scholars who invented Hangul for him) knew a few Indic-derived scripts, I guess. 'Phags pa is not a person but a script

[zompist]

'Phags pa knew a few Indic-derived scripts. But I guess he deserves extra points for making the vowels more separated.

You mean, King Sejong (or the scholars who invented Hangul for him) knew a few Indic-derived scripts, I guess. 'Phags pa is not a person but a script

No, I mean 'Phags pa, the person, the National Preceptor, the Buddhist monk who invented the 'Phags pa script.

[Richard W]

I think Phags-pa and especially Hangeul should be considered full alphabets, so "invented only once" should be changed to "probably invented at least twice."

Fair enough. Still, I think Peter Daniels overemphasizes the alphabet/abjad distinction. And an abugida is pretty much equivalent to an alphabet.

'Phags pa knew a few Indic-derived scripts. But I guess he deserves extra points for making the vowels more separated.

The Phags-pa script is predated by the traditional Mongolian script, an alphabet. Its precursor, the 'Old Uyghur' script, was usually used as an alphabet. This is a definite case of matres lectionis becoming the vowels of an alphabet simply by making their usage mandatory. The Phags-pa script could optionally be used as an alphabet (as opposed to an abugida) just to copy the Mongolian method.

These are not the only dual mode scripts. Pali in the Thai script can be written either as a traditional abugida (the more official method) or as an alphabet (the commoner use, at least for prayer books). If one doesn't know the system one is looking at, one can misread some Pali words.

[WeepingElf]

'Phags pa knew a few Indic-derived scripts. But I guess he deserves extra points for making the vowels more separated.

You mean, King Sejong (or the scholars who invented Hangul for him) knew a few Indic-derived scripts, I guess. 'Phags pa is not a person but a script

No, I mean 'Phags pa, the person, the National Preceptor, the Buddhist monk who invented the 'Phags pa script.

Ah, I didn't know that the script was named for its inventor. Thank you for clarifying this.

[Moose-tache]

Phags-pa is based on Tibetan. I never noticed that the Mongolic script predates it. TIL!

Zompist brings up a fun is-a-hotdog-a-sandwich argument: when does an abugida become an alphabet? Phags-pa and Hangeul give vowels the same visual weight as consonants. In Hangeul they are always mandatory, and in Phags-pa they can stand alone without a consonant. But some abugidas really blur the line as well, and a conlanger could blur it even further.

EDIT: King Sejong's biography says that he spent his youth doing nothing but reading, and he was allowed to get away with it because he wasn't the heir apparent yet. Assuming that wasn't fudged after the fact, he could well have taken the lead in inventing Hangeul. We'll never know...

[bradrn]

Zompist brings up a fun is-a-hotdog-a-sandwich argument: when does an abugida become an alphabet? Phags-pa and Hangeul give vowels the same visual weight as consonants. In Hangeul they are always mandatory, and in Phags-pa they can stand alone without a consonant. But some abugidas really blur the line as well, and a conlanger could blur it even further.

I’ve had that discussion before, I think. As I recall Richard W says that vowels have become mandatory in Lao, such that it’s functionally an alphabet. There are other neat examples too. But I’d have to find the original conversation for all the details.

[Travis B.]

Apparently reason why people categorize 'Phags-pa as an abugida is that it has an inherent vowel and initial vowels have special forms, unlike Hangeul.

[Moose-tache]

How do we feel about the idea that /f/ is a post-neolithic innovation?

https://www.science.org/doi/10.1126/science.aav3218

[Raphael]

How do we feel about the idea that /f/ is a post-neolithic innovation?

https://www.science.org/doi/10.1126/science.aav3218

Huh. Sounds weird. Not at all sure I'm buying it. But I guess it would be kinda cool, in a weird way, if it would be true.

[zompist]

How do we feel about the idea that /f/ is a post-neolithic innovation?

https://www.science.org/doi/10.1126/science.aav3218

Whenever I see someone doing statistics on languages, I get antsy. Languages are not independent variables. Language families, maybe.

The maps are certainly interesting, but then there's more inconvenient facts: a) present-day hunter-gatherers are not necessarily representative of paleolithic ones, and b) there's such a thing as areal features.

[jal]

It's a bogus article anyway. No serious scientist would write a conclusion that includes "reveal":

Our findings reveal that the transition from prehistoric foragers to contemporary societies has had an impact on the human speech apparatus, and therefore on our species’ main mode of communication and social differentiation: spoken language.

If they'd have written "suggest" or "supports the possibility", ok. But writing it like it's a forgone conclusion? :facepalm:


JAL

[malloc]

According to the relevant Wikipedia articles, ancient China and India both had words for massive numbers, such as 載 zǎi for 10⁴⁴ and महौघ mahaugha for 10⁶². Given the absence of anything like trillion-dollar corporations or high energy physics, what exactly were they using such large numbers for? Furthermore, how did they derive these numbers etymologically?