Conlangery Podcast (Latest Ep: #94: Face and Politeness)

[finlay]

I don't find it that difficult, but then I'm quite good at listening and repeating sounds. I'm more just a bit surprised to hear it that way... I very often just pronounce the ん as a nasalised vowel like in French, as that's how I often hear it.

[clawgrip]

I don't find it that difficult, but then I'm quite good at listening and repeating sounds. I'm more just a bit surprised to hear it that way... I very often just pronounce the ん as a nasalised vowel like in French, as that's how I often hear it.

It depends on the person, of course, but words like 単位 (たんい) tan'i /tãː(j)i/, 禁煙 (きんえん) kin'en /kĩːjeɴ/and as you said 新大久保 (しんおおくぼ) Shin-Ōkubo /ɕĩː(j)oːkubo/ are hard for many to get right. A little easier are ones followed by y- s- or z-, e.g. 金曜日 (きんようび) kin'yōbi /kĩːjoːbi/, (たんす) tansu /tãːsu/, or 銀座 (ぎんざ) Ginza /ɡĩːza/.

[finlay]

i hadn't realised they did it with Ginza - perhaps i just don't listen that carefully. They certainly find coda m/n very difficult to tell apart.

As for the actual podcast, I was having to try very hard not to start practising clicks on the train today. Thanks I guess.

I had another couple of observations, but I've forgotten them mostly. One was that vowel harmony doesn't imply vowel rounding; you can have vowel harmony by height instead of by rounding or front/back. William even mentioned ATR harmony, which is indeed difficult to explain or understand, but doesn't have to involve front rounded vowels.

Another was one that I've made elsewhere: the Japanese pronunciation of /g/ as [ŋ]. I haven't listened hard enough to find out exactly what situations it's allowed in (it may only be intervocalically), but it seems to take one unstable sound (g) and replace it with an even more unstable one (onset ŋ)...

[Ollock]

Super awesome to get Jim Henry on the show: Conlangery #49: The Noun Phrase

[Drydic]

Better late than never?

[Ollock]

Conlangery #50: The Technology of Literacy

[Ollock]

Hey, we reviewed Radius Solus's language in this one: Conlangery #51: Language History

[finlay]

An ok episode. I'd have liked to have heard more about how the neogrammarian hypothesis isn't necessarily true; aiui, sound changes take a while to "complete" and affect every word. They follow an s-shaped curve, affecting a small number at first and a lot all at once, but can abort at any time, so that you reasonably often get irregularities, either only a few affected at once or only a few which don't get affected.

[Pthagnar]

He's called Radius Solis. You know, like Latin. Julius Caesar, aqueducts, centurions and that.

[Astraios]

Don't be ridiculous, everything in Latin ends with -us.

[Bob Johnson]

quidquidus latinus dictus situs altus viditus

also: wow i'm more than half a year behind. currently on #19

[Drydic]

Don't be ridiculous, everything in Latin ends with -us.

don't be absurd

1/3 end in ae

[Jipí]

1/3 end in ae

*ii

[GrinningManiac]

Romanes eunt domus

[Drydic]

That's painful to look at.

[Ollock]

An ok episode. I'd have liked to have heard more about how the neogrammarian hypothesis isn't necessarily true; aiui, sound changes take a while to "complete" and affect every word. They follow an s-shaped curve, affecting a small number at first and a lot all at once, but can abort at any time, so that you reasonably often get irregularities, either only a few affected at once or only a few which don't get affected.

Hmm, to what extent does that really happen? Examples? I think the point of us saying you should assume the Neogrammarian hypothesis is that it makes it much easier to deal with the consequences of sound changes. There are a lot of small details in linguistic theory that most conlangers don't really need to be concerned with. Diachronic conlanging is enough work without also having to deal with irregular or aborted sound changes.

He's called Radius Solis. You know, like Latin. Julius Caesar, aqueducts, centurions and that.

I know. I always get the name wrong. I do not speak Latin and will likely never learn it.

[Drydic]

An ok episode. I'd have liked to have heard more about how the neogrammarian hypothesis isn't necessarily true; aiui, sound changes take a while to "complete" and affect every word. They follow an s-shaped curve, affecting a small number at first and a lot all at once, but can abort at any time, so that you reasonably often get irregularities, either only a few affected at once or only a few which don't get affected.

Hmm, to what extent does that really happen? Examples? I think the point of us saying you should assume the Neogrammarian hypothesis is that it makes it much easier to deal with the consequences of sound changes. There are a lot of small details in linguistic theory that most conlangers don't really need to be concerned with. Diachronic conlanging is enough work without also having to deal with irregular or aborted sound changes.

Actually that's a big source of fun when you get to more advanced stages of diachronics, and can actually help when you want a certain form but a different form requires a different rule. It's a dirty little secret, yes, and noobs shouldn't try it, but eventually it can be helpful.

[Whimemsz]

The whole S-shaped curve and interruptibility of sound changes in real life is a major finding of Labovian sociolinguistics, so a number of books on historical linguistics and sociolinguistics should cover it (where I first read it was in Jean Aitchison's Language Change: Progress or Decay?). Historical linguists still act as though the Neogrammarian principle is correct, however, simply because it imposes a structure and discipline on their methodology--otherwise you can just point to any holes in your proposed correspondences and say "welp it must have been irregular"; assuming that sound changes are completely regular forces linguists to very carefully consider any potential "exceptions" to their proposed histories. It's also the case, of course, that the end result of most sound changes is in fact what the Neogrammarians described: the sound in question has changed regularly, in all words in which it occurred in the given environment. It's just that the way the change happened (irregularly moving through the lexicon and varying by speaker and sociolinguistics etc.) isn't how the Neogrammarians thought it worked.

[Ollock]

A little departure to celebrate a year of episodes: Conlangery#52: Conlangery at the Movies

[Ollock]

Conlangery #53: Topicalization

[clawgrip]

I'll have to listen to this one sometime. My language Himmaswa is very much a topic-prominent language.

[Torco]

I'll have to listen to this one sometime. My language Himmaswa is very much a topic-prominent language.

same here.

[clawgrip]

One big thing this episode seems to have left out is the idea of embedded topics. An example, taken from the Democratic Party of Japan's website:

提言は、近年の社会のＩＣＴ（情報通信技術）への依存度はますます高まっており…
Teigen-wa, kinnen-no shakai-no ICT (jōhō tsūshin gijutsu)-e-no izondo-wa masumasu takamatte-ori …
proposal-TPC recent.years GEN society GEN ICT (information communication technology) DAT NMZ degree.of.dependence-TPC steadily heighten-ing …
"As for the proposal, as for society’s degree of dependence on ICT in recent years, it is steadily increasing …"
"(On the topic of (the purpose of) the proposal), in recent years, society is steadily becoming increasingly dependent on ICT …"

The first topic is teigen 'the proposal', but in order to comment on the proposal, a second topic, namely kinnen-no shakai-no ICT-e-no izondo "society’s degree of dependence on ICT in recent years" needs to be introduced in the comment, and itself followed by a comment within the comment. This is partly the reason that the first topic is followed by a comma while the second one isn't.

This is an important and interesting part of topic-comment languages, (at least of Japanese, anyway), so it's a shame it wasn't covered.

[finlay]

Just listening to the newest ep about reduplication (which you seem to have forgotten to advertise on here), and I keep thinking about Japanese reduplication, which is common enough to have its own kanji.

So you get things like:
山=yama=mountain
山々=yamayama= mountain range (i think)

時=toki=time
時々=tokidoki= sometimes
The latter also shows rendaku.

And it's exceedingly common in expressive and onomatopoeic words.

Someone like clawgrip with better command of japanese can probably fill you in on the details, though.

[Ollock]

Just listening to the newest ep about reduplication (which you seem to have forgotten to advertise on here),

Oh! I'll have to fix that.

and I keep thinking about Japanese reduplication, which is common enough to have its own kanji.

So you get things like:
山=yama=mountain
山々=yamayama= mountain range (i think)

時=toki=time
時々=tokidoki= sometimes
The latter also shows rendaku.

On the character -- I am aware of this. Chinese also has a "ditto mark" like this, though the form is different.

On these derivations -- interesting. I very much like tokidoki

And it's exceedingly common in expressive and onomatopoeic words.

Yes. We mentioned in the episode and way back in the ideophones episode that reduplication in ideophones is extremely common cross-linguistically. I imagine the same is true for simple onomatopeia.