If natlangs were conlangs

[Rounin Ryuuji]

I'm sure somebody must have mentioned it already, but Japanese has a frightening orthography that does things with Kanji I would never have dared! Why would the creator give so many readings to 生 — it can be なま (nama) or せい (sei) on its own, but also い (i) in 生き (いき - iki) and う (u) in 生む(うむ - umu) and their derivatives. There's also a Sino-Japanese reading しょう (shō), though it's fairly rare, to my understanding. I know it's nice being alive, but this is rather absurd.

[Kuchigakatai]

I'm sure somebody must have mentioned it already, but Japanese has a frightening orthography that does things with Kanji I would never have dared! Why would the creator give so many readings to 生 — it can be なま (nama) or せい (sei) on its own, but also い (i) in 生き (いき - iki) and う (u) in 生む(うむ - umu) and their derivatives. There's also a Sino-Japanese reading しょう (shō), though it's fairly rare, to my understanding. I know it's nice being alive, but this is rather absurd.

Related anime clip:

https://www.youtube.com/watch?v=nHd3xfDzTg8

Don't use English or katakana for everything! I can't remember those!
Don't write 鎮魂歌 (chinkonka) 'elegy' and read it レクイエム rekuiemu (requiem)!
Don't write 禁忌 (kinki) 'forbidden, taboo' and read it タブー tabuu (taboo)!
Don't write 聖戦 (seisen) 'holy war' and read it ジハード jihaado (jihad)!
Greek mythology, the Bible, Norse mythology, Japanese mythology, don't start talking about them just because you did a little research!
If you don't properly explain, I can't understand what you mean!

Hittite...

[Rounin Ryuuji]

Well, that was a thing.

[zompist]

OK, inventor of French, it's great that you've got some derivational morphology worked out, but really...

fier 'trust'
> fiance (nominalization) 'trust' > 'engagement'
> fiancer (verbalization) 'get engaged'
> fiançailles (nominalization) 'engagement'

Don't tell me, next is fiançaillage 'engagement'.

[Rounin Ryuuji]

I would not bat an eye at that being a French word.

Also, English is a... very interesting idea for a Germanic language, but I have a feeling the orthography is a little too creative — especially with how half-Gallicised it is, obscuring all those nice historical sound changes — and the phonology might sound a little nicer without so many /ai/ and /au/ diphthongs all over; I admit that, while /maus/ and /mais/ being a singular-plural patterning, writing one mouse and the other mice is taking even this degree too far.

I approve of keeping the dental fricatives, however.

[bradrn]

OK, inventor of French, it's great that you've got some derivational morphology worked out, but really...

fier 'trust'
> fiance (nominalization) 'trust' > 'engagement'
> fiancer (verbalization) 'get engaged'
> fiançailles (nominalization) 'engagement'

Don't tell me, next is fiançaillage 'engagement'.

My personal favourite (in English) is emerge → emergency.

I would not bat an eye at that being a French word.

Also, English is a... very interesting idea for a Germanic language, but I have a feeling the orthography is a little too creative — especially with how half-Gallicised it is, obscuring all those nice historical sound changes — and the phonology might sound a little nicer without so many /ai/ and /au/ diphthongs all over; I admit that, while /maus/ and /mais/ being a singular-plural patterning, writing one mouse and the other [/i]mice[/i] is taking even this degree too far.

I approve of keeping the dental fricatives, however.

Ah yes, orthographies are a regular topic of discussion here. The usual mentions are English, French, Irish, Tibetan, Kiowa and the frighteningly creative SENĆOŦEN; lesser-known ones are Gwoyeu Romatzyh, Northern Qiang and Mediæval Arpitan.

[Ares Land]

OK, inventor of French, it's great that you've got some derivational morphology worked out, but really...

fier 'trust'
> fiance (nominalization) 'trust' > 'engagement'
> fiancer (verbalization) 'get engaged'
> fiançailles (nominalization) 'engagement'

Don't tell me, next is fiançaillage 'engagement'.

I hadn't even realized the connection with confiance and fier! And I think the only reason fiançaillage isn't a word is because engagements aren't as important as they used to be.

[bradrn]

The Papuan language Burmeso has what is possibly the most perverse system of noun classes I have ever seen (Donohue 2001, Foley 2018). To start us off, there are six classes:

I: Masculine and general animate; largest class (‘contains almost half of all nouns’)
II: Feminine
III: Miscellaneous (‘containing nouns for body parts, insects and lizards, many aspects of material culture like axes and canoes, some foods, and many natural phenomena’)
IV: Body parts, mass nouns (including e.g. ‘head’, ‘sun’)
V: Sago trees and bananas only
VI: Arrows, coconuts and rice only

You may think that this system might be just a teensy bit on the lopsided side. Yet this is just the beginning! To continue, the verb agrees in noun class with its absolutive argument. Matching the six noun classes, there are six agreement markers: j-, g-, s-, b-, n-, t-. Now, at this point, you might be tempted into thinking that each noun class is marked with one of these prefixes. Do not be fooled! Such a crude system is clearly not nearly suitable enough for the refined linguistic tastes of the Burmeso people. Instead, they use the following rather more linguistically interesting system:

burmeso agreement.png (11.71 KiB) Viewed 13485 times

…yep, Burmeso halves its effective agreement marker inventory by separating verbs into two classes, then makes those three agreement markers stretch to six noun classes! The verbal classes, you will no doubt be pleased to hear, are lexically determined, with no particular regularity. Thankfully, the two sets at least have the same patterns, with j-, g-, s- in A corresponding to b-, n-, t- with B. Oh, and I should probably mention that pronouns agree using exactly the same markers; as Foley says, ‘first person is treated like a feminine noun, while second person like a masculine one’. Possessive prefixes, too, use the same markers: g- for first person singular, j- for second and third person singular, and s- for non-singular.

Ah, but this is not the end of the story! For Burmeso also has a cross-cutting gender system, marked on adjectives (except for the ones which don’t agree at all, and the ones which agree in number only, and the three which agree in both gender and noun class, as well as attributive verbs, which also agree in both gender and noun class). There are three genders: masculine, feminine and neuter. ‘A-ha!’, I hear you cry, ‘but surely masculine and feminine genders line up with noun classes I and II!’. I am sorry to say, but this is not so, though there is a tendency to correlation. Additionally, each gender is divided into two sub-classes, so Burmeso actually has six genders: M1, M2, F1, F2, N1 and N2. Gender agreement on adjectives is via seven suffixes: -ab, -an, -o(ra), -od, -od(o), -or, -or(o). Yet, despite having more suffixes than genders, the system somehow still manages to be underspecified. In the singular, the sub-groups are merged:

M1/2: -ab
F1/2: -an
N1/2: -o(ra)

And in the plural, the main groups are merged, except for neuter:

M/F1: -od(o)
M/F2: -or
N1: -od
N2: -or(o)

[Otto Kretschmer]

Someone who created Proto Indo European certainly disliked chemistry in high school.

And Chinese WTF? Are we supposed to learn a symbol for every word?

[dɮ the phoneme]

What?

How in the world did a system like this develop?

[Vijay]

Is it that surprising? How did any noun classification system develop?

At least the categories would seem to make cultural sense. Sago and bananas are indigenous staples in New Guinea; coconuts and rice seem to have been introduced from elsewhere.

[Richard W]

Is it that surprising? How did any noun classification system develop?

And the gender subclasses are called animate and inanimate, so the gender mergers aren't so weird.

[Otto Kretschmer]

The creator of Ubykh was very creative. Just two vowels .

[bradrn]

Is it that surprising? How did any noun classification system develop?

And the gender subclasses are called animate and inanimate, so the gender mergers aren't so weird.

Nominally, yes, that’s what Donohue calls them; but Foley points out that there’s actually not very much about them which is ‘animate’ or ‘inanimate’.

The creator of Ubykh was very creative. Just two vowels .

Also Margi, Arrernte and (possibly) Ndu languages.

[Richard W]

Nominally, yes, that’s what Donohue calls them; but Foley points out that there’s actually not very much about them which is ‘animate’ or ‘inanimate’.

I suspect you don't appreciate the difference between PIE *h₁n̥gʷnis and *péh₂wr̥, 'fire'.

[Nortaneous]

The creator of Ubykh was very creative. Just two vowels .

Also Margi, Arrernte and (possibly) Ndu languages.

Tocharian A is the worst conlang - vertical vowel systems are bad enough, but developing them in Indo-European?

[Richard W]

And Chinese WTF? Are we supposed to learn a symbol for every word?

Isn't that how English is alleged to work? One problem is that schemes like ideographic description sequences don't quite work.

[bradrn]

Nominally, yes, that’s what Donohue calls them; but Foley points out that there’s actually not very much about them which is ‘animate’ or ‘inanimate’.

I suspect you don't appreciate the difference between PIE *h₁n̥gʷnis and *péh₂wr̥, 'fire'.

Not only do I not appreciate it, I didn’t even know PIE had two separate words for it! Anyway, here’s Foley’s exact words:

Donohue divides [noun classes] into sub-classes according to animacy, claiming that there is an animate and an inanimate subclass for each gender, but looking at the membership of the classes, I see no grounds for that. His animate neuter class contains no nouns with animate reference, and while it is true that his masculine animate class contains nouns with male human referents, it also contains nouns with inanimate referents, no different from the inanimate masculine class. So I will designate them just with numerals as the subclasses of the genders, i. e. M1.

And Chinese WTF? Are we supposed to learn a symbol for every word?

Isn't that how English is alleged to work?

…no, it isn’t.

[Richard W]

…no, it isn’t.

The Chinese also work out how to pronounce compound ideographs they haven't seen before.

[dɮ the phoneme]

German has fantastically interesting syntax and nice morphology as well, but god it sounds ugly. There are so many great sounding Germanic languages as well, why in the world would you make a Germanic lang and then have it sound like that?