Conlang Random Thread

[trailsend]

It's a conlang; reality can be whatever I want it to be.

Echoing bradrn, the thing you have full control over as a creator is the conlang's data; you get to say what statements are grammatical, and what statements are ungrammatical, and in which contexts and for what purposes you might say a statement.

You can't declare what the best analysis of some part of the language is by authorial fiat, though. If there's some analysis you want to fit best, that's fine, but you have to accomplish that by tweaking the data so that your desired analysis is actually the best fit—otherwise it doesn't mean much.

You can't snap a part of speech into existence with your godlike authorial reality powers, but you can snap a bunch of words and syntactic structures into existence such that the best analysis someone can give for them is to posit a particular part of speech.

[bradrn]

It's a conlang; reality can be whatever I want it to be.

Echoing bradrn, the thing you have full control over as a creator is the conlang's data; you get to say what statements are grammatical, and what statements are ungrammatical, and in which contexts and for what purposes you might say a statement.

You can't declare what the best analysis of some part of the language is by authorial fiat, though. If there's some analysis you want to fit best, that's fine, but you have to accomplish that by tweaking the data so that your desired analysis is actually the best fit—otherwise it doesn't mean much.

You can't snap a part of speech into existence with your godlike authorial reality powers, but you can snap a bunch of words and syntactic structures into existence such that the best analysis someone can give for them is to posit a particular part of speech.

This is an excellent statement of what I was trying to say.

Also, I see this is your first post, so welcome to the board trailsend! Have some pickles and tea!

[Rounin Ryuuji]

Oh, are they new? I thought I'd seen that name before. Looks like I hadn't.

[bradrn]

Oh, are they new? I thought I'd seen that name before. Looks like I hadn't.

Oddly enough I had the same thought until I checked.

[Jonlang]

I've actually done a fair amount of work on my PQL conlang family, which seems to have been going on forever, with little progress. So, I'm feeling like I need to do something else for a bit, just to clear my mind of it and do something a bit more fun and weird.

I'm going to make a personal conlang with no regard for naturalism whatsoever. The plan is to choose a phoneme inventory with some of the rarer sounds (that I feel I can comfortably produce), particularly vowels (it won't have [a], [i], [e], [ɛ], [u], [o], or [ɔ], for example). Grammatically, I'm unsure where to go - I may have to look at some other languages outside of Europe for some inspiration, so I'm open to suggestions. I may go down the tri-consonantal route because I feel like this may be easier for a personal conlang, with words easily derived and recognisably related, but maybe not, I may look at Japanese (another language I'm interested in but never really studied).

[jal]

Do you, or anyone else, have examples of languages that are often analyzed as having a part of speech of modal particles?

The first Google hit is the Wikpedia article on modal particles. That would be a good start :).

To me prepositions and particles are weakly differentiated.

You mean "I have insufficient linguistic knowledge to tell what 'prepositions' and 'particles' are typically defined as", and that's fine I guess, but even Wikipedia has comprehensive overviews of both, (the latter as 'grammatical particle'), so gaining knowledge is just a few google searches away :). But don't use a lack of understanding of typical use as an excuse to do bogus analyzing.

They're both function words and they both take an argument of some kind.

Like others have written, particles do not take arguments, and prepositions don't really need to have an argument (or can have a null argument, or are magically converted into adverbs without an argument, depending on what analysis you prefer).

No, I meant modal particles. I have never seen a language claimed to have that as a part of speech.

I think you are too focussed on this "part of speech" thing. Many languages have modal particles, that aren't necessarily analyzed as something else (see Wikipedia link above). So if your language has otherwise not easily analyzable words with a modal meaning, they're modal particles.


JAL

[jal]

I'm going to make a personal conlang with no regard for naturalism whatsoever. (...) Grammatically, I'm unsure where to go - I may have to look at some other languages outside of Europe for some inspiration, so I'm open to suggestions.

Well, if there's no naturalism needed, I would not look at existing languages at first, but make a list of language features (possibly looking at WALS categories) that you personally like, or otherwise find attractive (for example because they're easy to learn), and start from that. You could always look at natural languages (or other conlangs) when you're out of ideas.


JAL

[masako]

help me decide how to sort my glyphs!

LINK

[Man in Space]

help me decide how to sort my glyphs!

LINK

I vote for the components-then-pataka method.

[trailsend]

Also, I see this is your first post, so welcome to the board trailsend! Have some pickles and tea!

Thank you!

Oh, are they new? I thought I'd seen that name before. Looks like I hadn't.

No, you likely have. I've been a member of various incarnations of the ZBB for a good while (I think I first registered in 2008?), but I hadn't reregistered since the last server change.

[cedh]

Oh, are they new? I thought I'd seen that name before. Looks like I hadn't.

No, you likely have. I've been a member of various incarnations of the ZBB for a good while (I think I first registered in 2008?), but I hadn't reregistered since the last server change.

Welcome back!

[bradrn]

Oh, wait a minute, I’ve just realised why your name seemed familiar — you’re the guy who made Hiding Waters! I remember it was recommended in this very thread about a year ago, but I never got around to reading through the grammar.

And incidentally, this also resolves a question I had a while ago:

Quick language identification question: I seem to remember reading about a language, probably Papuan, with pervasive multiple infixation (and possibly nested infixation as well). However, I can’t seem to find it now. Does anyone here happen to know which language that could be?

And now I realise why we never figured out the language I was asking about: it was Hiding Waters, which is a conlang. (Not Papuan at all, of course.)

[trailsend]

Welcome back!

Thank you! Great to see familiar faces about.

Oh, wait a minute, I’ve just realised why your name seemed familiar — you’re the guy who made Hiding Waters! I remember it was recommended in this very thread about a year ago, but I never got around to reading through the grammar.

That's me! And gosh, that's tremendously kind of quinterbeck

And incidentally, this also resolves a question I had a while ago:

Quick language identification question: I seem to remember reading about a language, probably Papuan, with pervasive multiple infixation (and possibly nested infixation as well). However, I can’t seem to find it now. Does anyone here happen to know which language that could be?

And now I realise why we never figured out the language I was asking about: it was Hiding Waters, which is a conlang. (Not Papuan at all, of course.)

Ah, yes! The bipartite roots are one of the few features of HW that it's held onto all the way from its inception. Will I ever manage to find a reasonable diachronic explanation for them? ...someday. someday

[Rounin Ryuuji]

Oh, are they new? I thought I'd seen that name before. Looks like I hadn't.

No, you likely have. I've been a member of various incarnations of the ZBB for a good while (I think I first registered in 2008?), but I hadn't reregistered since the last server change.

Oh, that would probably explain things.

[bradrn]

And incidentally, this also resolves a question I had a while ago:

Quick language identification question: I seem to remember reading about a language, probably Papuan, with pervasive multiple infixation (and possibly nested infixation as well). However, I can’t seem to find it now. Does anyone here happen to know which language that could be?

And now I realise why we never figured out the language I was asking about: it was Hiding Waters, which is a conlang. (Not Papuan at all, of course.)

Ah, yes! The bipartite roots are one of the few features of HW that it's held onto all the way from its inception. Will I ever manage to find a reasonable diachronic explanation for them? ...someday. someday

Oh, I’d say they hardly even need explanation — bipartite roots are reported from all over the world (Klamath, Nakh-Daghestanian, Limbu, Puare, possibly Skou), though from what I can see there are no cases as clear as Hiding Waters. Even so, it certainly doesn’t seem any more unusual than, say, triconsonantal roots.

As for diachronics, that’s easy — plenty of languages (especially Papuan, but others also) have adjunct+verb constructions, where the the adjunct is drawn from a special word class which only occurs with verbs. Simply restrict each adjunct to occurring with only one or two verbs, add a rule that all verbs must occur with an adjunct, and then fuse the whole thing into one word — voilà, bipartite stems. Really the only unusual thing about this is the rule that all verbs have an adjunct; normally adjunct+verb constructions are restricted to a handful of very general verbs like do. But all the other steps are well-attested.

[Zju]

If a language has unconjugated modal verbs, how do you know that they are verbs and not adverbs or prepositions or particles or something?

I think you are approaching this the wrong way. Whether a word with a modal meaning is a "verb" or an "adverb" or a "particle" or something else is a matter of language analysis - does the word behave like a verb syntactically or like an adverb? What are its diachronics? If there's no distinction between these two (e.g. because both auxiliary verbs and adverbs have the same syntactic position), does it really matter syncronically whether it's one or the other?

My absolute favourite article along these lines: http://alex.francois.online.fr/data/Ale ... -in-Lg.pdf. I think every conlanger should read it. (I’ve already linked it a handful of times here.)

Reading that paper, I honestly wonder doesn't English also have strong and weak nouns? Cf.:

* Word is missing.
A word is missing.

Father is missing.
A father is missing.

Granted, they'd have different syntactic functions than those in Hiw (as father can be modifier in NP).

[trailsend]

As for diachronics, that’s easy — plenty of languages (especially Papuan, but others also) have adjunct+verb constructions, where the the adjunct is drawn from a special word class which only occurs with verbs. Simply restrict each adjunct to occurring with only one or two verbs, add a rule that all verbs must occur with an adjunct, and then fuse the whole thing into one word — voilà, bipartite stems. Really the only unusual thing about this is the rule that all verbs have an adjunct; normally adjunct+verb constructions are restricted to a handful of very general verbs like do. But all the other steps are well-attested.

That's approximately the plan, in fact! Obligatory verbal complements on many verbs, which got extended by analogy to appear on almost all verbs, before getting reanalyzed as part of the verb. (I've been rigging much of the lexicon with this in mind; you can find a lot of roots that have a certain kind of semantic commonality and share pre-stems, because they descend from the same verb used with different complements, and a lot of roots with a different kind of semantic commonality that share post-stems, because they descend from different verbs used with the same complement.)

The tricky bit has been telling that story in a way that also accounts for the incorporation system, and lets the rest of the predicate template fall together too. I'd love some help working through the details—I'll post a thread one of these days if you think there'd be interest.

[quinterbeck]

—I'll post a thread one of these days if you think there'd be interest.

Yes please!

[Nortaneous]

Oh, I’d say they hardly even need explanation — bipartite roots are reported from all over the world (Klamath, Nakh-Daghestanian, Limbu, Puare, possibly Skou), though from what I can see there are no cases as clear as Hiding Waters. Even so, it certainly doesn’t seem any more unusual than, say, triconsonantal roots.

As for diachronics, that’s easy — plenty of languages (especially Papuan, but others also) have adjunct+verb constructions, where the the adjunct is drawn from a special word class which only occurs with verbs. Simply restrict each adjunct to occurring with only one or two verbs, add a rule that all verbs must occur with an adjunct, and then fuse the whole thing into one word — voilà, bipartite stems. Really the only unusual thing about this is the rule that all verbs have an adjunct; normally adjunct+verb constructions are restricted to a handful of very general verbs like do. But all the other steps are well-attested.

Or fossilized N+V constructions that outlasted the N component as in Coptic.

[bradrn]

I think you are approaching this the wrong way. Whether a word with a modal meaning is a "verb" or an "adverb" or a "particle" or something else is a matter of language analysis - does the word behave like a verb syntactically or like an adverb? What are its diachronics? If there's no distinction between these two (e.g. because both auxiliary verbs and adverbs have the same syntactic position), does it really matter syncronically whether it's one or the other?

My absolute favourite article along these lines: http://alex.francois.online.fr/data/Ale ... -in-Lg.pdf. I think every conlanger should read it. (I’ve already linked it a handful of times here.)

Reading that paper, I honestly wonder doesn't English also have strong and weak nouns? Cf.:

* Word is missing.
A word is missing.

Father is missing.
A father is missing.

Granted, they'd have different syntactic functions than those in Hiw (as father can be modifier in NP).

This is really interesting. I wonder if there are any non-kin terms which act as ‘strong nouns’ in English?