Conlang Random Thread

[bradrn]

I don't know, do these look naturalistic?

They look fine to me. Of course, I can’t say anything about your phonaesthetic preferences.

That's what I thought too, but then I keep looking at it and doubting myself. I have tried to find naturalistic languages where all the case endings end in the same vowels but even they have a little variation.

Far stranger things have happened: e.g. the Austronesian languages where all nouns begin with /n/ (due to fusion of a former article).

[Travis B.]

I think the system you have is fine, but if you really want more variation what I would suggest is removing the final consonant of one of the animate singular and one of the inanimate singular endings (e.g. have the animate accusative and inanimate genitive end in a bare stem vowel).

[malloc]

Far stranger things have happened: e.g. the Austronesian languages where all nouns begin with /n/ (due to fusion of a former article).

Whoa really? Do you remember the name of this language, apart presumably from starting with /n/?

[bradrn]

Far stranger things have happened: e.g. the Austronesian languages where all nouns begin with /n/ (due to fusion of a former article).

Whoa really? Do you remember the name of this language, apart presumably from starting with /n/?

Sakao, amongst others. Apparently it’s quite widespread in the region. However, it looks like in all the languages where this occurs, some subclasses of nouns are excluded, most usually animate nouns (which took a different article originally). Also, it seems common for the initial /n/ to disappear in closely-bound constructions like noun compounds.

[Ahzoh]

I think the system you have is fine, but if you really want more variation what I would suggest is removing the final consonant of one of the animate singular and one of the inanimate singular endings (e.g. have the animate accusative and inanimate genitive end in a bare stem vowel).

I was thinking, maybe the i-stem and u-stem inanimate genitive singular cases could be long vowels, but I started to cook with hyper-reducing the cases. Kinda reminds me like Pashto or Hindi/Urdu

Animate a-stem Singular Plural Nominative -am -ī Vocative -ē -ī Accusative -as -an Genitive -as -an Equative -īli -īli

Animate i-stem Singular Plural Nominative -im -ī Vocative -ē -ī Accusative -is -in Genitive -is -in Equative -īli -īli

Animate u-stem Singular Plural Nominative -um -ū Vocative -ē -ū Accusative -us -un Genitive -us -un Equative -īli -īli

Inanimate a-stem Singular Plural Nominative -as -an Vocative -as -an Accusative -as -an Genitive -aḫ -ī Equative -īli -īli

Inanimate i-stem Singular Plural Nominative -is -in Vocative -is -in Accusative -is -in Genitive -iḫ -ī Equative -īli -īli

Inanimate u-stem Singular Plural Nominative -us -un Vocative -us -un Accusative -us -un Genitive -uḫ -ū Equative -īli -īli

gībā asmas
wing-CNS.INAN.NOM.PL eagle.GVS-AN.GEN.SG
"the eagle's wings"

[jal]

my limited knowledge of sound changes

You don't need 'knowledge' of sound changes per se. I assume you have enough knowledge to know that k > l doesn't go, but m > b does, for example. But it's impossible to suggest more differentiated endings without knowing more about the phonology and phonitactics of the language proper.

EDIT: You could also look at harmonizing patterns, like having the -ili in the plurals become -ini (l > n isn't an unusual sound change) because of the n in the other forms.


JAL

[Ahzoh]

my limited knowledge of sound changes

You don't need 'knowledge' of sound changes per se. I assume you have enough knowledge to know that k > l doesn't go, but m > b does, for example.

What I mean is turning something like /lɛsp-ʔətsɒpʰ-ə-sə/ into /lɛsp:su:ɛsi:/. These aren't crazy sound changes, but they're not something I would immediately come up with.

Nor would I immediately come up with the sort of shenanigans Tiramisu did to turn <inpandārāzna> into <pindās>

These kinda things would take me months, maybe years to get such inspiration. Yet people can casually rattle these off onto a quick-n-dirty scratchpad.

And it's not like I haven't spent significant amounts of time pouring over a variety of attested and theoretically plausible sound changes.

But it's impossible to suggest more differentiated endings without knowing more about the phonology and phonitactics of the language proper.

• Syllable structure is CV(:)(C), Superheavy syllables (CV:C) are dispreferred and usually result from vowel coalescence which in turn form from intervocalic elision of weak consonants (glides and laryngeals). They may be left alone (/ˈka.pi.ʔas/ > /ˈka.pi.as/ > /ka.ˈpa:s/) or shortened (/na.ˈhab.bis/ > /na.ˈab.bis/ > /ˈna:b.bis/ > /ˈnab.bis/).
• Weak consonants that are adjacent to strong consonants (in front or behind) elide and lengthen the preceding vowel.
• Stress falls on the heaviest non-final syllable; if there are no heavy non-final syllables then stress is on the penultimate (two syllables) or antepenultimate (more than two syllables).
• There cannot be more than two unstressed light syllables in a row; it is not allowed and at least one of the vowels must elide or lengthen
• Word-final short vowels are subject to elision, especially if it's in the third weak syllable in a row; the second-final short vowel might elide or lengthen instead.

Phonotactics are essentially any strong consonant can appear in any position, though mixed voicing of obstruents of the same POA and MOA are not allowed; the first element assimilates in voicing. There also can't be mixed clusters of central and lateral obstruents; the first element of the cluster assimilates. Lastly, stops preceding fricatives of the same POA become fricatives.

/m n ŋ/ <m n n̮>
/pʰ pʼ b tʰ tʼ d kʰ kʼ g ʔ/ <p ṗ b t ṭ d k ḳ g ʔ>
/f s sʼ z ɬ ɬʼ ɮ x/ <f s ṣ z ś ṣ́ ź ḫ>
/w r l j// <w r l y>

And, of course, /e e:/ come mainly from the elision of pharyngeals, while /Vj Vi/ bcome /i:/ and /Vw Vu/ become /u:/.

[jal]

Nor would I immediately come up with the sort of shenanigans Tiramisu did to turn <inpandārāzna> into <pindās>

I have to opposite problem, I always see these wild possibilities and then something like Alũbetah happens :D.


JAL

[Ahzoh]

My latest verb morphology is something of a blend of Semitic and Ojibwe. A notable rip off of Ojibwe is the form of the verb known as the "conjunct", which is used for dependent and relative clauses among other things.

Strong Independent Person Realis Irrealis Imperative 1st Singular na-rmaṣ na-rammaṣ 2nd Singular ma-rmaṣ ma-rammaṣ ramṣ-ū 3rd Singular ta-rmaṣ ta-rammaṣ 1st Plural na-rmaṣ-ā na-rammaṣ-ā 2nd Plural ma-rmaṣ-ā ma-rammaṣ-ā ramṣ-ā 3rd Plural ta-rmaṣ-ā ta-rammaṣ-ā

Strong Conjunct Person Realis Irrealis 1st Singular ramaṣ-ni rammaṣ-ni 2nd Singular ramaṣ-mi rammaṣ-mi 3rd Singular ramaṣ rammaṣ 1st Plural ramaṣ-nin rammaṣ-nin 2nd Plural ramaṣ-min rammaṣ-min 3rd Plural ramaṣ-ā rammaṣ-ā

Though I think maybe the conjunct verb forms should all have suffixes

Strong Conjunct Person Realis Irrealis 1st Singular ramaṣ-ni rammaṣ-ni 2nd Singular ramaṣ-mi rammaṣ-mi 3rd Singular ramaṣ-ti rammaṣ-ti 1st Plural ramaṣ-nin rammaṣ-nin 2nd Plural ramaṣ-min rammaṣ-min 3rd Plural ramaṣ-tin rammaṣ-tin

Parraḫni ramṣā!
When I speak, you listen/hear!

[malloc]

I have to opposite problem, I always see these wild possibilities and then something like Alũbetah happens .

Cool, that is some interesting allophony.

The hexadecimal approach is certainly appealing from an abstract conceptual standpoint and the fact that the numerals work so well. That said, it also feels distinctly artificial even by the standards of my highly stylized minimalist project. At the very least, I would need to work on humanizing the system as it were.

[bradrn]

That said, it also feels distinctly artificial even by the standards of my highly stylized minimalist project. At the very least, I would need to work on humanizing the system as it were.

Perhaps it would be sufficient to adopt a more handwritten style? I played around a bit with my broad-edged nib and came up with this alternate style, as well as an attempt at a ballpoint adaption:

malloc-numerals.jpeg (102.06 KiB) Viewed 882 times

(Please forgive the mistakes…)

[malloc]

That really is amazing. You could make quite a good career as a calligrapher. That said, my point is more that human languages rarely use binary or anything approximating it. The system I've sketched makes sense if you're already familiar with mathematics and especially computer science but would feel quite alien to someone intuitively counting with their fingers. Nobody would naturally think to count "one two two-one four four-one..." after all.

[bradrn]

That really is amazing. You could make quite a good career as a calligrapher.

Thanks! Though if you look at it more closely it’s a bit too irregular to be really good.

Nobody would naturally think to count "one two two-one four four-one..." after all.

Yes, but counting ‘one, two, three, four, five… fifteen, sixteen, one-zero’ is a lot less unnatural. To me the internal structure of the numerals is just a neat design detail, sort of like how Hangeul letters have phonetically justified shapes.

[foxcatdog]

The refined phonology.

[Ahzoh]

Contemplating another adverbial case marker, which I call the instrumental-locative. Mainly used to describe means by which certain actions are done (go by foot, hit/strike by hand, kill it with fire, etc.), but it can also indicate direction (pour into the vessel) or path (gandē namātā "through the mountain(s) we went"). Like the equative, it is another means by which adverbs are conveyed.

The morpheme is -ʕ(V), which due to sound changes results in the case being the same in every paradigm and syncretic in number

Code: Select all

 SG /  PL   >  SG /  PL
-aʕ / -aːʕa > -eː / -eː
-iʕ / -iːʕa > -eː / -eː
-uʕ / -uːʕa > -eː / -eː

Maybe I'll have limited passives such as "X died by Y" = "X was killed by Y"

Bonus sentence:
Mātnin gandē napraḫ ʔAsmas ʔit=tarmaṣ nīya
[ˈmɑ̂ːt.nin ˈgɑ́n.deː ˈnɑ́p.rɑx ˈʔɑ́s.mɑs ʔit.ˈtár.masʼ ˈnîː.jɑ] (Imperial)
[ˈmɑːtʰ.nin ˈkɑn.teː ˈnɑpʰ.rɑx ˈʔɑs.mɑs ʔit.ˈtʰar.mat͡s ˈniː.jɑ] (Common)
"As we went through the mountain(s), I spoke/called out to Eagle and he listened to/heard me"

[malloc]

Yes, but counting ‘one, two, three, four, five… fifteen, sixteen, one-zero’ is a lot less unnatural. To me the internal structure of the numerals is just a neat design detail, sort of like how Hangeul letters have phonetically justified shapes.

The current proposal is essentially binary up to sixteen rather than having separate roots for numbers up to sixteen. The latter approach would certainly make the number system less crudely artificial, but it would lose the minimalism that originally motivated the proposal.

[bradrn]

Yes, but counting ‘one, two, three, four, five… fifteen, sixteen, one-zero’ is a lot less unnatural. To me the internal structure of the numerals is just a neat design detail, sort of like how Hangeul letters have phonetically justified shapes.

The current proposal is essentially binary up to sixteen rather than having separate roots for numbers up to sixteen. The latter approach would certainly make the number system less crudely artificial, but it would lose the minimalism that originally motivated the proposal.

I think it would be possible to make number words up to 16 which are made up of recognisable binary submorphemes yet remain coherent words of their own, similar to what I tried to do with the written numerals. After all, lots of natlangs have number words made up of smaller components.

[Travis B.]

Yes, but counting ‘one, two, three, four, five… fifteen, sixteen, one-zero’ is a lot less unnatural. To me the internal structure of the numerals is just a neat design detail, sort of like how Hangeul letters have phonetically justified shapes.

The current proposal is essentially binary up to sixteen rather than having separate roots for numbers up to sixteen. The latter approach would certainly make the number system less crudely artificial, but it would lose the minimalism that originally motivated the proposal.

I think it would be possible to make number words up to 16 which are made up of recognisable binary submorphemes yet remain coherent words of their own, similar to what I tried to do with the written numerals. After all, lots of natlangs have number words made up of smaller components.

It should be remembered that it is quite common in natlangs to form 6 through 9 with 5 plus 1 through 4 and 11 through 19 with 10 plus 1 through 9.

[rotting bones]

In technical contexts, English uses numbers with a base of 2 (edit: or 16). I've been thinking Ancient Galactic uses different bases in different ordinary contexts.

[malloc]

In technical contexts, English uses numbers with a base of 2 (edit: or 16). I've been thinking Ancient Galactic uses different bases in different ordinary contexts.

How so? Programmers use binary and hexadecimal, but not usually spoken and certainly not for cardinal numbers. Not even the most hard-core techbro would count to "four sixteens and five".