(Conlangs) Q&A Thread - Quick questions go here [2010-2020]

[Shemtov]

Does anybody know the phonology of Old Prussian? I'm asking here because I want to make a conlang that postulates that Old Prussian survived into a "Modern Prussian" language, and while I found a free grammar, it doesn't have the phonology.

[Chagen]

Is it plausible for a language to mandatorily mark number on demonstratives/determiners but not nouns themselves?

[Jackk]

Is it plausible for a language to mandatorily mark number on demonstratives/determiners but not nouns themselves?

I believe this is pretty much true for spoken French - almost all nouns mark number only on the demonstrative:
la table /la tabl/ ; les tables /le tabl/

(There are a few exceptions but these are comparatively rare and, afaik, a closed​ set: animal/animaux, œuf/œufs, œil/yeux are the only ones I can think of off the top of my head.)

[Sumelic]

Is it plausible for a language to mandatorily mark number on demonstratives/determiners but not nouns themselves?

I believe this is pretty much true for spoken French - almost all nouns mark number only on the demonstrative:
la table /la tabl/ ; les tables /le tabl/

(There are a few exceptions but these are comparatively rare and, afaik, a closed​ set: animal/animaux, œuf/œufs, œil/yeux are the only ones I can think of off the top of my head.)

There are also many adjectives and a fair number of nouns ending in "al" that follow the "animal/animaux" pattern. But still quite a small amount relative to regular nouns. One odd thing about French is that there is even number neutralization for a fair amount of the 3rd-person pronoun system: the feminine 3s disjunctive pronoun "elle" sounds identical to 3plr "elles", and the subject pronouns for both genders usually sound identical in singular (elle, il) and plural (elles, ils), only being distinguished before words that start with vowels, by the presence of a following liaison consonant /z/.

[Jackk]

Is it plausible for a language to mandatorily mark number on demonstratives/determiners but not nouns themselves?

I believe this is pretty much true for spoken French - almost all nouns mark number only on the demonstrative:
la table /la tabl/ ; les tables /le tabl/

(There are a few exceptions but these are comparatively rare and, afaik, a closed​ set: animal/animaux, œuf/œufs, œil/yeux are the only ones I can think of off the top of my head.)

There are also many adjectives and a fair number of nouns ending in "al" that follow the "animal/animaux" pattern. But still quite a small amount relative to regular nouns. One odd thing about French is that there is even number neutralization for a fair amount of the 3rd-person pronoun system: the feminine 3s disjunctive pronoun "elle" sounds identical to 3plr "elles", and the subject pronouns for both genders usually sound identical in singular (elle, il) and plural (elles, ils), only being distinguished before words that start with vowels, by the presence of a following liaison consonant /z/.

Ah yes, of course! I remember the il vs ils lack of distinction annoying me very much while I was learning.

Another one that amuses me is l'os /lɔs/ vs les os /le.zo/ , where the noun (meaning bone) is spelt identically but pronounced differently.

[LinguoFranco]

How would you romanize /ɕ/ if <x> is being used for /x/ and you don't want to use diacritics?

[Iyionaku]

<ch>, <zy> or <si> seem to be plausible choices.

Or you go full Swedish and use <kj>.

I for myself have used <c> for that sound in Caelian.

Does anybody know the phonology of Old Prussian? I'm asking here because I want to make a conlang that postulates that Old Prussian survived into a "Modern Prussian" language, and while I found a free grammar, it doesn't have the phonology.

Pages 16 to 20 of this document might be found helpful by you.

[Creyeditor]

How would you romanize /ɕ/ if <x> is being used for /x/ and you don't want to use diacritics?

Depending on the rest of your system, you could use <s> or <c>. <q> is an option if you want a more exotic option.
If you are into digraphs, you use <sh> (English style), <sy> (Malay style) or sj (Dutch style).

[LinguoFranco]

How would you romanize /ɕ/ if <x> is being used for /x/ and you don't want to use diacritics?

Depending on the rest of your system, you could use <s> or <c>. <q> is an option if you want a more exotic option.
If you are into digraphs, you use <sh> (English style), <sy> (Malay style) or sj (Dutch style).

Well, I'm thinking I could use <j>, but I think that may make the language look to Spanishy. I'm already using <s> and <c>.

The only reason I don't want to use digraphs is because they don't appear anywhere else in my language.

[Imralu]

Is it plausible for a language to mandatorily mark number on demonstratives/determiners but not nouns themselves?

Apart from the invariable indefinite article he, all Māori determiners mark plural or singular and only a very small class of nouns mark plural.

te whare = the house
ngā whare = the houses
tēnei whare = this house
ēnei whare = these houses
tōku whare = my house
ōku whare = my houses
tō te tangata whare = te whare o te tangata = the person's house
ō te tangata whare = ngā whare o te tangata = the person's houses
tō ngā tāngata whare = te whare ō ngā tāngata = the people's house
ō ngā tāngata whare = ngā whare ō ngā tāngata = the peoples houses

I can't remember, but I think there are around ten or fewer nouns that have a separate plural form, generally formed by lengthening a vowel. I think they're all words for people. The ones I can remember now are.

te tangata = the person
ngā tāngata = the people
te matua = the parent
ngā mātua = the parents
te wahine = the woman
ngā wāhine = the women
te tamaiti = the child
ngā tamariki = the children

[Omzinesý]

How would you romanize /ɕ/ if <x> is being used for /x/ and you don't want to use diacritics?

Depending on the rest of your system, you could use <s> or <c>. <q> is an option if you want a more exotic option.
If you are into digraphs, you use <sh> (English style), <sy> (Malay style) or sj (Dutch style).

Well, I'm thinking I could use <j>, but I think that may make the language look to Spanishy. I'm already using <s> and <c>.

The only reason I don't want to use digraphs is because they don't appear anywhere else in my language.

Go etymological!

If /S/ derives from
/s/ before /i/, you could mark it with <s> before <i> and <si> elsewhere.
/sk/ cluster, you can mark it <sk> or <sc> or something like that
/sx/, you can mark it <sh> or <sx>
/x/ before front vowels, you can mark it with <x> before them and say <xi> elsewhere.

[holbuzvala]

Have I created an accidentally erg/abs system? Here is a paradigm for verbs, intransitive and transitive, with human and inanimate arguments:

Ta-tsene
INAN.SBJ-COME
"It comes."

Ki-tsene
HUMAN.SBJ-COME
"He/she comes."

Ta-ki-zaga
INAN.SBJ-HUMAN.OBJ-SEE
"It sees him."

Ki-ta-zaga
HUMAN.SBJ-INAN.OBJ-SEE
"He sees it."

As the subject and object marker are only distinguished by place, with the 'most objectlike' appearing closest to the verb, have I inadvertantly made an ergstive absolutive system here? (If we treat the erative/absolutive marking as being proximity to the verb root as opposed to an affix)

NB. "He gives it to him" = ki-ta-ki-gaba
That's because the recipient has the most proto-object properties, so that human marker will occur closer to the verb than the 'direct' object.

[Micamo]

Would "he sees him" be "ki-ki-zaga"?

[holbuzvala]

Yes (as the language currently stands).

[Frislander]

have I inadvertantly made an ergstive absolutive system here?

No, at least not from this evidence. You don't have any other distinguishing feature between the two sets. Therefore we have no way of telling whether the prefix in the intransitive verb falls into one slot or the other. The only way we could tell was if there was some other kind of morphology which comes in between the two sets (à la Bantu): in that case, if the additional morphology comes before the intransitive markers then it is erg-abs, but nom-acc if it comes after. So if you had a past tense prefix ma- (where Ki-ma-ta-zaga is "He saw it", then Ma-ki-tsene would be erg-abs and Ki-ma-tsene would be nom-acc.

[sangi39]

I'd agree with Frislander on this. In the examples given it would be ambiguous at to whether the intransitive markers are in some "agent" or "patient" slot.

One thing you could do to make it erg-abs is to have something similar to Zulu, but messed around with a bit. In Zulu, the subject and object prefixes for the majority of noun classes are identical. However, the object markers for the first person singular and plural and the second person plural differ by tone, and the second person singular takes -kú- instead of ù-, and the object prefi for class one nouns is -ḿ-, as opposed to the subject marker ú-.

This is fairly similar to what you might expect to see with an animacy hierarchy in place. Less animate nouns usually have similar markings for when they're the agent and when thy're the patient, while higher animate nouns and pronouns have different markings.

Zulu is nom-acc as far as I know, but you could make it erg-abs by having intransive verbs take the object prefixes rather than the subject prefixes. For the vast majority of noun classes, this wouldn't make a difference, e.g. izinja zihamba (the dogs go) would remain as such, since the class 10 subject and object markers are both zi-. However, uhamba (you [sing.] go) would switch to kuhamba.

Some prefixes in Zulu, like the progressive -ya-, also appear before object markers (something Frislander pointed to), so here we would see a difference in izinja ziyahamba (the dogs are going), switching to izinja yazihamba with uyahamba (you [sing.] are going) likewise switching to yakuhamba.

[holbuzvala]

Thanks sangi39 and Frislander - that clears it up. I had planned for all verb modifiers to follow the verb stem, so it won't be à la Bantu, alas (as flavourful as it is).

Though, I then must ask, does the following paradigm of noun marking make it split erg abs?

Spoiler:

Key:
A = agent
S = subject
P = patient
O = other (beneficiary, destination, etc.)
X = marked

Edited by Sangi39

[Frislander]

Thanks sangi39 and Frislander - that clears it up. I had planned for all verb modifiers to follow the verb stem, so it won't be à la Bantu, alas (as flavourful as it is).

Though, I then must ask, does the following paradigm of noun marking make it split erg abs?



Key:
A = agent
S = subject
P = patient
O = other (beneficiary, destination, etc.)
X = marked

The image doesn't show up here. In the editing thing of Imgur it should give you a link you can use in BBCode, the programming language used here on this site: you can't put plain URLs in the image tags and expect it to work. Anyway, following up the link, yes I suppose, though you've picked the Monster Raving Loony candidate for your places, and having your "ergative" only on abstract nouns is probably the most unnatural place to put it I've seen.

[holbuzvala]

There is a logic to each, however, whereby things are marked where you don't EXPECT to find them as that argument of a discourse.

One would expect humans to be agents, so they are marked for patient and other.
One would expect animate things (incl rivers) to be agents or patients, but not usually other. (When was the ladt time you did something for a river?)
Inanimate things can just as easily be agents, patients, or 'other' so are unmarked.
Abstract things (ideas, feelings, substances), don't usually enact any change in other things, and are thus marked as agents, but not as patients or other.
And lastly, one expects a place to be a destination or location, but much rarer to see them as patients and agents.
And everything is unmarked for being a subject as that just seems tidy.
What do you think?
(Asp, just because there are x's everywhere doesnt mean that the marker will be the same. They just show that there is overt marking on the noun)

[qwed117]

It's most similar to my understanding of active-stative alignment. That's a relatively rare alignment linguistically, but does occur, though, I'm not sure if it does in the form that you ask about.