(Conlangs) Q&A Thread - Quick questions go here

[Salmoneus]

Bare nouns, huh, good god y'all, what are they good for?

So, I have the following problem in Mamabam, that I have been thinking about a lot. Obligatory inflection adds a whole CVC syllable to all words. I don't like this, it makes most sentences feel somewhat clunky. Now, I came across the concept of bare nouns. In some natlangs, otherwise obligatory inflection can be absent in certain contexts. For nouns this is called Bare Nouns in the literature. As I already have some idea for verbs, I was wondering if you all have some ideas for contexts in which the otherwise obligatory inflection on nouns could be absent.

Any more information on the structure of the language or the type of information conveyed by inflection?

The two obvious times you wouldn't need inflection would be when the information is already conveyed by another part of speech (like argument marking on verbs, or agreement marking on adjectives or articles) or when the information is already known from context (like topics) or seen as of little importance (like backgrounded elements).

[Knox Adjacent]

True that, but in the meantime may I suggest inflection regarding the contiguity of a determiner à la Macro-Jê?

[Omzinesý]

I'm making suffixes that cause lenition/ellipsis of some preceding consonants

lak + at -> la:t
lan + at -> lã:t

But I wouldn't like to lose all intervocalic instances of those consonants. How could it historically happen that only some suffixes cause the lenition/ellipsis?

lak + ap -> lakap

taking inspiration from your own mother tongue, perhaps lenition only occurred in closed syllables so -at would cause lenition while -āt wouldn’t. Later on length distinctions were lost leading to seemingly identical suffixes (or at least identical in their phonotactic structure) with different outcomes.

Maybe possible. If the syffixes appear in the same slot, I thought they would be at least analogically affected by each other, but maybe it is not such a problem than I first thought.

Finnish, for example has many infinitives that have developed at different times. Haspelmath's paper on historical development of infinitives is btw interesting.

tule-ta -> tulla 'to come'
tule-maan -> tulemaan 'to come'

Maybe it was a simpler problem than what I thought.

taking inspiration from your own mother tongue, perhaps lenition only occurred in closed syllables so -at would cause lenition while -āt wouldn’t. Later on length distinctions were lost leading to seemingly identical suffixes (or at least identical in their phonotactic structure) with different outcomes.

You probably mean the Finnish historically closed syllables (pataan below). āt is of course not closed.

pata-ssa -> padassa (dialectally paassa) 'in the pot'
pata-han -> pataan 'to the pot'

It could even be that vowel-initial (more bleached) suffixes are added to a consonant stem while consonant-initial suffixes are added to a vowel-final stem. The lenition process operates only once.

lak-at -> la:t
lak+kap -> lakakap -> laka:p -> lakap

[Omzinesý]

Bare nouns, huh, good god y'all, what are they good for?

So, I have the following problem in Mamabam, that I have been thinking about a lot. Obligatory inflection adds a whole CVC syllable to all words. I don't like this, it makes most sentences feel somewhat clunky. Now, I came across the concept of bare nouns. In some natlangs, otherwise obligatory inflection can be absent in certain contexts. For nouns this is called Bare Nouns in the literature. As I already have some idea for verbs, I was wondering if you all have some ideas for contexts in which the otherwise obligatory inflection on nouns could be absent.

Maybe genitive attributes have Suffixaufnahme and the possessed noun could be uninflected.

X Y-GEN-LOC
'in Y's X'

----

If I remember correctly Somali drops the marked nominative affixes in some subordinate constructions and both the main arguments are like accusatives.

----

Of course some cases only appear with animete definite nous, Spanish a.

[Creyeditor]

Any more information on the structure of the language or the type of information conveyed by inflection?

Nouns are marked for gender (masculine, feminine, animate, inanimate), number (singular, plural), definiteness (definite, indefinite), and case (ergative, nominative-absolutive-oblique, accusative).
Relatedly, verbs are marked for mood (realis, irrealis, imperative), tense (past, present, future), aspect etc (durative, ingressive, resultative, semelfactive, gnomic, causative, diminutive-iterative), and voice (active, passive, reflexive, reciprocal). Some of the aspect names are probably wrong or at least misleading, since this is an old document.

[Salmoneus]

Maybe genitive attributes have Suffixaufnahme and the possessed noun could be uninflected.

X Y-GEN-LOC
'in Y's X'

Two things happen in Irish that this reminds me of.

Firstly, although the definite article is usually mandatory, it is not permitted when a noun is modified by a genitive (although the genitive noun CAN have a definite article). Thus, an tUachtarán, "the President", but Uachtarán na hÉireann, "the President of Ireland".

Secondly, in double genitives the first genitive is made nominative, unless the modified noun is indefinite and forms a unit of meaning with the first genitive, or the first genitive is definite and the second genitive is indefinite and they form a unit of meaning together, or the first genitive is a verbal noun, in which cases the first genitive is indeed kept in the genitive. Apparently.

[Salmoneus]

Any more information on the structure of the language or the type of information conveyed by inflection?

Nouns are marked for gender (masculine, feminine, animate, inanimate), number (singular, plural), definiteness (definite, indefinite), and case (ergative, nominative-absolutive-oblique, accusative).
Relatedly, verbs are marked for mood (realis, irrealis, imperative), tense (past, present, future), aspect etc (durative, ingressive, resultative, semelfactive, gnomic, causative, diminutive-iterative), and voice (active, passive, reflexive, reciprocal). Some of the aspect names are probably wrong or at least misleading, since this is an old document.

Well, some rules you could have might include:

- topics are unmarked for case (case marking on non-topics combined with voice on the verb can make clear which case the topic must have)

- definite nouns are unmarked for number and gender (if it's already clear from context what you're referring to, you don't need these details to narrow it down)

- nouns modified by an adjective are unmarked for number, definiteness and/or case (clear from agreement on the adjective)

- nouns modifed by a genitive construction are unmarked for definiteness (as in Irish; they're semantically at least partially definite anyway)

- nouns acting as genitives are unmarked for number and gender (these are backgrounded and less information is needed)

- nouns modified by counters are unmarked for number and gender (these are specified by the counter)

- nouns in prepositional phrases are unmarked for case (the preposition already marks them as oblique)

- nouns in relative clauses are unmarked for case (made clear by the structure of the relative clause)

- nouns in conjunction are unmarked for case (shared with other noun in conjunction)

- nouns in conjunction are unmarked for definiteness if the same as the other noun in conjunction

- inanimate nouns not marked for gender when patients, and animate nouns not marked for gender when agents

etc.


Of course, the easiest would simply be to NOT have obligatory marked gender on nouns (which has to be super-rare, I'd have thought?), NOT to have no zero-marked gender, and NOT to have no zero-marked case, since these are all weird, rare features anyway, and if you insist on doing all these things in the same language you can't be too shocked when you end up with a lot of excess markers on things!

[Creyeditor]

Thank you all for all the ideas. I definitely did not think about possession/adjectives that way before. I will also look at determiner-like words. Circumpositions were something that I thought before do not need obligatory case.

A bit on my motivation. This is from Mamambam, which was supposed to be a naturalistic heartlang but it didn't work out because I didn't really like the result. I have a personal distaste against changing or deleting stuff from older conlangs. So, instead I started adding rules to the grammar document. I already decided that some verbal structures are serial-verb-like or compound-like and there will be additional declension classes that have less obligatory marking. Thank you all again for helping make this work.

[Omzinesý]

Maybe genitive attributes have Suffixaufnahme and the possessed noun could be uninflected.

X Y-GEN-LOC
'in Y's X'

Two things happen in Irish that this reminds me of.

Firstly, although the definite article is usually mandatory, it is not permitted when a noun is modified by a genitive (although the genitive noun CAN have a definite article). Thus, an tUachtarán, "the President", but Uachtarán na hÉireann, "the President of Ireland".

Secondly, in double genitives the first genitive is made nominative, unless the modified noun is indefinite and forms a unit of meaning with the first genitive, or the first genitive is definite and the second genitive is indefinite and they form a unit of meaning together, or the first genitive is a verbal noun, in which cases the first genitive is indeed kept in the genitive. Apparently.

Damn
Irish rules can be complex

[Omzinesý]

I'm thinking about the adjectives of my IE lang.

They agree their head in gender number SG.M, SG.F and PL.

A simple solution would be
SG.M -o
SG.F -a
PL -i

But it would be very boring. Any ideas how to make it more interesting?

[Salmoneus]

Maybe genitive attributes have Suffixaufnahme and the possessed noun could be uninflected.

X Y-GEN-LOC
'in Y's X'

Two things happen in Irish that this reminds me of.

Firstly, although the definite article is usually mandatory, it is not permitted when a noun is modified by a genitive (although the genitive noun CAN have a definite article). Thus, an tUachtarán, "the President", but Uachtarán na hÉireann, "the President of Ireland".

Secondly, in double genitives the first genitive is made nominative, unless the modified noun is indefinite and forms a unit of meaning with the first genitive, or the first genitive is definite and the second genitive is indefinite and they form a unit of meaning together, or the first genitive is a verbal noun, in which cases the first genitive is indeed kept in the genitive. Apparently.

Damn
Irish rules can be complex

Ugh.
and that was my simplification of the main rules of a simplified summary I read of a theoretical Irish, not accounting for differences between dialects...

[Kesshin]

What is the name of a noun that refers to a type of something, or the concept of something? I've been referring to it as a "mass noun" but I'm not sure that's right.

For example, from Daas:
Ue le kio li.
Water is blue.

N'dea ez ae'dae.
Humans can/could be good people.

[Omzinesý]

What is the name of a noun that refers to a type of something, or the concept of something? I've been referring to it as a "mass noun" but I'm not sure that's right.

For example, from Daas:
Ue le kio li.
Water is blue.

N'dea ez ae'dae.
Humans can/could be good people.

Which noun in those examples is the one in question?

[Kesshin]

Which noun in those examples is the one in question?

"Water" and "humans'. Sorry about the confusion, I should've made that clear.

[Salmoneus]

Generic nouns.

[Creyeditor]

The first one might be linked to some kind of verbal aspect (e.g. gnomic aspect) more than to the noun in question, i.e. water (as a mass noun) is always blue.
The second one is what some linguists refer to as a kind reading or kind reference. You don't refer to a certain single human, or any human (some might not be able to become good people) and also not to a specific group of humans. You are also not referring to all humans or some humans as a subset of all humans. Instead, you create a kind category human that you can refer to. Oh, and of course a big IINM because this is not my favorite part of semantic theory.

[Visions1]

I'm making suffixes that cause lenition/ellipsis of some preceding consonants

lak + at -> la:t
lan + at -> lã:t

But I wouldn't like to lose all intervocalic instances of those consonants. How could it historically happen that only some suffixes cause the lenition/ellipsis?

lak + ap -> lakap

Why not a coronal vs. peripheral distinction? So:

Code: Select all

lak + at = la.at | lap + at = lahat
lak + ap = lakap | lap + ak = lapak
lat + ap = la.ap | la  + ap = lap
lan + at = lã.ãt | lam + ap = lãʋãp
lam + at = lamat | lan + ap = lanap

In other words:
VCVT, VTVC -> VHVT, VHVC
VNVD, VMVC -> ṼHṼT
Otherwise everything stays the same.

[Visions1]

I'm thinking about the adjectives of my IE lang.

They agree their head in gender number SG.M, SG.F and PL.

A simple solution would be
SG.M -o
SG.F -a
PL -i

But it would be very boring. Any ideas how to make it more interesting?

Add a consonant somewhere(s).

[Salmoneus]

I'm thinking about the adjectives of my IE lang.

They agree their head in gender number SG.M, SG.F and PL.

A simple solution would be
SG.M -o
SG.F -a
PL -i

But it would be very boring. Any ideas how to make it more interesting?

...if it's an "IE-lang", would you not be evolving your adjectives from PIE? If so, "interesting" will obviously be taken care of automatically, surely?

[Solarius]

Does anyone have a good method for comparing letterforms to make sure they aren't too similar? I recently designed a syllabary, but with around 120 characters it's hard to figure out a good method.

Also, I'm a little unsure about how similar is too similar when you start to get to this number of letterforms.