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

[Keenir]

I don't know how the Semitic languages did it--

They had lots and lots of time.

:)

[Khemehekis]

I recall reading a long exercise in evolving a language with triconsonantal roots the way the Afro-Asiatic languages were believed to have evolved. It was on either the ZBB or Fiat Lingua.

[Ahzoh]

I recall reading a long exercise in evolving a language with triconsonantal roots the way the Afro-Asiatic languages were believed to have evolved. It was on either the ZBB or Fiat Lingua.

There is tiramisu's guide on the old ZBB but it is sadly quite incomplete.

[Keenir]

I recall reading a long exercise in evolving a language with triconsonantal roots the way the Afro-Asiatic languages were believed to have evolved. It was on either the ZBB or Fiat Lingua.

back on...I want to say it was the ZBB, but i can't swear to it, I copied down how Semitic-style root systems could arise (from my copy of The Unfolding Of Language, and I and Ahzoh tried our hands at making a conlang from that starting point...I ran aground, but Ahzoh has been successful in it from then to now.

[Ahzoh]

I recall reading a long exercise in evolving a language with triconsonantal roots the way the Afro-Asiatic languages were believed to have evolved. It was on either the ZBB or Fiat Lingua.

back on...I want to say it was the ZBB, but i can't swear to it, I copied down how Semitic-style root systems could arise (from my copy of The Unfolding Of Language

I have read that and The Origin and Development of Nonconcatenative Morphology to glean what I could. But still I remain directionless.

I ran aground, but Ahzoh has been successful in it from then to now.

I guess that's the power of weaponized autism for you. You spend 10 years obsessing over the same topic.

I would hardly say I'm successful in my goal. I still end up creating and relying too much on irreducible morphemes.

[Imralu]

I'm not really sure how to gloss my Tobarese morpheme -iā most effectively. It's a subject marker that, like the other subject markers, can either be "bare" or "filled" (or "completed" or "specified").

---

Other Subject Markers
I'll first show how other subject markers work.

For example, the rank 5 subject marker -ī signifies a "powerless" entity or group of entities. Because inanimate objects are most often in this rank, it can kind of be thought of as meaning "it", as in:

Hóžī
hóž-ī
be.good-R5
It is good.
The powerless entity is good.


Bánī
bán-ī
be.house-R5
It is a house.
The powerless entity is a house.

When the entity needs to be specified, the subject marker can be "filled" by a following base. (Note that the chroneme shifts from the vowel to a following consonant unless that consonant is word-final.)

Hóžibbán
hóž-ī-bán
be.good-R5-be.house
The house is good.
The powerless entity which is a house is good.

Bánihhóž
bán-ī-hóž
be.house-R5-be.good
The good one is a house.
The powerless entity which is good is a house.

In a discourse, a "filled" subject sets what that subject marker refers to until it is reassigned, meaning that the -ī in hóžī below specifically refers to the house.

Dágibbán hóžī.
dág-ī-bán hóž-ī
be.big-R5-be.house be.good-R5
The house is big and itthe house is good.

---

-iā
This contrasts with the subject marker -iā which refers not to a single entity but to an entire proposition.

Dágibbán hóžiā.
dág-ī-bán hóž-iā
be.big-R5-be.house be.good-iā
The house is big and thatthat the house is big is good.
The house is big, whichthat the house is big is good.


The difference between -iā and the other subject markers is not just that its semantic target (propositions) differs from the other subject markers, but also that it can be "filled", "completed", "specified" whatever by a clause, not just a base.

Hóžiaddágibbán.
hóž-iā-dág-ī-bán
be.good-iā-be.big-R5-be.house
The fact that the house is big is good. / It's good that the house is big.

(Hóžiā(-) is how "fortunately" is usually expressed.)

In this latter case, I've been glossing it as -SR because it's equivalent to a subordinator/complementiser "that", but when it is "bare" and refers back anaphorically to a previously mentioned proposition, as in dágibbán hóžiā, the gloss -SR makes fairly little sense.

I'm tentatively thinking of glossing it "PROP" for propositional, which of course needs to be defined (just as R for rank needs to be defined). Can anyone think of anything better?

---

-iō
A further complication is that -iā is factive or realis and it has the counterpart -iō which is non-factive. Essentially, -iā means "the fact that", subordinating or referring to a proposition which remains an assertion of fact, whereas -iō is non-assertive, more like "the idea of".

Hóžioddágibbán.
hóž-iō-dág-ī-bán
be.good-iō-be.big-R5-be.house
The house being big would good. / It is good for the house to be big.

Mō námiallílā.
I am glad that you are here.
I like that you are here.
I like the fact of you being here.

Mō námiollížā.
I would like it if you were here.
I like the idea of you being here.

(The change from l to ž is not a typo, but it's related to how demonstratives/locatives work, which I won't explain right now.)

Any ideas for nice, concise ways to gloss -iā and -iō?

I guess that's the power of weaponized autism for you.

Industrial-strength, steel-reinforced, twice-refined, double-filtered, weapons-grade autism with three blades for a closer cut.

[Creyeditor]

How about FACT and NFACT? Factive is what people often use to refer to embedded clauses independent of their (c)overtness., e.g. "I know it/that you sleep" is factive.

[Imralu]

How about FACT and NFACT? Factive is what people often use to refer to embedded clauses independent of their (c)overtness., e.g. "I know it/that you sleep" is factive.

Thanks! I had thought of that and then looked it up and "factive" often seems to be used for an evidential type (close-ish) or also verbs that express producing a result (e.g. make, build, kill) (not close), and as a synonym for "result" in case grammar (very far).

I have just found "factivizer" used to describe the suffix -eč in Cholón, glossed as -FAC, and it is a subordinating morpheme. I guess because FACT can be used for so many things, it needs to be clearly defined anyway, but FACT and NFACT or FAC and NFAC are at least pretty clear labels that are easy to remember and helpful to conceptualise and understand what the morphemes do. Thank you!

[Salmoneus]

I'd probably just go with something like RPA (realis propositional anaphora) and IPA (irrealis propositional anaphora), since that seems to be what they mean. Since you're abbreviating it for the gloss anyway, you don't really need to coopt a catching two-syllable term for it, I don't think.

[Khemehekis]

@Ahzoh: How would Vrkhazhian handle the "not because" distinction in this translation exercise?

[Ahzoh]

@Ahzoh: How would Vrkhazhian handle the "not because" distinction in this translation exercise?

probably some illative plus an adversative or a double use of an adversative. Or a negative marker plus an adversative.

[LinguoFranco]

So, I am toying with adding ATR harmony into a conlang, but doing it in a way that's different from most natlang examples.

I heard that in such a system, /e/ can contrast with /a/ instead of instead of /ɛ/. Why not take this farther and have /u/ contrast with /ɔ/ instead of /ʊ/?

So, /e u/ are +ATR, /a ɔ/ are -ATR, and /i/ is an opaque, neutral vowel.

This isn't anything too weird, is it?

[Khemehekis]

@Ahzoh: How would Vrkhazhian handle the "not because" distinction in this translation exercise?

probably some illative plus an adversative or a double use of an adversative. Or a negative marker plus an adversative.

Cool.

[Creyeditor]

So, I am toying with adding ATR harmony into a conlang, but doing it in a way that's different from most natlang examples.

I heard that in such a system, /e/ can contrast with /a/ instead of instead of /ɛ/. Why not take this farther and have /u/ contrast with /ɔ/ instead of /ʊ/?

So, /e u/ are +ATR, /a ɔ/ are -ATR, and /i/ is an opaque, neutral vowel.

This isn't anything too weird, is it?

Well, it combines a number of rare things, making it ... Well, the pun doesn't work in English (German 'selten' rare -> 'seltsam' weird).
Re-pairing is the term that I have seen in use for the phenomenon you describe. It usually occurs of there is a gap in the vowel inventory. Languages that don't have a [+ATR] low vowel re-pair it with a [+ATR] mid vowel. In principle that could also apply to high and mid vowels. This is rarer because gaps in the high and mid vowels are rarer in ATR vowel harmony systems. This also applies to your neutral vowel /i/ which I would expect to not have a [-ATR] counterpart if it is a neutral vowel.
If we take this into consideration, we would arrive at the following vowel inventory for your conlang:

/i u/
/e ɔ/
/a/

I am not saying this is implausible but I would expect phonologist to analyze it differently. If the two mid vowels do not form a natural class for any phonological process, I would expect /e/ to be transcribed as a low vowel or /ɔ/ to be transcribed as a high vowel. Of course, this might be different if you add the missing mid vowels. This would make the vowel harmony system more complicated since it would involve either covert contrast or neutralizations.

[Ahzoh]

Can clitics take secondary stress? I know they are characterized by lacking prosodic independence, which usually means unaccented/unstressed. But I hear nothing about whether they can or can't take secondary or tertiary stress as a feature of a word's demand for rhythmic feet.

[Creyeditor]

Time for another diabolc duality declaration. There are (at least) two different clitic concepts in theoretical linguistics: adphrasal affixes and 'outer' affixes.

• Adphrasal affixes are clitics in the sense that they attach to a phrase and not to a word. This means that (a) they do not always occur next to the head that they modify and (b) they occur next to a variety of different word classes. These are clitics in the sense of Martin Haspelmath.
• Outer affixes are affixes that are prosodically least integrated in a given language. This means that they undergo less phonological processes compared to other affixes in the same language and usually occur at the left or right edge of a word. Due to the tendency of affix order to correlate with semantics, these often come from the same set of grammatical categories. Haspelmath refers to prosodic integration as welding, IIRCA.

Note, however, that these two do not need to coincide. There are outer affixes that only occur on nouns (or verbs) and there are adphrasal affixes that are tightly integrated into the prosodic word. My hunch is that the two do not even show a weak correlation.
Adphrasal affixes have no reason to not take secondary stress. And even outer affixes can take secondary stress as long as there are other phonological processes that they do not undergo compared to other affixes.
Sorry for the long-ish answer, it's just a question that I have been thinking about.

[Ahzoh]

Time for another diabolc duality declaration. There are (at least) two different clitic concepts in theoretical linguistics: adphrasal affixes and 'outer' affixes.

• Adphrasal affixes are clitics in the sense that they attach to a phrase and not to a word. This means that (a) they do not always occur next to the head that they modify and (b) they occur next to a variety of different word classes. These are clitics in the sense of Martin Haspelmath.
• Outer affixes are affixes that are prosodically least integrated in a given language. This means that they undergo less phonological processes compared to other affixes in the same language and usually occur at the left or right edge of a word. Due to the tendency of affix order to correlate with semantics, these often come from the same set of grammatical categories. Haspelmath refers to prosodic integration as welding, IIRCA.

Note, however, that these two do not need to coincide. There are outer affixes that only occur on nouns (or verbs) and there are adphrasal affixes that are tightly integrated into the prosodic word. My hunch is that the two do not even show a weak correlation.
Adphrasal affixes have no reason to not take secondary stress. And even outer affixes can take secondary stress as long as there are other phonological processes that they do not undergo compared to other affixes.
Sorry for the long-ish answer, it's just a question that I have been thinking about.

I have a class of words that behave like prefixes attached to the first words of an entire clause and modify the entire clause (they behave like conjunctions)

ur(ru)= "and, nor, also (adverb)"
ek(ke)= "or, else/otherwise (adverb)"
kiz(za)= "but, except"

(āti) naparraḫti "we will interrogate"
(tutti) naparraḫtisi "we will not interrogate"
āti=ki naparraḫti? "will we interrogate?"

ur=naparraḫti "and we will interrogate"
ur=āti naparraḫti "and we will interrogate"
ur=tutti naparraḫtisi "and we will not interrogate"
ur=ki naparraḫti? "And will we interrogate?"
ur=āti=ki naparraḫti? "and will we interrogate?"

They're unstressed when attached to even-syllabed words, but secondarily stressed on odd-syllabled words (opposite situation when the negative suffix is attached)

[Creyeditor]

That sounds sensible and very naturalistic.

[Knox Adjacent]

Can anyone guess what distinction I might have intended when I added both a dubitative and an uncertain mood to my list to later buff out? Cuz I can't.

[Creyeditor]

Maybe a difference in perdspective? "I (subjectively) doubt that ..." vs. "It is (objectively) uncertain if". This could be some kind of evidentiality-related distinction then.