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

[Shemtov]

Is it naturalistic for a language to have a separate piece of morphology added on transitive verbs for causatives, that is without a specified indirect object, and another for ditransitivation, that is, with a specified indirect object?

[sangi39]

Hello. While developing my protolang's phonology, I've been wondering: are proto-languages likely to develop allophony from the very beginning or not?
For instance, my proto-lang's phonetic inventory is thus:
/m n ɲ ŋ/
/p b t̪ d̪ c ɟ k g/
/ɸ θ ɬ s̠ ɹ̠̊˔ ç h/
/l/
/i iː ʏ ʏː u uː/
/e eː ø øː o oː/
/æ æː ɐ ɐː/
But there's also a bunch of allophony:
/ŋ/- = -[ŋ.g]-.
/b d ɟ g/- = -[β̞ ð̞ j ɰ]-.
/ɸ θ ɬ s̠ ɹ̠̊˔ ç/- = -[β ð ɮ z̠ ɹ̠˔ ʝ]-.
/h/- = -[x]-.
(-)/(')e/- = -[ə].
Is it naturalistic? (I mean, to have allophones from the very beginning.) Are allophones more likely to be developed by language descendants rather than by the proto-language itself?

Strictly speaking, a "proto-language" is either a) an actual language spoken by actual people some time in the past that is the ancestor of some other language, or b) the reconstructed version of that language based on comparison of descendant languages. So, for example, Proto-Indo-European was most probably a real language spoken by real people at some point in the past, but the PIE we see today are just best guesses as to what it might look like.

That said, since proto-languages are just languages, it doesn't seem likely that they wouldn't have had some sort of allophony going on, so "to have allophones from the very beginning" makes perfect sense, and could actually lead to more well-developed descendant languages.

[eldin raigmore]

Is it naturalistic for a language to have a separate piece of morphology added on transitive verbs for causatives, that is without a specified indirect object, and another for ditransitivation, that is, with a specified indirect object?

I am practically certain that which you propose is both naturalistic and realistic.
But I gather you really want an answer citing a couple of supporting references.
I don’t have such references at the front of my brain right now. I suspect it would be just as easy or difficult for you to look them up as for me.

—————

If you want to add an argument to a monotransitive clause, it would be, I think, either;
* a direct Instigator (made the causee agent-of-effect Performer do the thing), or;
* an indirect Instigator (persuaded the causee agent-of-effect Performer to do the thing), or;
* a Beneficiary or Maleficiary (entity who the speaker wants to say was/is/will be saliently affected).

Why would it be a Recipient?
Though maybe your conlang’s “ditransitivization morphology” might be kinda vague-ish on the difference between a Recipient and a Beneficiary/Maleficiary?

Or maybe you have one or a few real-life examples?
“The quarterback threw the football”
“The quarterback threw-DITRANS the football to the wide receiver”
Or something like that?

[Dormouse559]

Hello. While developing my protolang's phonology, I've been wondering: are proto-languages likely to develop allophony from the very beginning or not?
[…]

[…]

That said, since proto-languages are just languages, it doesn't seem likely that they wouldn't have had some sort of allophony going on, so "to have allophones from the very beginning" makes perfect sense, and could actually lead to more well-developed descendant languages.

It also should be noted that allophony is sound change. A sound change that isn't (yet) phonemic is called allophony. You can also get allophony from some kinds of phoneme mergers, so again, more sound change. Point being that asking whether a proto-language should have allophony is like asking whether it should undergo sound change.

[Shemtov]

Is it naturalistic for a language to have a separate piece of morphology added on transitive verbs for causatives, that is without a specified indirect object, and another for ditransitivation, that is, with a specified indirect object?

I am practically certain that which you propose is both naturalistic and realistic.
But I gather you really want an answer citing a couple of supporting references.
I don’t have such references at the front of my brain right now. I suspect it would be just as easy or difficult for you to look them up as for me.

—————

If you want to add an argument to a monotransitive clause, it would be, I think, either;
* a direct Instigator (made the causee agent-of-effect Performer do the thing), or;
* an indirect Instigator (persuaded the causee agent-of-effect Performer to do the thing), or;
* a Beneficiary or Maleficiary (entity who the speaker wants to say was/is/will be saliently affected).

Why would it be a Recipient?
Though maybe your conlang’s “ditransitivization morphology” might be kinda vague-ish on the difference between a Recipient and a Beneficiary/Maleficiary?

Or maybe you have one or a few real-life examples?
“The quarterback threw the football”
“The quarterback threw-DITRANS the football to the wide receiver”
Or something like that?

What I'm thinking of doing is have one piece of morphology that makes a transitive verb a causative, that is, if possible, the outcome is still transitive as the object is made to do an intransitive action: CAUSE-hit "To make go wild" CAUSE-kill "To make bloodthirsty". The other morpheme is used to mean "Cause to kill" or is used in the case of transitives that take an object automatically. Eat, which always takes a dummy object, CAUSE2-eat DO, "To feed" but the first causative cannot be applied to "Eat" or anything that doesn't have an intransitive equivalent that can be caused. I guess the first morpheme is a detransitiver that then retransitivizes. Or could I make it that the first causative only that way works for verbs that can be detransitized. then retransitivized, but it a simple ditransitivizer for verbs like "eat", and another morpheme that acts like a simple ditransitivizer for verbs that can be detransitized. then retransitivized. Which is more natural? Is either really "naturalistic"? I don't want exact examples- I trust that if someone says they've encountered it before, or there's a similar idea in X language, that probably has an ANADEW version of what I'm proposing.

[Omzinesý]

Is it naturalistic for a language to have a separate piece of morphology added on transitive verbs for causatives, that is without a specified indirect object, and another for ditransitivation, that is, with a specified indirect object?

I am practically certain that which you propose is both naturalistic and realistic.
But I gather you really want an answer citing a couple of supporting references.
I don’t have such references at the front of my brain right now. I suspect it would be just as easy or difficult for you to look them up as for me.

—————

If you want to add an argument to a monotransitive clause, it would be, I think, either;
* a direct Instigator (made the causee agent-of-effect Performer do the thing), or;
* an indirect Instigator (persuaded the causee agent-of-effect Performer to do the thing), or;
* a Beneficiary or Maleficiary (entity who the speaker wants to say was/is/will be saliently affected).

Why would it be a Recipient?
Though maybe your conlang’s “ditransitivization morphology” might be kinda vague-ish on the difference between a Recipient and a Beneficiary/Maleficiary?

Or maybe you have one or a few real-life examples?
“The quarterback threw the football”
“The quarterback threw-DITRANS the football to the wide receiver”
Or something like that?

What I'm thinking of doing is have one piece of morphology that makes a transitive verb a causative, that is, if possible, the outcome is still transitive as the object is made to do an intransitive action: CAUSE-hit "To make go wild" CAUSE-kill "To make bloodthirsty". The other morpheme is used to mean "Cause to kill" or is used in the case of transitives that take an object automatically. Eat, which always takes a dummy object, CAUSE2-eat DO, "To feed" but the first causative cannot be applied to "Eat" or anything that doesn't have an intransitive equivalent that can be caused. I guess the first morpheme is a detransitiver that then retransitivizes. Or could I make it that the first causative only that way works for verbs that can be detransitized. then retransitivized, but it a simple ditransitivizer for verbs like "eat", and another morpheme that acts like a simple ditransitivizer for verbs that can be detransitized. then retransitivized. Which is more natural? Is either really "naturalistic"? I don't want exact examples- I trust that if someone says they've encountered it before, or there's a similar idea in X language, that probably has an ANADEW version of what I'm proposing.

I'm not sure what you are asking.
In causes like (1) "X made Y kill Z", C is the causer, Y is the causee, and Z is the object.
Are you asking if there are languages where (2) "X made Y kill" without an explicit object Z have a different causative morpheme/auxiliary from (1) which has three arguments?

[Salmoneus]

I think a confusion may be the choice of examples - the derivation from "I hit you" to "I make you wild" seems almost random, not a regular inflection. [if "I hit you" > "I make you wild", what does "I punch you" or "I kick you" inflect to?]

However, you seem to have answered your own question regarding naturalism.


Let's take an easier verb, like "wash" or "swim". In English, these are both homophonous with their univalent partner, so let's mark that explicitly:
I swam-BV the river; I wash-UV the cat
vs
I swam-UV; I wash-UV

Now, some languages will have different verbs for "swim-BV" and "swim-UV"; others, like English, will rely on syntactic context (like, eg, prohibiting object-dropping for ambiguous verbs - you would very, very rarely in English say "I washed!" when you meant you washed something else, you'd either use a dummy pronoun ("I washed it!" or "I washed something!") or you'd explicitly detransitivise the verb either by making a stative with a participle ("I've done (some/the) washing!") or by making an accomplishment with a phrasal verb ("I've washed up!")).

But some languages will have regular derivational processes to turn native bivalents into univalents - antipassives, essentially. There are also of course passives, which do the same but alter the argument structure. Passives make the second argument of the bivalent into the first (and only) argument of the univalent, while antipassives make the first argument of the bivalent into the first (and only) argument of the univalent.


Other languages have regular derivational processes to INCREASE valency. One type of such a process is a causative. These turn univalents into bivalents, and bivalents into trivalents by making the first argument of the bivalent or univalent into the second argument of the trivalent or bivalent:
I sleep > you make me sleep
I eat cabbage > you make me eat cabbage
I swam the river > you made me swim the river


Your question appears to be: can a language that has both an antipassive and a causative apply both to the same verb at the same time?

To which the answer is: of course, why would that not be possible?

Consider English:
I swam the river (bivalent) > I swam (zero-marked antipassive univalent) > you made me swim (causative of an antipassive)
we can also do the same with the passive, though this is sometimes less elegant in English:
I swam the river > the river was swum (passive) > you made the river be swum (causative of a passive)

Now, I would say, not all languages with both an antipassive and a causative will allow them to be combined. The construction will often be rare, hence little-used, hence may go extinct. Also, many languages limit the number of "slots", either by position around the verb (eg only one prefix) or by category (eg only one valency-altering affix), and this may prohibit these more complicated combinations. I don't know what percentage of verbs with both allow them to be combined. But surely there must be some, because there's no absolute reason why that combination of morphemes should be impossible.

I suppose a second question is then: could a language have a fused affix for the causatives of antipassives?

Again: of course, any morphemes can become fused over time.


[one difference this language would have to have from English would be that verbs would have to be 'natively' either bivalent or univalent, so that you'd know whether, for instance, "you make me eat" requires only the causative of a native univalent, or the causative of the antipassive of a native bivalent.]

[Shemtov]

I gave up on the idea, as I realized that even if it was naturalistic, it would make the language clunky.

[Ahzoh]

I don't know how my language can do fractions above 1/10 (e.g. 1/11 or 1/27) since my fractions are entirely nouns.

my numbers also behave like this:
20 = 2x10
200 = 2x100
2000 = 2x1000
2222 = 2 2x10 2x100 2x1000

This only works since the cardinals and ordinals are adjectives, not nouns.

[holbuzvala]

@Ahzoh

I mean, in English we can just say 'three four-hundredths' for 3/400. So if you wanted 1/27, you could say 'one twenty-and-seventh' or simply 'one twenty-seventh'.

However, I'm not entirely sure I've understood your question. What do you mean that your fractions are 'entirely nouns'?

[eldin raigmore]

In English “half” and “quarter” can be nouns or verbs or adjectives or adverbs. (Both ad-verbal and ad-adjectival adverbs.)

—————

I don’t see why the names of unit fractions should be closely related to ordinal numbers.
I’d be willing to bet there’s a natlang in which there is no such relationship; but @tm none comes to mind.

[Reyzadren]

^There is no such relationship in that other natlang that I speak (you win!). It just uses per, ie: number per number, or like percent.

imo using ordinals for fractions is just silly

[eldin raigmore]

^There is no such relationship in that other natlang that I speak (you win!). It just uses per, ie: number per number, or like percent.

Good! And Thanks! And would you care to PM me the name of that other natlang?

imo using ordinals for fractions is just silly

I’m sure it comes from “the third part” and “the fifth part” and so on.
Like the barker crying “One nickel! Half a dime! The twentieth part of a dollar!”

—————

But I’m working on the numeral system of a conlang, in which ordinal numbers are the roots, and cardinal numbers are derived or inflected from ordinals. (So I definitely don’t want unit fractions to be homophones and homographs for ordinals!)

And for fractions, I want 1/n, (n-1)/n, and (n+1)/n, to be a single word each, for n running from two to twelve (its base), or by twelves from a dozen to a gross, or the cube of twelve. (Not sure I want 1/2 and (2-1)/2 to have both forms.)

And I want them, and also the counting numbers, to have verbal forms meaning in effect “to multiply by”.

And I wonder how much of that is ANADEW?

[Ahzoh]

I mean, in English we can just say 'three four-hundredths' for 3/400. So if you wanted 1/27, you could say 'one twenty-and-seventh' or simply 'one twenty-seventh'.

No, because one twenty seventh or "seventh ten two (one)" would be interpreted as 20/7.

However, I'm not entirely sure I've understood your question. What do you mean that your fractions are 'entirely nouns'?

They behave like nouns, decline like nouns, are modified by adjectives like nouns and are posssessed by other nouns like nouns.

[Micamo]

You should do a conlang where the word for a fraction is just its index in a breadth-first traversal of the Stern-Brocot tree.

[clawgrip]

I mean, in English we can just say 'three four-hundredths' for 3/400. So if you wanted 1/27, you could say 'one twenty-and-seventh' or simply 'one twenty-seventh'.

No, because one twenty seventh or "seventh ten two (one)" would be interpreted as 20/7.

However, I'm not entirely sure I've understood your question. What do you mean that your fractions are 'entirely nouns'?

They behave like nouns, decline like nouns, are modified by adjectives like nouns and are posssessed by other nouns like nouns.

I would like to help, but I'm a little confused here. Can you provide some actual examples in your conlang?

Seven ten two = 27?
Seven ten two one = 20/7?
If I see how your grammar works, maybe I can help you come up with a workaround.

[Nloki]

Hi. I've come up with three brief questions I'd be very thankful for someone to answer.
•Are Nominative-Accusative languages likely to evolve into Ergative-Absolutive or viceversa?
Are both types of alignment systems more common to evolve into split Ergative?
•Is a conlang evolved from an a priori conlang also another a priori conlang? Or rather an a posteriori [i/] conlang (since its based in another language (construced or not)).
•Would it be a good idea to create new daughter languages within Tolkien's elvish languages, for instance, evolving Quenya to obtain a daughter elvish Valinórean language or enlarging Vanyarin Quendya? (Of course not pretending the resulting language to be completely yours).

[Omzinesý]

Hi. I've come up with three brief questions I'd be very thankful for someone to answer.
•Are Nominative-Accusative languages likely to evolve into Ergative-Absolutive or viceversa?
Are both types of alignment systems more common to evolve into split Ergative?
•Is a conlang evolved from an a priori conlang also another a priori conlang? Or rather an a posteriori [i/] conlang (since its based in another language (construced or not)).
•Would it be a good idea to create new daughter languages within Tolkien's elvish languages, for instance, evolving Quenya to obtain a daughter elvish Valinórean language or enlarging Vanyarin Quendya? (Of course not pretending the resulting language to be completely yours).

Short answers

* No statistical evidence. Both directions are possible. Passive-like constructions can become abs-erg and antipassive-like constructions mom-acc.
My understanding is that the ergative is often generalized as the case of agentive intransitives as well. That can too lead to real nom-acc.
* Those terms aren't too scientific. My understanding is that a-posteriori langs are based on natural languages.
* That's up to you! What do you like. One shouldn't care on others opinions on such questions.

[LinguistCat]

I mean, in English we can just say 'three four-hundredths' for 3/400. So if you wanted 1/27, you could say 'one twenty-and-seventh' or simply 'one twenty-seventh'.

No, because one twenty seventh or "seventh ten two (one)" would be interpreted as 20/7.

However, I'm not entirely sure I've understood your question. What do you mean that your fractions are 'entirely nouns'?

They behave like nouns, decline like nouns, are modified by adjectives like nouns and are posssessed by other nouns like nouns.

I would like to help, but I'm a little confused here. Can you provide some actual examples in your conlang?

Seven ten two = 27?
Seven ten two one = 20/7?
If I see how your grammar works, maybe I can help you come up with a workaround.

There's always seven of ten two or similar constructions (ten two, of, seven? depending on how things get ordered in possessive phrases)

[wintiver]

Can a syllable timed language have long and short vowels? If so this would mean you have one length for the long and one for short vowels, which would approximate the way stress-time languages function right? Or is having fixed long and fixed short syllable length the middle-ground where moras live?

I did some reading but I can't get clear on it.

Thank you for your time.