Questions on grammatical number & agreement

If you're new to these arts, this is the place to ask "stupid" questions and get directions!
Post Reply
Doranwen
rupestrian
rupestrian
Posts: 19
Joined: 18 Jul 2016 02:49

Questions on grammatical number & agreement

Post by Doranwen »

As promised, I'm back with more questions.

So I think I understand the basic singular-plural distinction pretty well at this point. :) Dual makes sense, for all it's somewhat rare. (Same for trial, quadral, paucal.) I sort of understand singulative-collective as well. What I've got some questions about is regarding a mixed system (and trying to sort out how I want to do number in my conlang).

In a mixed system, do you just have two groups of nouns, and you have to memorize which one belongs to which type? From what I saw of the Wikipedia article, it looked like you basically had to just know which was which, and also that some nouns could take the singulative, become a noun in their own right, and have a plural applied to it. (Most of the complications on there seemed to be due to Welsh just being plain complicated in general, lol.)

If a word like "grass" is collective, then the singulative would mean "a blade of grass" or "a piece of grass", right? If so, how would one then go about expressing "blades of grass"? (The sort of thing like above? Singulative -> separate noun -> apply plural?)

For that matter, I found quantifiers very challenging, to think of another way to express them than exactly how English does it. I'd rather not just copy English, but pick features that are the right amount of complexity to suit me and that are interesting. I couldn't figure out how you'd say things like "several cows" or "a few blades of grass" or "some rice" besides just plunking a quantifier word down next to the noun. (Well, there was the partitive case, but that one confused me a lot, lol. I haven't put it in my list of cases yet because of that.)

As for what number system I want to use, I know I like the idea of being able to form a singulative, and I like singular/plural for sure. For that matter, I like dual too. (I'm inclined to include a vestigial trial, used only on words referring to the Trinity, as part of my religious usage of this language--yeah, I know no natural language has that, lol, but I'm not too concerned--but I'm not sure how one would go about adding in a feature that is no longer productive, and that is used only on specific words.)

If I use singular/plural, but also have a way to mark singulative, does that make it a mixed system? Is this too complex? I really like the idea of it. Or perhaps there's another way to get rid of the need to use specific unit nouns in order to count mass nouns: "grain" (of rice), "blade" (of grass), etc. It gets so muddled in my brain that I can't sort out what I'm trying to do, or whether there's another (better) way to do something that I'm just not thinking of... This is one of the areas I got things overly complicated last time I tried this, mostly because I didn't fully understand the systems and definitely didn't have any idea about how to properly handle quantifiers.

Also, if I'm trying to go for fusional rather than agglutinating, that would mean I'd need to develop plural forms of each case, so I don't have to have two separate affixes? (No wonder I don't want to make too many cases; the number of inflections would soon be absolutely insane.) And I would assume any other grammatical number forms would also need some sort of marking...

One thing I do know is I liked the idea of leaving a noun singular if a numeral is applied to it (why mark if you don't need to?). Like "dogs" + "three" becomes "three dog" or "dog three". As far as that word order, I haven't decided which way it will go; most of the features I have settled on so far correspond with adjective-noun order in natural languages as far as I could see on WALS, but there are so few for a couple of the features that it's hard to draw any definitive conclusions. I would assume that number should agree in case with the noun it's modifying, in order to promote the free word order I want, though, am I correct? But now that I think of it, if I'm using inflections rather than separate to keep things fusional, then it would make sense to mark "three dogs" as it won't take up any extra space, since I have to mark the case for dogs, yes?

Though with absolutive, if I'm not marking that, then the plural would be extra marking that I could then have left out if the numeral is there--maybe that's a dialect thing, even, if I want to create a sort of lower register with variant grammar. The higher registers requiring the plural to always be marked, the lower register using singular whenever possible, or something like that? Could even be a difference between spoken/written.
User avatar
Inkcube-Revolver
cuneiform
cuneiform
Posts: 126
Joined: 05 Nov 2015 23:20
Location: Miami, FL

Re: Questions on grammatical number & agreement

Post by Inkcube-Revolver »

Hi there! I'm not too well-versed on the topic, but since you're going with something fusional in this case rather than agglutinating, I'd make a distinction between mass nouns that are countable and uncountable. It just makes it easier as far as I can tell to clear that with at least some words. Also, establishing whether your language has articles or not might be helpful, because although it might be grammatically incorrect in almost every language with them to say "a water," you can refer to "a (drop) of water" by using a genitive for water and possibly omitting the word for "drop" or another word accompanying it that implies singularity, but only if that's what you would go for. Latin does away with that with the word aquae literally meaning "of water" in its singular genitive of aqua, doing so lacking articles.

So, in making up some gibberish (a thing I love doing) as a demonstration, we have:

opi "water," and
id opi
(paucal article)+(noun) "some water."

You can have it be te pui opne (indefinite article)+(noun.NOM.)+(noun.GEN.Sing.) "a drop of water" just like that. "Drops of water" would be pluralizing the subject of the sentence, so puli opne would be "drops of water," the word for water still being singular but "drops" being pluralized. Another option is to have it be simply opne with singular-plural distinctions.
If you'd want to go even further, have a special ending or affix that only deals with mass nouns, so if you deem opne too vague or general, then put something like -tka to make it opitka, literally "one (something) of water."

sashitka "a blade of grass"
hashitka a spark (lit. "a something of speed")
yukotka a beam, ray of light (lit. "a something of light")

You could in theory even do it in reverse:
sashi "a blade of grass" > sashio "grass"
hashi "a spark" > hashio "lightning"
opi "a drop of water" > opia "water"
yuko "a ray of light" > yukio "light"

I'm sure a natural language somewhere out there does it like that, but no name comes to mind.

I've dabbled in some mass nouns like this, trying to make sense with "a blade of grass" as well in one of my own conlangs, but hopefully this helps! You can even use the gibbery words I used if you want to help with their distinctions made, and good luck on your conlang! It'd be really interesting to see what you end up using, especially with that vestigial trial number system, that sounds fascinating and the possibilities are nearly endless.
I like my languages how I like my women: grammatically complex with various moods and tenses, a thin line between nouns and verbs, and dozens upon dozens of possible conjugations for every single verb.
Keenir
mayan
mayan
Posts: 2399
Joined: 22 May 2012 03:05

Re: Questions on grammatical number & agreement

Post by Keenir »

Doranwen wrote:As promised, I'm back with more questions.

As for what number system I want to use, I know I like the idea of being able to form a singulative, and I like singular/plural for sure. For that matter, I like dual too. (I'm inclined to include a vestigial trial, used only on words referring to the Trinity, as part of my religious usage of this language--yeah, I know no natural language has that, lol, but I'm not too concerned--but I'm not sure how one would go about adding in a feature that is no longer productive, and that is used only on specific words.)
you just answered your own question - give it a trial (paucal, I thought was serviceable as a 3), but limit it to only a few words.

the fact it only shows up on words relating to the Trinity, actually offers a reason why: when it stopped being productive, it survived in words pertaining to the Trinity; but in other words, substitute words - which were still productive - took the place of the trial/paucal.
At work on Apaan: viewtopic.php?f=6&t=4799
User avatar
Lambuzhao
korean
korean
Posts: 5405
Joined: 13 May 2012 02:57

Re: Questions on grammatical number & agreement

Post by Lambuzhao »

Keenir wrote:
Doranwen wrote:As promised, I'm back with more questions.

As for what number system I want to use, I know I like the idea of being able to form a singulative, and I like singular/plural for sure. For that matter, I like dual too. (I'm inclined to include a vestigial trial, used only on words referring to the Trinity, as part of my religious usage of this language--yeah, I know no natural language has that, lol, but I'm not too concerned--but I'm not sure how one would go about adding in a feature that is no longer productive, and that is used only on specific words.)
you just answered your own question - give it a trial (paucal, I thought was serviceable as a 3), but limit it to only a few words.

the fact it only shows up on words relating to the Trinity, actually offers a reason why: when it stopped being productive, it survived in words pertaining to the Trinity; but in other words, substitute words - which were still productive - took the place of the trial/paucal.

A three-horse (or other draft animal) driven wagon/chariot/wagon was called a troika in :rus:. Reference to the group of horses might be another, non-religious (albeit only one more) use of a Triune/trial number.

In Ancient Rome, one type of government was the Triumvirate. If you had a similar government, the body of three rulers/archons could be referred to in the Triune/trial number.

On the other hand, if you have a government based on the principle of trias politica (separation of powers into three branches), then you could also refer to that government in the Triune/trial number.


If your :con: world has a region dominated by three cities, a tripolis (Cf. :ell: Τρίπολις Tripoli), that could be referred to in a Triune/trial number. My :con: world Tirga had a political alliance of thirteen independent Castaldies/Nomes called collectively the Triscaidecapolis.

Yet another option is an introductory curriculum at a medieval university involving the study of grammar, rhetoric, and logic. This was and is referred to as the Trivium. If you had an educational system involving a foundation of three important areas, you could refer to it in the Triune/trial number.

Clover and Poison Ivy are both notorious three-leaved plants. If you were to have these plants, or some other plant known by its three leaves (or three roots, or three cotyledons, or triple-fruits, or three seeded-fruit), it could also be described using a Triune/trial number.

A famous river in your :con: world that had three branches could be referred to in the Triune/trial number. Another possibility is if you had three great lakes together (Cf. Lake Huron, Lake Michigan, Lake Superior).

A group of three volcanoes, geysers or mountains (the very hanzi for mountain 山 has three peaks) could likewise be referred to in the Triune/trial number. The Tharsis Montes of Mars ( Arsia Mons, Pavonis Mons, and Ascraeus Mons) are a great non-Terran example of this.




Though these are, admittedly, a handful of examples, the list could go on and on and on.
:wat:
Doranwen
rupestrian
rupestrian
Posts: 19
Joined: 18 Jul 2016 02:49

Re: Questions on grammatical number & agreement

Post by Doranwen »

Inkcube-Revolver wrote:Hi there! I'm not too well-versed on the topic, but since you're going with something fusional in this case rather than agglutinating, I'd make a distinction between mass nouns that are countable and uncountable. It just makes it easier as far as I can tell to clear that with at least some words.
Ahh, good point.
Inkcube-Revolver wrote: Also, establishing whether your language has articles or not might be helpful, because although it might be grammatically incorrect in almost every language with them to say "a water," you can refer to "a (drop) of water" by using a genitive for water and possibly omitting the word for "drop" or another word accompanying it that implies singularity, but only if that's what you would go for. Latin does away with that with the word aquae literally meaning "of water" in its singular genitive of aqua, doing so lacking articles.
Oh yes, I have decided that: no articles, lol. I am using a few cases (trying to limit to a reasonable number and finding ways to express some things with other means than separate individual cases, as you can see in my first thread. One of my aims is free word order, as much as possible, so I'd like to find ways to eliminate articles entirely and still be able to distinguish the different types of number. (Though I suppose if it doesn't work otherwise, I could always use articles and decline them right along with the nouns so there's agreement in case they can't be next to each other for some reason. But I'd still like to see if there's a way to eliminate the articles first. I'm not too fond of either articles or prepositions, the latter being one of the reasons why I have cases.)

The challenge I can see is that, if I'm having cases, does that mean I need separate singular/plural versions of cases, if I'm trying to keep it fusional? And if so, then how do I handle dual and the trial I want to incorporate? (And then again, what do I do about the singulative idea--deriving a single item from a mass noun?) Putting it all together is the real tricky part. I honestly don't mind big charts of inflections and declensions (I think that sounds like FUN), I'm just trying to figure out what's reasonable without going overboard, if there are options I've overlooked, and which charts I would need.
Inkcube-Revolver wrote:It'd be really interesting to see what you end up using, especially with that vestigial trial number system, that sounds fascinating and the possibilities are nearly endless.
Well, I'm sure I'll be back a lot as I sort stuff out, and once I have a better idea of all the things I need to take into account, I'll start actually developing the charts and suffixes and pulling some of the vocabulary I'd created for the old version of this conlang and updating it as necessary to use in this one, so I can start using examples from the language itself.
Keenir wrote:you just answered your own question - give it a trial (paucal, I thought was serviceable as a 3), but limit it to only a few words.

the fact it only shows up on words relating to the Trinity, actually offers a reason why: when it stopped being productive, it survived in words pertaining to the Trinity; but in other words, substitute words - which were still productive - took the place of the trial/paucal.
Numerals still being productive, I'd imagine that "three trees" rather than "trees.TRI" (or whatever the glossing would be for that) would take the place nicely.
Lambuzhao wrote:If your :con: world has a region dominated by three cities, a tripolis (Cf. :ell: Τρίπολις Tripoli), that could be referred to in a Triune/trial number. My :con: world Tirga had a political alliance of thirteen independent Castaldies/Nomes called collectively the Triscaidecapolis.

Yet another option is an introductory curriculum at a medieval university involving the study of grammar, rhetoric, and logic. This was and is referred to as the Trivium. If you had an educational system involving a foundation of three important areas, you could refer to it in the Triune/trial number.
Ironically, despite wanting to include a vestigial number, this is more of a heartlang than anything; I'm not particularly into conworlding and haven't done much in that regard. I mean, I've sort of set bits of culture that I would include in it, but not a full-fledged world or even a nation or anything--it's just not where my interest lies. But I'll toy with the idea of having the trial be used in set names for things in our word; I mean, there's the "Three Sisters" of Native American gardening (corn, beans, squash), and the Trivium is a great idea. But I'm starting to sort of see how the productivity/non-productivity works with this--if it's only part of set names, then it's no longer productive, but if one can use it for anything, whether it's been said before or not, then it's still productive and an active part of the language. Am I right?
Last edited by Doranwen on 19 Jul 2016 22:42, edited 1 time in total.
Ebon
sinic
sinic
Posts: 354
Joined: 02 Jul 2016 20:55

Re: Questions on grammatical number & agreement

Post by Ebon »

Doranwen wrote:
Oh yes, I have decided that: no articles, lol. I am using a few cases (trying to limit to a reasonable number and finding ways to express some things with other means than separate individual cases, as you can see in my first thread. One of my aims is free word order, as much as possible, so I'd like to find ways to eliminate articles entirely and still be able to distinguish the different types of number. (Though I suppose if it doesn't work otherwise, I could always use articles and decline them right along with the nouns so there's agreement in case they can't be next to each other for some reason. But I'd still like to see if there's a way to eliminate the articles first. I'm not too fond of either articles or prepositions, the latter being one of the reasons why I have cases.)
You could try tacking the article on the noun in some way. Definite articles in Swedish are suffixes, if I remember correctly; would something similar work with your language?
Keenir
mayan
mayan
Posts: 2399
Joined: 22 May 2012 03:05

Re: Questions on grammatical number & agreement

Post by Keenir »

Doranwen wrote:The challenge I can see is that, if I'm having cases, does that mean I need separate singular/plural versions of cases, if I'm trying to keep it fusional?
nope.

to use the earlier names, you could say "Alice has two teas. Bob has one lemonade. Chris has two waters." Even if you say "Alice has two teas & Bob has one lemonade" you don't need separate versions of cases -- as long as you can tell who has what. (yes, I'm sure Alice has no problem sharing)

. But I'm starting to sort of see how the productivity/non-productivity works with this--if it's only part of set names, then it's no longer productive, but if one can use it for anything, whether it's been said before or not, then it's still productive and an active part of the language. Am I right?
welllllllllllllll...yes and no. it might be that the TRI is only used for religious events and religious objects; some linguists might analyse that as still being productive, others might say its nonproductive outside of its niche, etc.
At work on Apaan: viewtopic.php?f=6&t=4799
User avatar
Inkcube-Revolver
cuneiform
cuneiform
Posts: 126
Joined: 05 Nov 2015 23:20
Location: Miami, FL

Re: Questions on grammatical number & agreement

Post by Inkcube-Revolver »

Yes, with fusional languages they typically have singular and plural distinctions in their cases for how a word behaves in a sentence, but there are exceptions. With free word order, as long as you have the cases and their forms established, you should be able to have sentences come together without much worry save for any irregularities. Such irregularities could be dual and trial forms that, in this instance, would be vestigial, and could occur in only some words, like a word for "god" and "deity."
You can even go about distinguishing "god" with dual and trial forms pertaining to one main religion, and another word also meaning "god," but of a lesser or foreign variety lacking the duals and trials and still having a regular plural.

I think the examples that I gave with the words I made up in my previous reply are a good place to start, and you could modify the system of how mass nouns behave. You can have the genitive partake in the working of mass nouns like I gave with aquae earlier, or have a separate declension for the mass nouns altogether. Gender, linguistically speaking, could also help organize this, so that words decline differently when referring to countables and uncountables, such as a "cup" or "air," which I imagine to some extent a natural language could or would do.

This is all just kind of guesswork, and, again, I'd like to see where this goes. If this is any help at all, let me know.
I also think that the best way to go about this is to just have a set of words and play with them, get some results. I don't tend to worry too much about phonology because I get stumped on that aspect, but actually sitting there and playing with words usually yields me satisfying results instead of trying to do the math in my head. Just something to keep in mind if you haven't already done so.

(Also, belated apologies for my lengthy replies!)
I like my languages how I like my women: grammatically complex with various moods and tenses, a thin line between nouns and verbs, and dozens upon dozens of possible conjugations for every single verb.
Trailsend
moderator
moderator
Posts: 1451
Joined: 18 Aug 2010 05:22

Re: Questions on grammatical number & agreement

Post by Trailsend »

Oh I'll show you a lengthy reply ;)
Doranwen wrote:If a word like "grass" is collective, then the singulative would mean "a blade of grass" or "a piece of grass", right? If so, how would one then go about expressing "blades of grass"?...

For that matter, I found quantifiers very challenging, to think of another way to express them than exactly how English does it. I'd rather not just copy English, but pick features that are the right amount of complexity to suit me and that are interesting. I couldn't figure out how you'd say things like "several cows" or "a few blades of grass" or "some rice" besides just plunking a quantifier word down next to the noun...

If I use singular/plural, but also have a way to mark singulative, does that make it a mixed system? Is this too complex? I really like the idea of it. Or perhaps there's another way to get rid of the need to use specific unit nouns in order to count mass nouns: "grain" (of rice), "blade" (of grass), etc. It gets so muddled in my brain that I can't sort out what I'm trying to do, or whether there's another (better) way to do something that I'm just not thinking of... This is one of the areas I got things overly complicated last time I tried this, mostly because I didn't fully understand the systems and definitely didn't have any idea about how to properly handle quantifiers.
There's a trap that conlangers sometimes fall into when they start learning about a lot of the theory that linguists have constructed over time. Remember that linguistics (at least as practiced here for the most part) is descriptive: it aims to describe what languages are doing. This is why so many concepts can seem so messy, and why contradictions are so common; the world is very complex, and it can be difficult to find the most useful, most elegant description of the crazy things that a language gets up to.

For conlangers, though, this provides a powerful degree of freedom: at the end of the day, your language will do whatever it wants, and it will be up to you and other observers to figure out how to talk about it.

What this means is that if you find yourself caught in a loop of "but if there's a singulative and I have a collective noun and I need to put in a numeral but there's more than one than what is it supposed to be and what am I trying to accomplish anymore", you can always stop, set the theory aside, and start from the foundation.

What do you say?

Think of a situation. Think of a problem that you would need to solve, using language. Then put the words down in a way that you like. Come up with a similar situation, and put together the words that you like for that situation. And after you do this a few times, look back, and see if linguistics has names for what the language is doing.

As an example:

Suppose it's late September, and the leaves are starting to change color. Your problem is that you find the changing leaves beautiful, and you want to convey your thoughts to someone else.

1. telen-tim essai-us hess-te nunesek
leaf-ABS.??? depart-PTCP gold-ILL change
The leaves are starting to turn gold.

You have a framework for some general cases, so you know telentim should be absolutive as the single argument of nunesek. You also know you'd like to do something with number, and you like your suffixes fusional, so the -tim suffix should represent some combination of the absolutive case, and the number of the leaves. The specifics aren't important; you just know for now that telentim is how you talk about all the leaves that are changing.

What are other situations where you might want to talk about leaves in various quantities?

Maybe you're walking in the woods, and you want your friend to look at the cool leaf you found.

2. telen-ta hess-e ok sun-ni unek
leaf-ABS.??? gold-GEN that tree-ABL rest_upon
There's a golden leaf on that tree.

In (1), you were talking about all the leaves, a world's worth. This time you're just talking about the leaf you found, so using a different suffix feels right. So you use -ta, and know that it means some combination of the absolutive and the number of the leaf.

You continue coming up with examples like this, noticing and playing with patterns along the way. You use -tim to talk about all the dishes you have to wash, and for all the dogs at the dog park. You use -ta for the coffee mug you misplaced, and your brother's friend.

And then you find something different.

The changing leaves have started falling, and you come across a tree that has lost all but two of its leaves. Something about the light, and the look of the tree in the dusk just hits you, and you want to find a way to describe what you felt when you saw the tree.

2. anai telen-ti sun-ni unek
two leaf-ABS.??? tree-ABL rest_upon
There are two leaves on the tree.

The examples you've created so far had you thinking that -ta was singular number, and -tim was plural. If that were true, then you should use -ti here, because there's more than one leaf. But that feels wrong—you don't want to talk about these leaves the same way you talked about the changing leaves that were everywhere. You want -ti to be something different.

But what is it?

Maybe it's because there's a numeral here. Maybe -ti shows up anytime you specify the exact number of leaves. You play with that idea, create more examples: it feels right to use -ti when you need to explain that there's only one slice of cake left, or that you'll be ready to leave in three minutes. But somehow using -tim feels better than -ti when you talk about your friend's eleven siblings.

Maybe -ti is just for groups of a certain size or lower?

You experiment with that some, but you find you don't really like it either. Sometimes a set of three feels fine with -ti, like when you need three more minutes to get ready. But sometimes it doesn't—like when your brother leaves you to unload the rest of the car even though there's still three more boxes to carry.

Eventually you decide that the feeling behind -ti is not that there's a specific number of things, but that there's so few of them.

There were so few leaves on the tree, just one more slice of cake, just three more minutes until you could go. But three was an awful lot of boxes to carry yourself, and eleven is an awful lot of siblings to have, so -ti didn't fit.

At this point you start hunting around to find if there's an established name for something like this, and you come across the paucal number. This is great, because you have a name for the thing that you can use when people ask. But you also notice that your -ti doesn't quite act the same as the paucal number does in some of the languages that use the term; some languages seldom use the paucal to talk about numbers greater than ten, but you find no problem using -ti to talk about the twelve cents you have in your pocket.

And that's fine. The language is doing its thing, and "paucal" is just the best name you can give to what it's doing. Previously, if you had considered adding a paucal number, you might have wondered how to handle the paucal number when you were already using a numeral (because then don't you already know how many there are?), but now you know that your language uses the paucal to add additional information: it's not just that there's thing of the thing, it's that there's just three of the thing.


I find that this kind of exercise—of falling back on situations and thinking about how you'd like to talk about them—can help bounce you out of "theory angst", and illuminate practical solutions to theoretical problems.
任何事物的发展都是物极必反,否极泰来。
Keenir
mayan
mayan
Posts: 2399
Joined: 22 May 2012 03:05

Re: Questions on grammatical number & agreement

Post by Keenir »

Trailsend wrote:I find that this kind of exercise—of falling back on situations and thinking about how you'd like to talk about them—can help bounce you out of "theory angst", and illuminate practical solutions to theoretical problems.
*nods* [tick] [+1]

example statements, basically.
Inkcube-Revolver wrote: Such irregularities could be dual and trial forms that, in this instance, would be vestigial, and could occur in only some words, like a word for "god" and "deity."
You can even go about distinguishing "god" with dual and trial forms pertaining to one main religion, and another word also meaning "god," but of a lesser or foreign variety lacking the duals and trials and still having a regular plural.
Turkish does that, as well; in addition to words for the Abrahamic God, there's a word that's used both for referring to bygone deities like Amun and Matar and Men (I kid you not) and Kybele,,,and also for god-like beings in movies.
(Also, belated apologies for my lengthy replies!)
I've seen replies that took up an entire page all by themselves. I don't think you wrote any of them...yet. looking forwards to when you do.
At work on Apaan: viewtopic.php?f=6&t=4799
Doranwen
rupestrian
rupestrian
Posts: 19
Joined: 18 Jul 2016 02:49

Re: Questions on grammatical number & agreement

Post by Doranwen »

Ebon wrote:You could try tacking the article on the noun in some way. Definite articles in Swedish are suffixes, if I remember correctly; would something similar work with your language?
Hmm, an idea--though I've got case endings as suffixes too, so would end up with two suffixes. Not a problem really, just something to think about. Thanks for the suggestion.
Keenir wrote:
Doranwen wrote:The challenge I can see is that, if I'm having cases, does that mean I need separate singular/plural versions of cases, if I'm trying to keep it fusional?
nope.

to use the earlier names, you could say "Alice has two teas. Bob has one lemonade. Chris has two waters." Even if you say "Alice has two teas & Bob has one lemonade" you don't need separate versions of cases -- as long as you can tell who has what. (yes, I'm sure Alice has no problem sharing)
Hmm... when there are numerals that makes it pretty easy to tell that it's plural, even if the noun itself doesn't have it marked. What about when there's no numeral to tell? Such as "Alice picked up the pencil" vs. "Alice picked up the pencils". Or "Alice picked up a pencil" vs. "Alice picked up pencils"/"Alice picked up some pencils". From English one could argue that the definite article meant you already knew if there was one pencil or two (does that mean if you knew there were two but that she picks up just one, that you have to codify "one" or specify which or else it'll be thought that she picked up both of them?). And English uses the indefinite article vs. leaving it out or using a quantifier to encode the difference, but I'd prefer to put more of it on the noun itself when possible.
Keenir wrote:
Doranwen wrote:But I'm starting to sort of see how the productivity/non-productivity works with this--if it's only part of set names, then it's no longer productive, but if one can use it for anything, whether it's been said before or not, then it's still productive and an active part of the language. Am I right?
welllllllllllllll...yes and no. it might be that the TRI is only used for religious events and religious objects; some linguists might analyse that as still being productive, others might say its nonproductive outside of its niche, etc.
Lol, I see what you mean. It's fun to imagine linguists arguing over structures in my conlang. Thanks for the clarification.
Inkcube-Revolver wrote:Yes, with fusional languages they typically have singular and plural distinctions in their cases for how a word behaves in a sentence, but there are exceptions. With free word order, as long as you have the cases and their forms established, you should be able to have sentences come together without much worry save for any irregularities. Such irregularities could be dual and trial forms that, in this instance, would be vestigial, and could occur in only some words, like a word for "god" and "deity."
I particularly like the idea of using dual forms to refer to couples of things, whether it be birds mating, a pair of socks, one's hands, etc. Basically, that you can't use dual for just "two", it has to be a pair of some sort. So the dual is becoming more restricted in its use and the trial is almost gone. Which would mean I'd need dual forms, but the trial forms would be definite irregularities.
Inkcube-Revolver wrote:I think the examples that I gave with the words I made up in my previous reply are a good place to start, and you could modify the system of how mass nouns behave. You can have the genitive partake in the working of mass nouns like I gave with aquae earlier, or have a separate declension for the mass nouns altogether. Gender, linguistically speaking, could also help organize this, so that words decline differently when referring to countables and uncountables, such as a "cup" or "air," which I imagine to some extent a natural language could or would do.
Yeah, I may just have a random particle that acts as a singulative (sort of like your -kta ending), and have it be added as a suffix before case endings. Then it can be turned into a separate word for usage purposes, like it could be used to form new coinages on the spot, but would also be codified into some words (like "blade of grass" might be a word in the dictionary, thanks to that particle). And the particle for the reverse, I'd sort of come up with--I want to use one that will encode groups, for instance. Y'know, all those terms of venery that English came up with, all replaced by one all-purpose particle! *giggles* It can go in my category of "particles that modify meaning", along with the equivalent of -phile and -phobe, and "inhabitant/member of", etc. I've got one for "language of" and have had all kinds of fun thinking about the difference between someone speaking "English" vs. speaking "American". (And then imagine the arguments over "language of South Africa", which of course evolves into the need for separate words for a lot of languages--or else resorting to specific regions within the country--which adds a fun quirk to the language names.)
Inkcube-Revolver wrote:This is all just kind of guesswork, and, again, I'd like to see where this goes. If this is any help at all, let me know.
I also think that the best way to go about this is to just have a set of words and play with them, get some results. I don't tend to worry too much about phonology because I get stumped on that aspect, but actually sitting there and playing with words usually yields me satisfying results instead of trying to do the math in my head. Just something to keep in mind if you haven't already done so.
Definitely helpful! During the day when I'm away from the computer, I'm letting the ideas play in my head, and feeling a lot more like I actually understand this stuff--and have ideas of what I want to do with it. The examples were helpful too. I've been doing a little playing with the words. If I had figured out what I wanted to do with verbs roughly, I'd start playing with it. But I'm still up in the air about what I want to encode or not, so I'm thinking I may need just one more post to get clarification on some of the concepts I'm looking at.
Inkcube-Revolver wrote:(Also, belated apologies for my lengthy replies!)
Lol, I never mind long answers. :)
Trailsend wrote:There's a trap that conlangers sometimes fall into when they start learning about a lot of the theory that linguists have constructed over time. Remember that linguistics (at least as practiced here for the most part) is descriptive: it aims to describe what languages are doing. This is why so many concepts can seem so messy, and why contradictions are so common; the world is very complex, and it can be difficult to find the most useful, most elegant description of the crazy things that a language gets up to.

For conlangers, though, this provides a powerful degree of freedom: at the end of the day, your language will do whatever it wants, and it will be up to you and other observers to figure out how to talk about it.
To some degree, yes--my problem is that if I never have any other example in front of me, I will totally do exactly what English does because I cannot think of any other way to do it. The descriptive linguistics theory that's in the LCK and such guides has been invaluable to me because it presents ideas that I never would've thought of in a million years. The problem then is that I can only think to use it as it's been described there. So all my discussion here and asking for help is me trying to expand my mental horizon, loosen up the tight definitions and examples and be more creative with the usage. When I was working with the old version of the language, I did come up with ways to use the quantifiers and all--the same way English did it. And I was unhappy with that because I knew that meant I was just copying English without choosing to--I didn't use individual words in front of nouns for quantifiers because that was the option I liked best, I did it because I couldn't think of any other way to possibly do it. For that matter, I still can't.
Trailsend wrote:What this means is that if you find yourself caught in a loop of "but if there's a singulative and I have a collective noun and I need to put in a numeral but there's more than one than what is it supposed to be and what am I trying to accomplish anymore", you can always stop, set the theory aside, and start from the foundation.
A little dialogue did shake me out of that, and InkcubeRevolver's different examples helped me settle on something very much like what I'd done before with it--except now I feel like I know what I'm doing with it, and I've decided how I want to handle the inflections. So I'm happy with that. :)
Trailsend wrote:What do you say?

Think of a situation. Think of a problem that you would need to solve, using language. Then put the words down in a way that you like. Come up with a similar situation, and put together the words that you like for that situation. And after you do this a few times, look back, and see if linguistics has names for what the language is doing.
I'll keep trying that. I did that a little for the last version of the language, when I ran into "I have no idea how to do this". But I ended up either copying English in a lot of cases, or doing things that ended up clunky and made no sense, and I was generally unhappy with the result but had no idea what to do with it to make it better. Sometimes I just need another pair of eyes on my thought processes to point out what I'm just not seeing, lol. (Like your comment on my other thread that it looked like I was talking about conjugation--of course it was. And clarifying that meant that I came across a blind spot that needed addressing, that I wouldn't have thought of without the right name.) What I'm finding really helpful is discussing all these possible ways, talking about ideas that I have--so I know if I'm just missing something obvious, and having names for what I'm doing really does help me try out ideas to an extent, but I'm still toying with them in my head. Once I know more or less the main dimensions I'm trying to include, and have an idea of how they'll relate to each other, I'll just start playing around with it. One of the most helpful things is from all this discussion I've decided am not going to go and create all the inflections in each chart at once. I think that was one mistake I made last time. I still need to know what the chart should hold (and how many I need), but I plan on creating them blank--and then just playing with the language, and creating them as I need them. I think that'll work a lot better. Thank you for all your help and suggestions, it's really been useful.
Trailsend wrote:I find that this kind of exercise—of falling back on situations and thinking about how you'd like to talk about them—can help bounce you out of "theory angst", and illuminate practical solutions to theoretical problems.
I'll try that more, now that I feel like I have most of the noun situations sorted out. At this point I just would love some ideas of how to express quantifiers like "some", "many", etc., if there's any way to do it besides English. And even if I do use the individual-word-in-front-of-noun method, would I possibly mark it in any way or would I maybe need to establish some word order rules for quantifiers and the nouns they modify? What's been done? The more ideas of how various languages have done things, the better I can pick what suits me, and play around with it.
Keenir wrote:
Inkcube-Revolver wrote: Such irregularities could be dual and trial forms that, in this instance, would be vestigial, and could occur in only some words, like a word for "god" and "deity."
You can even go about distinguishing "god" with dual and trial forms pertaining to one main religion, and another word also meaning "god," but of a lesser or foreign variety lacking the duals and trials and still having a regular plural.
Turkish does that, as well; in addition to words for the Abrahamic God, there's a word that's used both for referring to bygone deities like Amun and Matar and Men (I kid you not) and Kybele,,,and also for god-like beings in movies.
Oh neat! I like that idea. :D *makes note*
Keenir
mayan
mayan
Posts: 2399
Joined: 22 May 2012 03:05

Re: Questions on grammatical number & agreement

Post by Keenir »

Doranwen wrote:
Keenir wrote:
Doranwen wrote:The challenge I can see is that, if I'm having cases, does that mean I need separate singular/plural versions of cases, if I'm trying to keep it fusional?
nope.
to use the earlier names, you could say "Alice has two teas. Bob has one lemonade. Chris has two waters." Even if you say "Alice has two teas & Bob has one lemonade" you don't need separate versions of cases -- as long as you can tell who has what. (yes, I'm sure Alice has no problem sharing)
Hmm... when there are numerals that makes it pretty easy to tell that it's plural, even if the noun itself doesn't have it marked. What about when there's no numeral to tell? Such as "Alice picked up the pencil" vs. "Alice picked up the pencils". Or "Alice picked up a pencil" vs. "Alice picked up pencils"/"Alice picked up some pencils". From English one could argue that the definite article meant you already knew if there was one pencil or two (does that mean if you knew there were two but that she picks up just one, that you have to codify "one" or specify which or else it'll be thought that she picked up both of them?).
or "Alice picked up the pencil" could mean there were, on the table, a #2 pencil, a mechanical pencil, and a pastel pencil - but Alice (or the narrator) was of the opinion that only one of those three was a true pencil.

remember, you don't need to specify everything. some languages have no problem omitting information.

like in English - "I said hi to my neighbor." (when? is my neighbor an adult or child, a man or woman? was I walking or driving at the time?)

or in Mandarin - "I ate rice." (when? what kind of rice? what did I eat with? was I eating with anyone?)
Thanks for the clarification.
welcome.
To some degree, yes--my problem is that if I never have any other example in front of me, I will totally do exactly what English does because I cannot think of any other way to do it. The descriptive linguistics theory that's in the LCK and such guides has been invaluable to me because it presents ideas that I never would've thought of in a million years. The problem then is that I can only think to use it as it's been described there.
if I remember my copy of the LCK, there's multiple methods from several languages in there.
So all my discussion here and asking for help is me trying to expand my mental horizon, loosen up the tight definitions and examples and be more creative with the usage. When I was working with the old version of the language, I did come up with ways to use the quantifiers and all--the same way English did it. And I was unhappy with that because I knew that meant I was just copying English without choosing to--I didn't use individual words in front of nouns for quantifiers because that was the option I liked best, I did it because I couldn't think of any other way to possibly do it. For that matter, I still can't.
copying English isn't a bad thing, particularly when you're starting out. perhaps this early version of your conlang, is a mixed language of your conlang & English (not unheard of), and as you learn more about how English does things, you can cut away how much English there is in the mixed language, until all that remains is your conlang.

though also, just because English does something, doesn't mean you have to avoid that strategy like the plague.

What I'm finding really helpful is discussing all these possible ways, talking about ideas that I have--so I know if I'm just missing something obvious, and having names for what I'm doing really does help me try out ideas to an extent, but I'm still toying with them in my head. Once I know more or less the main dimensions I'm trying to include, and have an idea of how they'll relate to each other, I'll just start playing around with it.
then just playing with the language, and creating them as I need them.
both strategies are good, and none of them contradict each other.
(even making a chart can be a good starting point - if you use that as the skeleton of the tree, which you build statements upon...and making example sentences can show you where you go back and fix one or two spots on the chart)
Trailsend wrote:I find that this kind of exercise—of falling back on situations and thinking about how you'd like to talk about them—can help bounce you out of "theory angst", and illuminate practical solutions to theoretical problems.
I'll try that more, now that I feel like I have most of the noun situations sorted out. At this point I just would love some ideas of how to express quantifiers like "some", "many", etc., if there's any way to do it besides English.[/quote]

it might be apocraphal, but I think I heard of some languages having no "some" & "many" qualifiers...they get by with numbers. and others, Piraha most famously of them, have neither numbers nor "many" and "some", but only have "more" and "fewer".
And even if I do use the individual-word-in-front-of-noun method, would I possibly mark it in any way or would I maybe need to establish some word order rules for quantifiers and the nouns they modify? What's been done? The more ideas of how various languages have done things, the better I can pick what suits me, and play around with it.
WALS is useful...but a bit technical.
http://wals.info/languoid
At work on Apaan: viewtopic.php?f=6&t=4799
Doranwen
rupestrian
rupestrian
Posts: 19
Joined: 18 Jul 2016 02:49

Re: Questions on grammatical number & agreement

Post by Doranwen »

Keenir wrote:
Doranwen wrote:Hmm... when there are numerals that makes it pretty easy to tell that it's plural, even if the noun itself doesn't have it marked. What about when there's no numeral to tell? Such as "Alice picked up the pencil" vs. "Alice picked up the pencils". Or "Alice picked up a pencil" vs. "Alice picked up pencils"/"Alice picked up some pencils". From English one could argue that the definite article meant you already knew if there was one pencil or two (does that mean if you knew there were two but that she picks up just one, that you have to codify "one" or specify which or else it'll be thought that she picked up both of them?).
or "Alice picked up the pencil" could mean there were, on the table, a #2 pencil, a mechanical pencil, and a pastel pencil - but Alice (or the narrator) was of the opinion that only one of those three was a true pencil.

remember, you don't need to specify everything. some languages have no problem omitting information.

like in English - "I said hi to my neighbor." (when? is my neighbor an adult or child, a man or woman? was I walking or driving at the time?)

or in Mandarin - "I ate rice." (when? what kind of rice? what did I eat with? was I eating with anyone?)
Quite true. I have a really hard time leaving my language vague in areas where English is more specific. I'm not necessarily trying to eliminate all ambiguity (and I figure most languages have some way to disambiguate just about anything, even if it takes a bunch of extra words), but that's just really hard for me to do, lol. I have decided on some things I have no interest in including morphologically, however, even if it means omitting some information--politeness (like Japanese), information about the listener (like Basque), evidentiality (like some Austronesian languages), etc. They're neat features, but there's a limit to what I really want to throw in my language, and I can think of other ways to put all of that stuff in there besides morphologically, if I really need that info in a given sentence (syntax, personal pronouns, phrases).
Keenir wrote:if I remember my copy of the LCK, there's multiple methods from several languages in there.
Yeah, I keep going back to it and getting more information from it. The only thing I've found for quantifiers in it is to extend the correlatives table, but that doesn't help with "some pencils", which I actually do want to be able to say.
Keenir wrote:copying English isn't a bad thing, particularly when you're starting out. perhaps this early version of your conlang, is a mixed language of your conlang & English (not unheard of), and as you learn more about how English does things, you can cut away how much English there is in the mixed language, until all that remains is your conlang.

though also, just because English does something, doesn't mean you have to avoid that strategy like the plague.
True, true. I just don't like having to choose it as a default, because it's the only thing I know about. Much different when I have several options and I decide that the English method suits better than anything else. (I'd actually borrow more strategies from English except that the general idea of how my language is structured is different enough that English's methods just wouldn't fit with the rest of the language in most situations.)
Keenir wrote: (even making a chart can be a good starting point - if you use that as the skeleton of the tree, which you build statements upon...and making example sentences can show you where you go back and fix one or two spots on the chart)
Yeah, I figured I'd find things I wanted to change, eliminate, tweak, etc. But I need to have that structure there, initially, to have an idea of what I'm doing; I really can't just throw stuff out there and then go "oh, I guess I'm marking verbs by this, this, and this", lol. My brain just won't do that.
Keenir wrote:it might be apocraphal, but I think I heard of some languages having no "some" & "many" qualifiers...they get by with numbers. and others, Piraha most famously of them, have neither numbers nor "many" and "some", but only have "more" and "fewer".
But I like them! And want them! Lol. Yeah, I've heard of Piraha's lack of numbers, but it's not for me. I'm a math teacher/tutor, actually, so I want to be able to actually do math in my language. :D
Keenir wrote:WALS is useful...but a bit technical.
I find the very technical rather hard to wade through, but I've attempted it. I wasn't able to find anything on the quantifiers, though, but I might just be missing what it calls them... Sometimes knowing which feature is the one you're trying to learn about is challenging in and of itself.
Keenir
mayan
mayan
Posts: 2399
Joined: 22 May 2012 03:05

Re: Questions on grammatical number & agreement

Post by Keenir »

Doranwen wrote:Quite true. I have a really hard time leaving my language vague in areas where English is more specific.
it is tricky.
Keenir wrote:if I remember my copy of the LCK, there's multiple methods from several languages in there.
Yeah, I keep going back to it and getting more information from it. The only thing I've found for quantifiers in it is to extend the correlatives table, but that doesn't help with "some pencils", which I actually do want to be able to say.
the ALC (its sequel) has quantifiers in the chapter on logical languages, which may miss the point of what you're doing.
Keenir wrote:copying English isn't a bad thing, particularly when you're starting out. perhaps this early version of your conlang, is a mixed language of your conlang & English (not unheard of), and as you learn more about how English does things, you can cut away how much English there is in the mixed language, until all that remains is your conlang.
though also, just because English does something, doesn't mean you have to avoid that strategy like the plague.
True, true. I just don't like having to choose it as a default, because it's the only thing I know about. Much different when I have several options and I decide that the English method suits better than anything else. (I'd actually borrow more strategies from English except that the general idea of how my language is structured is different enough that English's methods just wouldn't fit with the rest of the language in most situations.)

Yeah, I figured I'd find things I wanted to change, eliminate, tweak, etc. But I need to have that structure there, initially, to have an idea of what I'm doing;
makes sense... maybe use the English feature/word as a placeholder until you find a replacement that you like.
Keenir wrote:WALS is useful...but a bit technical.
I find the very technical rather hard to wade through, but I've attempted it. I wasn't able to find anything on the quantifiers, though, but I might just be missing what it calls them... Sometimes knowing which feature is the one you're trying to learn about is challenging in and of itself.
*nods* the "what's the word for this?" happens to us all.
At work on Apaan: viewtopic.php?f=6&t=4799
Post Reply