Kovur Grammar - Thoughts, Questions, Comments?

If you're new to these arts, this is the place to ask "stupid" questions and get directions!
User avatar
sangi39
moderator
moderator
Posts: 3024
Joined: 12 Aug 2010 01:53
Location: North Yorkshire, UK

Kovur Grammar - Thoughts, Questions, Comments?

Post by sangi39 »

I've been trying to think about more than just the sounds of Kovur languages, but I'm honestly struggling to decide what I want to do here.

The only thing I do know right now is that because Kovur languages are, like human languages, primarily spoken, so parts of a given utterance appear linearly, one after the other (so no saying two words at once, or simultaneous verbal and non-verbal signs). They're not human though, so I'm not sure, for example, whether they'll build up sentences the same way (or a similar way to the tree), or whether they'll have the same, or similar, syntactic categories and word classes.

For word classes, at the very least, a division between "things" (concrete nouns) and "actions" ((dynamic?) verbs) makes sense, but I don't know whether to have, for example, abstract concepts or states to be treated as separate syntactic categories that behave differently in all Kovur languages.

Something I had been thinking about, is for Kovur languages to make a distinction between "proximal" and "distal" things, states, actions and concepts throughout the language, without using pronouns, I suppose similar to egophoricity in Tibetan. So, as an example, there would be some states and actions that would by default be attributed directly to the speaker in all circumstances while others would be associated with everything else. So you could have "(I) think" but not "(he) thinks", and instead you'd have to say "(he) has a thought", and similarly you could have "(he) will die" but "death (my) will come" instead of "(I) will die". If I followed through on that, I'd like to avoid the use of distinct personal pronouns, possibly have the system extend out to "proximal" and "distal" based on context e.g. the second person is represented by proximal forms by default, and the third person by the distal form, but the proximate can also be used to indicate the proximate third person when there's an obviative third person which would then be represented by the distal forms.

I think I might be able to build on this system and while making it flexible enough that different Kovur languages can handle this sort of thing with different mechanisms, but if anyone has any other thoughts on how these languages can behave differently from human languages, I'd love some feedback or input [:)]
You can tell the same lie a thousand times,
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

Re: Kovur Grammar - Thoughts, Questions, Comments?

Post by Creyeditor »

What I did, am doing, plan to do, with a non-human language is to focus on situations. There is thus no deep distinction between actions and things anymore. Just because you asked for input. This information might help.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
Clio
sinic
sinic
Posts: 228
Joined: 27 Dec 2012 23:45

Re: Kovur Grammar - Thoughts, Questions, Comments?

Post by Clio »

My thought isn't informed by any thoughts about human syntax or xenopsychology, but rather by an aesthetic impression I got from your other thread about Kovur phonology. Basically, I felt like, after looking at all the charts with tons of dorsal points of articulation and all the vowels, the Kovur languages gave off a very maximalist image. That's not a bad thing, but I think it would be pleasing artistically to see that trend towards "more-more-more" balanced out a bit in the syntax. The proximal/distal contrast is really appealing to me for precisely that reason: most or all sentences would involve not the typical human ternary person distinction but rather a binary one. So I find that attractive.

Are there any other morphosyntactic categories that you might delete (without running into clichés like "nounlessness")? I don't know, and I don't mean to suggest that removing categories is the only thing I'd like to see, but it could be interesting!
Niûro nCora
Getic: longum Getico murmur in ore fuit
scratchpad
User avatar
sangi39
moderator
moderator
Posts: 3024
Joined: 12 Aug 2010 01:53
Location: North Yorkshire, UK

Re: Kovur Grammar - Thoughts, Questions, Comments?

Post by sangi39 »

Creyeditor wrote: 31 May 2018 21:12 What I did, am doing, plan to do, with a non-human language is to focus on situations. There is thus no deep distinction between actions and things anymore. Just because you asked for input. This information might help.
I'm not quite sure I understand this fully, but what would these "situations" be, and how are they different from, say, nouns or verbs?
You can tell the same lie a thousand times,
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

Re: Kovur Grammar - Thoughts, Questions, Comments?

Post by Creyeditor »

They include actions plus things that are associated with the action. A sentence like 'the boy admires the girl' would be a single atomic lexeme. The sentence 'the woman admires the girl' would be completely different.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
User avatar
sangi39
moderator
moderator
Posts: 3024
Joined: 12 Aug 2010 01:53
Location: North Yorkshire, UK

Re: Kovur Grammar - Thoughts, Questions, Comments?

Post by sangi39 »

Clio wrote: 01 Jun 2018 05:34 My thought isn't informed by any thoughts about human syntax or xenopsychology, but rather by an aesthetic impression I got from your other thread about Kovur phonology. Basically, I felt like, after looking at all the charts with tons of dorsal points of articulation and all the vowels, the Kovur languages gave off a very maximalist image. That's not a bad thing, but I think it would be pleasing artistically to see that trend towards "more-more-more" balanced out a bit in the syntax. The proximal/distal contrast is really appealing to me for precisely that reason: most or all sentences would involve not the typical human ternary person distinction but rather a binary one. So I find that attractive.

Are there any other morphosyntactic categories that you might delete (without running into clichés like "nounlessness")? I don't know, and I don't mean to suggest that removing categories is the only thing I'd like to see, but it could be interesting!
First, just to say it, thanks for reading my other threads [:)] I did try to balance the "more POAs" in the Kovur speech abilities with "fewer MOAs", i.e. a lack of ability to pronounce sibilants, and of course no labiodentals or labial-dentals (the latter of which is already rare in human languages anyway), so I'm definitely aiming for some sort of balance in the language abilities of the Kovur overall.

Similarly, that was the aim with the proximal-distal idea. It adds something to the language (all verbs must be either one or other, so to express that a "proximal subject" performs "distal action" you have to use some sort of alternative construction), while it takes out, as you said, the ternary person distinction (well, it doesn't really, it's still there, but it's more contextual).

I do want to try and get rid of the ternary distinction a bit more, and I'm trying to think of how to deal with possession. The proximal-distal thing at the moment is specific to the speaker, so I'd assume that possession would be based on alienability almost exclusively in Kovur languages, in some form or another. Like, "hand" would, by default, refer to the speaker's hand, so to indicate "your hand", or "his hand", the speaker would have to refer to a hand that is "far from" themselves, e.g. "that arm" or "that arm over there". I don't know yet, though, how on would refer to alienable possessions, such as a house. The physical house might not be close to the speaker at the time of speaking, so Kovur languages might have get around this by saying something like "this house over there" or "the house that (I) have".

At the moment, though, I'm really not sure what other directions I could take the languages of the Kovur in. I honestly think this might be enough to start with, since it messes with verbs, possession, and I assume it would have an impact on the ideas of "subject", "object", "agent", "patient", etc. The ideas of "one who acts" and "one who is acted upon" might still exist, but because of how verbs are divided between personal/proximal/distal, the default for some verbs might be passive-like in terms of voice, so for the agent to appear as the subject, the voice of a verb might need to be changed.

(I'm hoping this makes sense)

Other parts of speech, like adjectives, articles, etc. might be dependent on the individual language, as they are in human languages. Some Kovur languages might have a distinct adjectival word class, while others might have adjectives instead be nouns, or verbs, or possibly adverbs.
You can tell the same lie a thousand times,
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
User avatar
sangi39
moderator
moderator
Posts: 3024
Joined: 12 Aug 2010 01:53
Location: North Yorkshire, UK

Re: Kovur Grammar - Thoughts, Questions, Comments?

Post by sangi39 »

Creyeditor wrote: 01 Jun 2018 23:04 They include actions plus things that are associated with the action. A sentence like 'the boy admires the girl' would be a single atomic lexeme. The sentence 'the woman admires the girl' would be completely different.
Are there no nouns or verbs at all, or noun-like or verb-like morphemes? Like, for example, could "the boy admires the girl" be modified with a noun-like morpheme meaning "actress" to mean "the boy admires the actress"? Or can situations only be modified by other situations, such that "the boy admires the girl" would become "the boy admires the actress" by means of some "situation" like "she is an actress"?
You can tell the same lie a thousand times,
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
Clio
sinic
sinic
Posts: 228
Joined: 27 Dec 2012 23:45

Re: Kovur Grammar - Thoughts, Questions, Comments?

Post by Clio »

sangi39 wrote: 01 Jun 2018 23:19 Similarly, that was the aim with the proximal-distal idea. It adds something to the language (all verbs must be either one or other, so to express that a "proximal subject" performs "distal action" you have to use some sort of alternative construction), while it takes out, as you said, the ternary person distinction (well, it doesn't really, it's still there, but it's more contextual).

I do want to try and get rid of the ternary distinction a bit more, and I'm trying to think of how to deal with possession. The proximal-distal thing at the moment is specific to the speaker, so I'd assume that possession would be based on alienability almost exclusively in Kovur languages, in some form or another. Like, "hand" would, by default, refer to the speaker's hand, so to indicate "your hand", or "his hand", the speaker would have to refer to a hand that is "far from" themselves, e.g. "that arm" or "that arm over there". I don't know yet, though, how on would refer to alienable possessions, such as a house. The physical house might not be close to the speaker at the time of speaking, so Kovur languages might have get around this by saying something like "this house over there" or "the house that (I) have".
I completely agree, and seeing how balanced the proximal-distal system is was really enjoyable. I think you definitely have more than enough opportunities from and consequences of the proximal-distal system to track down so as to give you plenty of leads and things to write about.

I wonder if possession might be influenced at all by some aspect of Kovur psychology. If they're based on wolves, do they have a pack structure? Could possession possibly relate to ownership within a pack: e.g., "my house" = house.PROX.in-pack; "your/his house" (of a pack member) = house.DIST.in-pack; "your house" (of a non-pack member) = house.PROX.out-pack; "his house" (of a non-pack member) = house.DIST.out-pack. Or maybe the proximal/distal distinction cannot be applied outside of the pack, and some fine distinctions have to be expressed by circumlocution. After all, just because humans tend to care who precisely owns something doesn't mean that the Kovur should; maybe pack ownership is simply more important to them.

Your other ideas about agency and voice also make sense, and I look forward to seeing them more fleshed-out!
Niûro nCora
Getic: longum Getico murmur in ore fuit
scratchpad
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

Re: Kovur Grammar - Thoughts, Questions, Comments?

Post by Creyeditor »

sangi39 wrote: 01 Jun 2018 23:33
Creyeditor wrote: 01 Jun 2018 23:04 They include actions plus things that are associated with the action. A sentence like 'the boy admires the girl' would be a single atomic lexeme. The sentence 'the woman admires the girl' would be completely different.
Are there no nouns or verbs at all, or noun-like or verb-like morphemes? Like, for example, could "the boy admires the girl" be modified with a noun-like morpheme meaning "actress" to mean "the boy admires the actress"? Or can situations only be modified by other situations, such that "the boy admires the girl" would become "the boy admires the actress" by means of some "situation" like "she is an actress"?
The first one, I would not want to be possible, for my case. It looks, kind of, like cheating to me. But of course that is not necessarily true for other people implementing a similar idea.
I guess the second one could be true: [the boy admires the girl] AND [there is an actress].
I am not sure how I handle pronouns. Maybe there are different conjunctions for similar subject and different subject conjunctions? Of course, situations have to be modifiable (maybe for tense, etc) and situations have to conjoinable (like conjunction, disjunction, temporal sequence), if you want to retain a certain expressive range.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
User avatar
sangi39
moderator
moderator
Posts: 3024
Joined: 12 Aug 2010 01:53
Location: North Yorkshire, UK

Re: Kovur Grammar - Thoughts, Questions, Comments?

Post by sangi39 »

Clio wrote: 02 Jun 2018 04:47
sangi39 wrote: 01 Jun 2018 23:19 Similarly, that was the aim with the proximal-distal idea. It adds something to the language (all verbs must be either one or other, so to express that a "proximal subject" performs "distal action" you have to use some sort of alternative construction), while it takes out, as you said, the ternary person distinction (well, it doesn't really, it's still there, but it's more contextual).
I completely agree, and seeing how balanced the proximal-distal system is was really enjoyable. I think you definitely have more than enough opportunities from and consequences of the proximal-distal system to track down so as to give you plenty of leads and things to write about.
The main hope is that it doesn't become too "strict", and there can be classifications of Kovur languages based on how they handle this distinction, similar to the different groups of morphosyntactic alignments amongst human languages.


Clio wrote: 02 Jun 2018 04:47
sangi39 wrote: 01 Jun 2018 23:19 I do want to try and get rid of the ternary distinction a bit more, and I'm trying to think of how to deal with possession. The proximal-distal thing at the moment is specific to the speaker, so I'd assume that possession would be based on alienability almost exclusively in Kovur languages, in some form or another. Like, "hand" would, by default, refer to the speaker's hand, so to indicate "your hand", or "his hand", the speaker would have to refer to a hand that is "far from" themselves, e.g. "that arm" or "that arm over there". I don't know yet, though, how on would refer to alienable possessions, such as a house. The physical house might not be close to the speaker at the time of speaking, so Kovur languages might have get around this by saying something like "this house over there" or "the house that (I) have".
I wonder if possession might be influenced at all by some aspect of Kovur psychology. If they're based on wolves, do they have a pack structure? Could possession possibly relate to ownership within a pack: e.g., "my house" = house.PROX.in-pack; "your/his house" (of a pack member) = house.DIST.in-pack; "your house" (of a non-pack member) = house.PROX.out-pack; "his house" (of a non-pack member) = house.DIST.out-pack. Or maybe the proximal/distal distinction cannot be applied outside of the pack, and some fine distinctions have to be expressed by circumlocution. After all, just because humans tend to care who precisely owns something doesn't mean that the Kovur should; maybe pack ownership is simply more important to them.
Grah! After all that "different syntax different species" stuff I forgot that they were descended from wolves [:P] I might not word it as an "in-pack" vs. "our-pack" distinction, but maybe two overlapping levels of proximity, e.g. PROX-PROX (inalienable mine, i.e. possessor nearby + possessed nearby) vs. PROX-DIST (alienable mine, i.e. possessor nearby + possessed far away) vs. DIST-PROX (inalienable yours or his/hers, i.e. possessor far away + possessed near [to the possessor]) vs. DIST-DIST (alienable yours or his/hers, i.e. possessor far away + possessed far away [from the possessor]).

I think overall, regardless of how the individual language handles possession, it would probably be based on a) distance to the possessor, and b) "distance" between the possessor and the possessed.



The one thing I had thought about this morning was whether what constitutes "proximal" and "distal" in each language might vary, e.g. whether proximal verbs cover the listener as well as the speaker, or whether the listener is covered by distal verbs. The two categories will always exist in each Kovur language, but different languages might place that dividing line in different places.


Clio wrote: 02 Jun 2018 04:47 Your other ideas about agency and voice also make sense, and I look forward to seeing them more fleshed-out!
I think my main worry at the moment is terminology, lol. I'm already getting the distal/proximal thing a bit messed up n my head because I'm using those two words to refer to maybe three different distinctions [:P]
You can tell the same lie a thousand times,
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
User avatar
sangi39
moderator
moderator
Posts: 3024
Joined: 12 Aug 2010 01:53
Location: North Yorkshire, UK

Re: Kovur Grammar - Thoughts, Questions, Comments?

Post by sangi39 »

Creyeditor wrote: 02 Jun 2018 08:16
sangi39 wrote: 01 Jun 2018 23:33
Creyeditor wrote: 01 Jun 2018 23:04 They include actions plus things that are associated with the action. A sentence like 'the boy admires the girl' would be a single atomic lexeme. The sentence 'the woman admires the girl' would be completely different.
Are there no nouns or verbs at all, or noun-like or verb-like morphemes? Like, for example, could "the boy admires the girl" be modified with a noun-like morpheme meaning "actress" to mean "the boy admires the actress"? Or can situations only be modified by other situations, such that "the boy admires the girl" would become "the boy admires the actress" by means of some "situation" like "she is an actress"?
The first one, I would not want to be possible, for my case. It looks, kind of, like cheating to me. But of course that is not necessarily true for other people implementing a similar idea.
I guess the second one could be true: [the boy admires the girl] AND [there is an actress].
I am not sure how I handle pronouns. Maybe there are different conjunctions for similar subject and different subject conjunctions? Of course, situations have to be modifiable (maybe for tense, etc) and situations have to conjoinable (like conjunction, disjunction, temporal sequence), if you want to retain a certain expressive range.
Ohhh, that makes sense. So there's a set of "situational units" ("stasemes", maybe, from Greek στάση), with meaning becoming more and more detailed through the use of more and more situational units. I'd definitely be interested to see how adjectives are handled, and, as you said, pronouns too.
You can tell the same lie a thousand times,
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: Kovur Grammar - Thoughts, Questions, Comments?

Post by Salmoneus »

It would not be possible to have a language where basic transitive sentences were not composable. The number of possible agents, patients and actions is far too great - most 'situations' would never arise more than once in your lifetime, so it would be impossible to learn them.

That's the whole point about language: that with a finite, and hence learnable, set of subunits (words) we can construct an infinite number of superunits (clauses) to meet any situation. If your superunits, 'situations', cannot be decomposed intelligibly into a small number of learnable subunits, they will not be themselves learnable.

That said, it is possible to make your verbs more specific than in most human languages. For instance, in English we have progressively specific "he depicted the fox", "he marked out the outline of the fox", "he drew the outline of the fox", and "he pencilled the outline of the fox". English usually gets more specific by embedding more assumptions about manners and instruments, but other languages might do it differently. Some do do it at least partially with relevance to the arguments - so there may be suppletive verbs for "a man spoke" and "a woman spoke", for instance. But it's much rarer, I think. Maybe your alien language embeds arguments into verbs more frequently. It would only, however, be able to do it with relatively few arguments, because the total number of verbs would have to remain low enough to be learnable.

This is basically what's happening in Sangi's suggestion. Note, however, that it's no good just saying [the boy admired the girl][there was an actress][there was a soldier] - because how do you tell which is admiring which? Because already when you're doing that, you've got a verb (boy/girl-admires), and two nouns (there-is-an-actress) and (there-is-a-soldier) [NB if 'actress' only appears in the 'word' meaning 'there is an actress', then it doesn't mean 'there is an actress', it just means 'actress' - without the possibility of contrast, there is no distinction!] - and to make it intelligible you need to make one of those nouns an agent and one a patient.

So the result is not fundamentally different from English: subject, verb, object. That's the problem with language: it's pretty much impossible to escape that framework. Because you're trying to describe events, and what things were involved in them, and what role the involved things had.

You can change what words look like. Or how words relate to meanings. Or how words are structured into sentences syntactically. Or the pragmatics of what sort of things are talked about. But ultimatly you can't really change the semantic framework of agent, patient, noun, verb, because all languages are talking about the same sort of things.

That said, the system you and sangi suggest would be distinctive and noticeable weird by human standards. But bear in mind: it's weird because it's a language with many noun classes, and unusually high levels of verbal suppletion by noun class*, not because it does anything to challenge the noun/verb dichotomy.





*which raises the question: if you're having to have all these nouns anyway, what good does massive verbal suppletion do other than making the language harder to learn? Well, the obvious answer would be that it's counterbalanced by massive nominal homophony.

So, for example:
dia samaw kaa - the soldier admires the actress
dia malak kaa - the mongoose fornicates with the zebra

Here, 'dia' can mean 'soldier' or 'mongoose'; 'kaa' can mean 'actress' or 'zebra'. How can you tell which? By looking at the verb! "samaw" means "admire (male-female human)", while "malak" means "fornicate (male-female non-human animal)". The verb tells you what class the noun is in, which in turn tells you which meaning of the noun is intended.

It's unlikely a human language would work like this, but everything there is attested in human languages. [It's rare for suppletion to go by the class of BOTH arguments - but it does happen in a minimal way with animacy (i.e. high>low vs. low>high actions may be distinguished suppletively).]

However, the big problem with this is that it's still a right bugger to learn. Until you learn what 'malak' means, you don't just not know what the mongoose did to the zebra, you don't even know that it's a mongoose and a zebra. You couldn't easily learn word-by-word - you'd need a lot of time to be able to learn to say anything. One way to help might be to give hints in the morphology. So, for instance m- might be an animal-class prefix, and 's-' might be a 'female object' class prefix. This wouldn't necessarily have to be systematic or completely predictable, but it would help learners considerably by letting them guess some words before they knew all of them.

OR: there could be a 'training wheels' system. Younger speakers might begin with vaguer verbs (like dia sman kaa - the soldier is aware of the actress) and gradually learn the more complicated ones...
User avatar
sangi39
moderator
moderator
Posts: 3024
Joined: 12 Aug 2010 01:53
Location: North Yorkshire, UK

Re: Kovur Grammar - Thoughts, Questions, Comments?

Post by sangi39 »

Salmoneus wrote: 02 Jun 2018 21:18 It would not be possible to have a language where basic transitive sentences were not composable. The number of possible agents, patients and actions is far too great - most 'situations' would never arise more than once in your lifetime, so it would be impossible to learn them.

That's the whole point about language: that with a finite, and hence learnable, set of subunits (words) we can construct an infinite number of superunits (clauses) to meet any situation. If your superunits, 'situations', cannot be decomposed intelligibly into a small number of learnable subunits, they will not be themselves learnable.

That said, it is possible to make your verbs more specific than in most human languages. For instance, in English we have progressively specific "he depicted the fox", "he marked out the outline of the fox", "he drew the outline of the fox", and "he pencilled the outline of the fox". English usually gets more specific by embedding more assumptions about manners and instruments, but other languages might do it differently. Some do do it at least partially with relevance to the arguments - so there may be suppletive verbs for "a man spoke" and "a woman spoke", for instance. But it's much rarer, I think. Maybe your alien language embeds arguments into verbs more frequently. It would only, however, be able to do it with relatively few arguments, because the total number of verbs would have to remain low enough to be learnable.

This is basically what's happening in Sangi's suggestion. Note, however, that it's no good just saying [the boy admired the girl][there was an actress][there was a soldier] - because how do you tell which is admiring which? Because already when you're doing that, you've got a verb (boy/girl-admires), and two nouns (there-is-an-actress) and (there-is-a-soldier) [NB if 'actress' only appears in the 'word' meaning 'there is an actress', then it doesn't mean 'there is an actress', it just means 'actress' - without the possibility of contrast, there is no distinction!] - and to make it intelligible you need to make one of those nouns an agent and one a patient.

So the result is not fundamentally different from English: subject, verb, object. That's the problem with language: it's pretty much impossible to escape that framework. Because you're trying to describe events, and what things were involved in them, and what role the involved things had.

You can change what words look like. Or how words relate to meanings. Or how words are structured into sentences syntactically. Or the pragmatics of what sort of things are talked about. But ultimatly you can't really change the semantic framework of agent, patient, noun, verb, because all languages are talking about the same sort of things.

That said, the system you and sangi suggest would be distinctive and noticeable weird by human standards. But bear in mind: it's weird because it's a language with many noun classes, and unusually high levels of verbal suppletion by noun class*, not because it does anything to challenge the noun/verb dichotomy.





*which raises the question: if you're having to have all these nouns anyway, what good does massive verbal suppletion do other than making the language harder to learn? Well, the obvious answer would be that it's counterbalanced by massive nominal homophony.

So, for example:
dia samaw kaa - the soldier admires the actress
dia malak kaa - the mongoose fornicates with the zebra

Here, 'dia' can mean 'soldier' or 'mongoose'; 'kaa' can mean 'actress' or 'zebra'. How can you tell which? By looking at the verb! "samaw" means "admire (male-female human)", while "malak" means "fornicate (male-female non-human animal)". The verb tells you what class the noun is in, which in turn tells you which meaning of the noun is intended.

It's unlikely a human language would work like this, but everything there is attested in human languages. [It's rare for suppletion to go by the class of BOTH arguments - but it does happen in a minimal way with animacy (i.e. high>low vs. low>high actions may be distinguished suppletively).]

However, the big problem with this is that it's still a right bugger to learn. Until you learn what 'malak' means, you don't just not know what the mongoose did to the zebra, you don't even know that it's a mongoose and a zebra. You couldn't easily learn word-by-word - you'd need a lot of time to be able to learn to say anything. One way to help might be to give hints in the morphology. So, for instance m- might be an animal-class prefix, and 's-' might be a 'female object' class prefix. This wouldn't necessarily have to be systematic or completely predictable, but it would help learners considerably by letting them guess some words before they knew all of them.

OR: there could be a 'training wheels' system. Younger speakers might begin with vaguer verbs (like dia sman kaa - the soldier is aware of the actress) and gradually learn the more complicated ones...
Going by extensions of "it's rare for suppletion to go by the class of BOTH arguments" are you suggesting that dividing verbs between egophoric vs. non egophoric or proximal vs. distal, is unreasonable?

I only ask, because in the systems I imagine for the various Kovur languages, there should be mechanisms through which an "inherently egophoric verb" can be expressed when the subject is 2nd person or 3rd person.
You can tell the same lie a thousand times,
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

Re: Kovur Grammar - Thoughts, Questions, Comments?

Post by Creyeditor »

Salmoneus wrote: 02 Jun 2018 21:18 It would not be possible to have a language where basic transitive sentences were not composable. The number of possible agents, patients and actions is far too great - most 'situations' would never arise more than once in your lifetime, so it would be impossible to learn them.

That's the whole point about language: that with a finite, and hence learnable, set of subunits (words) we can construct an infinite number of superunits (clauses) to meet any situation. If your superunits, 'situations', cannot be decomposed intelligibly into a small number of learnable subunits, they will not be themselves learnable.
Sorry, I was not clear enough. The communication system I thought of, does not necessarily result in "infinite number of superunits (clauses)". That was the idea. I should have mentioned that. Thefore the number of verbs might stay low. New situations might just not be easily describable.

Salmoneus wrote: 02 Jun 2018 21:18 That said, it is possible to make your verbs more specific than in most human languages. For instance, in English we have progressively specific "he depicted the fox", "he marked out the outline of the fox", "he drew the outline of the fox", and "he pencilled the outline of the fox". English usually gets more specific by embedding more assumptions about manners and instruments, but other languages might do it differently. Some do do it at least partially with relevance to the arguments - so there may be suppletive verbs for "a man spoke" and "a woman spoke", for instance. But it's much rarer, I think. Maybe your alien language embeds arguments into verbs more frequently. It would only, however, be able to do it with relatively few arguments, because the total number of verbs would have to remain low enough to be learnable.

This is basically what's happening in Sangi's suggestion. Note, however, that it's no good just saying [the boy admired the girl][there was an actress][there was a soldier] - because how do you tell which is admiring which? Because already when you're doing that, you've got a verb (boy/girl-admires), and two nouns (there-is-an-actress) and (there-is-a-soldier) [NB if 'actress' only appears in the 'word' meaning 'there is an actress', then it doesn't mean 'there is an actress', it just means 'actress' - without the possibility of contrast, there is no distinction!] - and to make it intelligible you need to make one of those nouns an agent and one a patient.
Okay, so let's be more specific about the 'nouns'. The way I though of it is the following: the speaker uses this staseme (I like that word), whenever (s)he sees an actress, that (s)he had not seen before (inside a reasonable context). You could also phrase this as 'an actress enters the stage' or something. It contrasts with other things an actress can do, like e.g. 'an actress vanishes'. The other idea would be that you use a specific AND if the the agent of some situation is the same as the argument of some other situation. Maybe there would also a specifc AND that indicates that the patient is the same as the argument of some other situation, and another one that indicates that all arguments are different. My idea would actually be that you cannot specify a marked agent and an marked patient in one utterance.

So [the boy admired the girl] ANDpatient [there was an actress] is okay, and [the boy admired the girl] ANDagent [there was a soldier], but [the boy admired the girl] ANDpatient [there was an actress] ANDagent [there was a soldier] is not okay.
Also, the direction (who is agent/patient) would be 'suppletive'. 'the boy admires the girl' is unrelated to 'the girl admires the boy', unless you introduce a kind of 'inverse' marker, in which case you might derive one from the other. Still, this would mean that the one that exists is an atomic situation.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: Kovur Grammar - Thoughts, Questions, Comments?

Post by Salmoneus »

Creyeditor wrote: 03 Jun 2018 11:22
Salmoneus wrote: 02 Jun 2018 21:18 It would not be possible to have a language where basic transitive sentences were not composable. The number of possible agents, patients and actions is far too great - most 'situations' would never arise more than once in your lifetime, so it would be impossible to learn them.

That's the whole point about language: that with a finite, and hence learnable, set of subunits (words) we can construct an infinite number of superunits (clauses) to meet any situation. If your superunits, 'situations', cannot be decomposed intelligibly into a small number of learnable subunits, they will not be themselves learnable.
Sorry, I was not clear enough. The communication system I thought of, does not necessarily result in "infinite number of superunits (clauses)". That was the idea. I should have mentioned that. Thefore the number of verbs might stay low. New situations might just not be easily describable.
But since the whole purpose of language is to describe new situations, that makes it all pointless.

Let's get some numbers into this. Let's start with a swadesh-size vocabulary. Around say 150 common nouns, and around 50 common verbs. We want to make 'statemes' out of these. How many statemes would we need? Well, at least 50, one for each verb. And then for each verb, we'd need a different version for each subject, and for each subject-verb we'd need a different version for each object. Leading to a vocabulary of 1.25 million stateme words. Not including the possibility of intransitives or solo nouns. You'd need 1.25 million words just to be able to converse at the level of the Swadesh list - 1.25 million words just to be able to talk about a tiny number of the most common, basic nouns engaged in the most tiny number of the most common, basic actions with one another. And have you tried having a conversation only in swadesh words? It's impossible. You'd need 1.25 million words and you'd still be at the linguistic level of a toddler. Virtually all situations would be "new" and "not easily describable". To have enough statemes to actually get through a conversation about daily events, you'd need literally trillions of statemes (as the number of participants and actions you want to describe increases, the number of suppletive stateme-forms you need increases exponentially).

[NB 'boy', 'girl' and 'admire' aren't in the swadesh list, for a start]

Which is impossible. The average person only uses a few thousand words in their daily life. They may passively understand, or occasionally use, a few tens of thousands. Most people won't encounter more than 100,000 words in their lifetime. So learning 1 million statemes, let alone 10 trillion, is going to be utterly, utterly impossible.


["boy" is #383 in frequency in English; "girl" is #364; "admire" is #3305. If you want your statemes to be able to specify 'boy' and 'girl' in them, you need at least 4 million statemes. If you want them to be able to specify admiration, you need at least 36 trillion stateme words. If you only want 100,000 statemes - a huge but theoretically possible number - you'd only be able to talk about the interactions between the, approximately, 45 most common things. That won't be enough to get you through breakfast.]
Okay, so let's be more specific about the 'nouns'. The way I though of it is the following: the speaker uses this staseme (I like that word), whenever (s)he sees an actress, that (s)he had not seen before (inside a reasonable context). You could also phrase this as 'an actress enters the stage' or something.
That's a long, long way, semantically, from admiration!
It contrasts with other things an actress can do, like e.g. 'an actress vanishes'. The other idea would be that you use a specific AND if the the agent of some situation is the same as the argument of some other situation. Maybe there would also a specifc AND that indicates that the patient is the same as the argument of some other situation, and another one that indicates that all arguments are different. My idea would actually be that you cannot specify a marked agent and an marked patient in one utterance.

So [the boy admired the girl] ANDpatient [there was an actress] is okay, and [the boy admired the girl] ANDagent [there was a soldier], but [the boy admired the girl] ANDpatient [there was an actress] ANDagent [there was a soldier] is not okay.
Let's rephrase that in normal terms. You're saying [admire] [actress-PAT] is OK, and [admire] [soldier-AGT] is OK, but not [admire][actress-PAT][soldier-AGT] (with the verb suppleting to show agreement with the noun class of the arguments). So basically you're just saying you only have intransitive verbs.

[again, glossing 'soldier' as 'there is a soldier' is just an unnecessarily confusing affectation]

The two questions then are: why? Why would your species be unable, over hundreds of thousands of years, to find a way to state both arguments in one sentence, given that they're able to state either of them? And second: then what? How DO they indicate transitives? Repetition of the verb? "Actress is admired and soldier admires"? Again, given that one of the verbs would be entirely superfluous, you'd have thought someone would have invented dropping it by now...
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

Re: Kovur Grammar - Thoughts, Questions, Comments?

Post by Creyeditor »

Salmoneus wrote: 03 Jun 2018 12:57 Let's get some numbers into this. Let's start with a swadesh-size vocabulary. Around say 150 common nouns, and around 50 common verbs. We want to make 'statemes' out of these. How many statemes would we need? Well, at least 50, one for each verb. And then for each verb, we'd need a different version for each subject, and for each subject-verb we'd need a different version for each object. Leading to a vocabulary of 1.25 million stateme words. Not including the possibility of intransitives or solo nouns. You'd need 1.25 million words just to be able to converse at the level of the Swadesh list - 1.25 million words just to be able to talk about a tiny number of the most common, basic nouns engaged in the most tiny number of the most common, basic actions with one another. And have you tried having a conversation only in swadesh words? It's impossible. You'd need 1.25 million words and you'd still be at the linguistic level of a toddler. Virtually all situations would be "new" and "not easily describable". To have enough statemes to actually get through a conversation about daily events, you'd need literally trillions of statemes (as the number of participants and actions you want to describe increases, the number of suppletive stateme-forms you need increases exponentially).
Just to be fair. If you take any triple of very common noun + very common verb + very common noun to form a sentence, you will often encounter nonsense, such as 'the man drinks the fish' (taking relatively high ranked Swadesh-list entries). Such a communication system would not need every singlie possible combination. People might use a certain number of words, that potentially leads to such an amount of situations, but realistically speaking, nobody describes a trillion situations in his everyday life.
Salmoneus wrote: 03 Jun 2018 12:57 Let's rephrase that in normal terms. You're saying [admire] [actress-PAT] is OK, and [admire] [soldier-AGT] is OK, but not [admire][actress-PAT][soldier-AGT] (with the verb suppleting to show agreement with the noun class of the arguments). So basically you're just saying you only have intransitive verbs.

[again, glossing 'soldier' as 'there is a soldier' is just an unnecessarily confusing affectation]

The two questions then are: why? Why would your species be unable, over hundreds of thousands of years, to find a way to state both arguments in one sentence, given that they're able to state either of them? And second: then what? How DO they indicate transitives? Repetition of the verb? "Actress is admired and soldier admires"? Again, given that one of the verbs would be entirely superfluous, you'd have thought someone would have invented dropping it by now...
First, it is important that the markers are not modifiers of the situation. [admire] [actress-PAT][soldier-AGT] would mean that the actress is admired (the patient of an admiration event and the argument of an actress-existence/whatever situation) and the soldier is the agent of the actress-existence situation (which does not make sense to me) and the only argument of an actor existence argument. It is important to note that all three situations are coordinated. There is no subordination or verb-argument relation.

The idea woud be that this form of communication derived from animal comuncation with a very restricted set of signals for a very restricted set of situations. At the next stage vocabulary was expanded (bounded by learnability, as you mentioned) and the modifaction and conjunction of situations was introduced. Situation repetition is -- at that stage -- probably the best way to indicate several arguments of a situation. I never claimed that this stage would be very stable, btw. Maybe it will eventually develop into a full fleged, human like, language system.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
User avatar
cedh
MVP
MVP
Posts: 386
Joined: 07 Sep 2011 22:25
Location: Tübingen, Germany
Contact:

Re: Kovur Grammar - Thoughts, Questions, Comments?

Post by cedh »

Creyeditor wrote: 03 Jun 2018 17:55
Salmoneus wrote: 03 Jun 2018 12:57 Let's get some numbers into this. Let's start with a swadesh-size vocabulary. Around say 150 common nouns, and around 50 common verbs. We want to make 'statemes' out of these. How many statemes would we need? Well, at least 50, one for each verb. And then for each verb, we'd need a different version for each subject, and for each subject-verb we'd need a different version for each object. Leading to a vocabulary of 1.25 million stateme words. Not including the possibility of intransitives or solo nouns. You'd need 1.25 million words just to be able to converse at the level of the Swadesh list - 1.25 million words just to be able to talk about a tiny number of the most common, basic nouns engaged in the most tiny number of the most common, basic actions with one another. And have you tried having a conversation only in swadesh words? It's impossible. You'd need 1.25 million words and you'd still be at the linguistic level of a toddler. Virtually all situations would be "new" and "not easily describable". To have enough statemes to actually get through a conversation about daily events, you'd need literally trillions of statemes (as the number of participants and actions you want to describe increases, the number of suppletive stateme-forms you need increases exponentially).
Just to be fair. If you take any triple of very common noun + very common verb + very common noun to form a sentence, you will often encounter nonsense, such as 'the man drinks the fish' (taking relatively high ranked Swadesh-list entries). Such a communication system would not need every singlie possible combination. People might use a certain number of words, that potentially leads to such an amount of situations, but realistically speaking, nobody describes a trillion situations in his everyday life.
Also, it seems to me that what Creyeditor and Sangi were talking about was verb suppletion by noun class, not by individual nouns. In human languages, there are usually less than about 10 noun classes (the higher numbers in Bantu are mostly due to singular vs. plural being counted separately). Let's suppose the Kovur typically distinguish a couple more, say, about 20 classes. That would bring down the maximum number of verbs needed to express the Swadesh list to about 20,000 if I've calculated correctly. Much more manageable (although it's probably still too much). And you wouldn't need any additional verbs to acommodate new nouns. Also, you could further take advantage of semantic restrictions - as Crey said, some types of agent > patient constellations are unlikely to occur, either in general (e.g. an inanimate thing acting on a human) or with specific types of action (e.g. a plant or a house as the subject of a motion verb or a perception verb), and might not ever get lexicalised. This might then lead to a restriction that entities belonging to certain noun classes couldn't ever be used in a certain role, because the language has no situation definitions where this role would be possible. In fact, this is an idea that's probably worth exploring! It's attested in human languages that certain inanimate nouns can't be transitive agents or so, but Kovur languages might take this a step further.

Another idea that could be used to keep the number of noun classes relevant for situation suppletion down: There could be secondary classifiers to be used within each situation set to distinguish between different types of referents that belong to the same major class. Since the Kovur are based on wolves, it'd make sense to base these finer distinctions on smell. This idea is not mine - Trailsend used smell-based noun classes in his werewolf language Feayran - but it's very obviously suitable here, and you can surely find a way to make it work differently than he did. An aspect I find interesting is that smell-based classification is not necessarily time-stable: the same person could, for instance, be tracked with different smell types depending on what they did before the current situation (e.g. rest vs. eat vs. run...) or depending on their emotional state (relaxed vs. stressed...)
User avatar
sangi39
moderator
moderator
Posts: 3024
Joined: 12 Aug 2010 01:53
Location: North Yorkshire, UK

Re: Kovur Grammar - Thoughts, Questions, Comments?

Post by sangi39 »

cedh wrote: 05 Jun 2018 09:32
Creyeditor wrote: 03 Jun 2018 17:55
Salmoneus wrote: 03 Jun 2018 12:57 Let's get some numbers into this. Let's start with a swadesh-size vocabulary. Around say 150 common nouns, and around 50 common verbs. We want to make 'statemes' out of these. How many statemes would we need? Well, at least 50, one for each verb. And then for each verb, we'd need a different version for each subject, and for each subject-verb we'd need a different version for each object. Leading to a vocabulary of 1.25 million stateme words. Not including the possibility of intransitives or solo nouns. You'd need 1.25 million words just to be able to converse at the level of the Swadesh list - 1.25 million words just to be able to talk about a tiny number of the most common, basic nouns engaged in the most tiny number of the most common, basic actions with one another. And have you tried having a conversation only in swadesh words? It's impossible. You'd need 1.25 million words and you'd still be at the linguistic level of a toddler. Virtually all situations would be "new" and "not easily describable". To have enough statemes to actually get through a conversation about daily events, you'd need literally trillions of statemes (as the number of participants and actions you want to describe increases, the number of suppletive stateme-forms you need increases exponentially).
Just to be fair. If you take any triple of very common noun + very common verb + very common noun to form a sentence, you will often encounter nonsense, such as 'the man drinks the fish' (taking relatively high ranked Swadesh-list entries). Such a communication system would not need every singlie possible combination. People might use a certain number of words, that potentially leads to such an amount of situations, but realistically speaking, nobody describes a trillion situations in his everyday life.
Also, it seems to me that what Creyeditor and Sangi were talking about was verb suppletion by noun class, not by individual nouns. In human languages, there are usually less than about 10 noun classes (the higher numbers in Bantu are mostly due to singular vs. plural being counted separately). Let's suppose the Kovur typically distinguish a couple more, say, about 20 classes. That would bring down the maximum number of verbs needed to express the Swadesh list to about 20,000 if I've calculated correctly. Much more manageable (although it's probably still too much). And you wouldn't need any additional verbs to acommodate new nouns. Also, you could further take advantage of semantic restrictions - as Crey said, some types of agent > patient constellations are unlikely to occur, either in general (e.g. an inanimate thing acting on a human) or with specific types of action (e.g. a plant or a house as the subject of a motion verb or a perception verb), and might not ever get lexicalised. This might then lead to a restriction that entities belonging to certain noun classes couldn't ever be used in a certain role, because the language has no situation definitions where this role would be possible. In fact, this is an idea that's probably worth exploring! It's attested in human languages that certain inanimate nouns can't be transitive agents or so, but Kovur languages might take this a step further.

Another idea that could be used to keep the number of noun classes relevant for situation suppletion down: There could be secondary classifiers to be used within each situation set to distinguish between different types of referents that belong to the same major class. Since the Kovur are based on wolves, it'd make sense to base these finer distinctions on smell. This idea is not mine - Trailsend used smell-based noun classes in his werewolf language Feayran - but it's very obviously suitable here, and you can surely find a way to make it work differently than he did. An aspect I find interesting is that smell-based classification is not necessarily time-stable: the same person could, for instance, be tracked with different smell types depending on what they did before the current situation (e.g. rest vs. eat vs. run...) or depending on their emotional state (relaxed vs. stressed...)
Well, my idea for the Kovur was literally just to have verbs divided into two classes universally, either based on "egophoricity" or "proximity":

1) "egophoric" (1st person) or "proximal" (1st person and 2nd person) verbs, which by default indicate that the agent (in the case of transitive verbs) is the speaker, or the speaker and/or listener, vs.
2) "non-egophoric" (2nd person and 3rd person) or "distal" (proximate 3rd person and obviative 3rd person) verbs, which by default indicate that the agent (in the case of transitive verbs) is not the speaker, but instead the listener and/or some other thing, or some other thing

In the case of an egophoric vs. non-egophoric constrast, a 2nd person agent would be indicated on non-egophoric verbs as "proximate" while a 3rd person agent would be indicated on non-egophoric verbs as "obviative". A 1st person agent would simply occur with egophoric verbs, which could be marked or unmarked depending on the language.

In the case of a proximal vs. distal constrast, a 1st person agent and proximate 3rd person agent would be marked as proximate on proximal and distal verbs respectively, while a 2nd person agent and obviative 3rd person agent would be marked as obviative on proximal and distal verbs respectively.

For a language with an egophoricity contrast to have, say a 1st person agent do something that is covered by a non-egophoric verb, they would need to employ some sort of mechanism, either syntactic or morphological, which would, for example, either convert the non-egophoric verb into a noun which could act as the patient for some egophoric verb, or the 1st person agent would have to be demoted some oblique case while the non-egophoric patient would be promoted to the subject, and perform some other action on the speaker represented by a non-egophoric verb.

So, for example, you might have the egophoric verb "go" and the non-egophoric verb "come", undergo the following:

a) An egophoric verb with an egophoric agent
go-EGO-FUT Jack-PAT
I will go to Mary

b) A non-egophoric verb with a non-egophoric agent
Jack-AGT come-DIST-FUT here-PAT
Jack will come to me (with "me" indicated by polysemous "here" that indicates both the speaker and the immediate surroundings)

Which are converted into:

c.i) An egophoric verb with a non-egophoric agent
Mary-AGT carry-DIST-FUT go-VN-PAT house-DAT steve-GEN
Mary will carry going to Steve's house
Mary will go to Steve's house

OR

c.ii)
house-AGT Steve-GEN take-DIST-FUT go-VN-PAT Mary-GEN
"Steve's house will take Mary's going"
Mary will go to Steve's house

and

d.i) A non-egophoric verb with an egophoric agent
give-EGO-FUT come-VN-PAT Jack-DAT
I will give coming to Jack
I will come to Jack

OR

d.ii)
Jack take-DIST-FUT come-VN-PROX
Jack will take my coming
I will come to Jack

In the last four examples, the original verb is converted into a noun which is then acted upon. In c.i and d.i the "intended" agent ("Mary" in c and "Jack" in d) remain as the syntactic agent and act upon the verbal noun. In c.ii and d.ii, however, the intended agent instead becomes the possessor of the verbal noun, which again appears as the patient, while the "intended" patient ("Steve's house" in c and "Jack" in d) are promoted to the syntactic agent and act upon the verbal noun. You can also see that the "intended" patient in c.i and d.i are treated as a third argument appearing in the dative instead of as the patient.





I can't speak for Creyeditor's conlang, or Sal's views on mine (he didn't answer my post so I don't know if he's addressing just Creyeditor or both Creyeditor and me), but I don't recall mentioning noun classes, just two distinct classes of verbs based on "egophoricity" or "proximity".
You can tell the same lie a thousand times,
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

Re: Kovur Grammar - Thoughts, Questions, Comments?

Post by Creyeditor »

sangi39 wrote: 05 Jun 2018 15:46I can't speak for Creyeditor's conlang, or Sal's views on mine (he didn't answer my post so I don't know if he's addressing just Creyeditor or both Creyeditor and me), but I don't recall mentioning noun classes, just two distinct classes of verbs based on "egophoricity" or "proximity".
You are right. Sorry for hijacking your thread.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
User avatar
sangi39
moderator
moderator
Posts: 3024
Joined: 12 Aug 2010 01:53
Location: North Yorkshire, UK

Re: Kovur Grammar - Thoughts, Questions, Comments?

Post by sangi39 »

Creyeditor wrote: 05 Jun 2018 20:02
sangi39 wrote: 05 Jun 2018 15:46I can't speak for Creyeditor's conlang, or Sal's views on mine (he didn't answer my post so I don't know if he's addressing just Creyeditor or both Creyeditor and me), but I don't recall mentioning noun classes, just two distinct classes of verbs based on "egophoricity" or "proximity".
You are right. Sorry for hijacking your thread.
Oh that wasn't my point at all [:)] If I had a problem with that I would have mentioned it earlier, but since we were talking about non-human syntax and I asked for any thoughts, and Sal had thoughts on those thoughts, it made sense to keep it going [:)]
You can tell the same lie a thousand times,
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
Post Reply