Short questions for the Omni-kan project

[Qwynegold]

One very important thing I would add is that languages very often have multiple complement types. For instance, the English sentence I wish [one was able [to turn back the clock]] has two different complement types: One was able to turn back the clock has the internal structure of a sentence, and could be used as a main clause, whereas to turn back the clock uses a different form lacking a subject, and thus cannot be used as a main clause. (I’ve seen the two types called balanced and deranked respectively, though I’m not sure how widely-used the terminology is.) In a language without deranked complement clauses — or in one where they are used in different places than English — that sentence might be rendered I wish one was able one turns back the clock. Also in many languages it is possible to delete the subject of a complement clause if equal to another argument (equi-NP deletion): I wish one was able turns back the clock. It’s worth thinking about exactly how many types of complement clauses Omni-Kan has, and what the semantic and syntactic distinctions are between them.

Ah yes. Omni-kan is usually pretty anal about this kind of stuff; because Asian languages often drop all kinds of stuff, and that makes speakers of European languages totally lost; so all sentences need to be constructed in a very clear way, so you know who is doing what. But I had decided that you can drop the subject in subordinate clauses when it's the same as in the main clause.

[Qwynegold]

So I wonder if anyone knows of other natlang suffixes used for male derivation. I would like it to come from one of the below languages, but I may consider other languages as well.

German has -rich. It's not terribly common and mainly used just for a few animals.

Thanks! I might change my male suffix to this, because right now I'm using -uire /uire/ from Latin, and I think people might have trouble pronouncing that.

[bradrn]

Note that poste mici puru saha "turn back the clock" is a subject complement here since it is describing the subject: mi "I". This might be part of what's tripping you up:

Aha. I had to reread that part in the WP article, because that's something I hadn't understood at all. So a subject complement is something that the subject is and becomes, while an object complement is something that the object becomes. And the complement co-occurs with the subject or object. This article only mentions subject/object complements that are predicative though. Are there other kinds of subject/object complement? (Disregarding what is says under the heading "As arguments", because there complement is just synonymous with argument.)

I hadn’t originally noticed this in Vardelm’s post, but this isn’t standard linguistic terminology. (Misleading terminology is a recurring problem with Wikipedia’s linguistic articles.) As far as I’m aware, a ‘subject complement’ is simply a complement clause in subject position (e.g. that Wikipedia misuses terminology bothers me), whereas an ‘object complement’ is one in object position (e.g. I am annoyed that Wikipedia misuses terminology). But even with these definitions, I haven’t really seen either term used with any regularity.

In that case, I will have to revise my earlier statement, which misread Vardelm:

Also, complements are supposed to take the object place in Omni-kan

Does that mean that subject complements are not allowed in Omni-kan? That is, you are not allowed to use a complement to describe the subject? I doubt that, but maybe there are languages that work this way.

Attested in Tariana, White Hmong and Panare. Even English doesn’t like subject complement clauses: we prefer it is possible that he is alive rather than that he is alive is possible.

What I meant to say is that Tariana, White Hmong and Panare disallow complement clauses in subject position; I wouldn’t know anything about restrictions in coreference between main and complement clauses.

[Vardelm]

Ah, I see. I had been thinking of only poste mici (back way) as the complement, and saha (clock) as a separate object.

Aahh-HAH! Yeah, that was probably tripping you up quite a bit.

Aha. I had to reread that part in the WP article, because that's something I hadn't understood at all. So a subject complement is something that the subject is and becomes, while an object complement is something that the object becomes. And the complement co-occurs with the subject or object.

That's my understanding yeah, but Bradrn mentions this:

I hadn’t originally noticed this in Vardelm’s post, but this isn’t standard linguistic terminology. (Misleading terminology is a recurring problem with Wikipedia’s linguistic articles.) As far as I’m aware, a ‘subject complement’ is simply a complement clause in subject position (e.g. that Wikipedia misuses terminology bothers me), whereas an ‘object complement’ is one in object position (e.g. I am annoyed that Wikipedia misuses terminology). But even with these definitions, I haven’t really seen either term used with any regularity.

How sure about this are you? A very quick Google search of "subject complement" shows results that agree with the Wikipedia article. Granted, those aren't academic linguistics articles, but it tells me that Wikipedia's explanations aren't created in a vacuum. To me, having the definitions mean describing the subject or object rather than what position they are in makes more sense.

This article only mentions subject/object complements that are predicative though. Are there other kinds of subject/object complement? (Disregarding what is says under the heading "As arguments", because there complement is just synonymous with argument.)

No idea on this. I'm sure even Bradrn would have a hard time researching & finding articles that shed light on the topic.

(Reverse psychology research now in session... )

Yes, thanks for the pedagogical explanations.

My pleasure and glad it helps! Writing something like this actually helps me to get some clarity on a topic as well. I have looked into complements, relative clauses, etc. before, but haven't thought quite this much about complements. Even so, there's plenty more to learn.

(Btw, I don't think there are much more "syntactically experted people".)

Thanks, but I'd point to Zompist (he's written a book, fer cryin' out loud!) and pretty much anyone who has looked into syntactic trees, transformations, etc. I tend to think of syntax in a fairly simple, mathematical kind of way. It's just simple formulas and then making substitutions for various portions of that. I haven't gotten in much trouble with that approach yet, and maybe that's all the formalisms are saying.


----------------------------------------------

What I meant to say is that Tariana, White Hmong and Panare disallow complement clauses in subject position; I wouldn’t know anything about restrictions in coreference between main and complement clauses.

This makes sense to me. I have a hard time believing languages would not allow subjects or objects to be described, but having syntax that doesn't allow them in certain positions seems quite likely.



----------------------------------------------

Ah, you've been working on a big thesis until now? Congratulations! It must be a big relief to have that over with.

Yep, January until July 9th, when I defended. My brain is enjoying the rest!

[bradrn]

I hadn’t originally noticed this in Vardelm’s post, but this isn’t standard linguistic terminology. (Misleading terminology is a recurring problem with Wikipedia’s linguistic articles.) As far as I’m aware, a ‘subject complement’ is simply a complement clause in subject position (e.g. that Wikipedia misuses terminology bothers me), whereas an ‘object complement’ is one in object position (e.g. I am annoyed that Wikipedia misuses terminology). But even with these definitions, I haven’t really seen either term used with any regularity.

How sure about this are you? A very quick Google search of "subject complement" shows results that agree with the Wikipedia article. Granted, those aren't academic linguistics articles, but it tells me that Wikipedia's explanations aren't created in a vacuum. To me, having the definitions mean describing the subject or object rather than what position they are in makes more sense.

I must admit, I’m not particularly sure; complement clauses aren’t my strength. All that I know is that I can’t recall those terms being used that way in anything I’ve read. Could you link to some other sources please?

This article only mentions subject/object complements that are predicative though. Are there other kinds of subject/object complement? (Disregarding what is says under the heading "As arguments", because there complement is just synonymous with argument.)

No idea on this. I'm sure even Bradrn would have a hard time researching & finding articles that shed light on the topic.

(Reverse psychology research now in session... )

Y’know, it almost worked on me until I noticed the little message…

(Btw, I don't think there are much more "syntactically experted people".)

Thanks, but I'd point to Zompist (he's written a book, fer cryin' out loud!) and pretty much anyone who has looked into syntactic trees, transformations, etc. I tend to think of syntax in a fairly simple, mathematical kind of way. It's just simple formulas and then making substitutions for various portions of that. I haven't gotten in much trouble with that approach yet, and maybe that's all the formalisms are saying.

This is how I think about syntax as well. I hate it when people start to drag out 7-level syntax trees and X-bars and various unintelligible acronyms for a sentence which, when all is said and done, is only seven words long and can be understood by any child. (Yes, I took that example from a real paper.)

[Vardelm]

I must admit, I’m not particularly sure; complement clauses aren’t my strength. All that I know is that I can’t recall those terms being used that way in anything I’ve read. Could you link to some other sources please?

https://en.wikipedia.org/wiki/Subject_complement
https://dictionary.cambridge.org/us/gra ... omplements
https://www.grammar-monster.com/glossar ... lement.htm

This is how I think about syntax as well. I hate it when people start to drag out 7-level syntax trees and X-bars and various unintelligible acronyms for a sentence which, when all is said and done, is only seven words long and can be understood by any child. (Yes, I took that example from a real paper.)

Excellent. Glad I'm not alone on that.

Y’know, it almost worked on me until I noticed the little message…

This will simply inform the methodology for my next, evil experiment!!!

[bradrn]

(Sorry for the delay; I completely forgot to reply to your post…)

I must admit, I’m not particularly sure; complement clauses aren’t my strength. All that I know is that I can’t recall those terms being used that way in anything I’ve read. Could you link to some other sources please?

https://en.wikipedia.org/wiki/Subject_complement
https://dictionary.cambridge.org/us/gra ... omplements
https://www.grammar-monster.com/glossar ... lement.htm

Ah, I should have specified academic sources. Your links use traditional English grammar, which basically ends up analysing English as a warped form of Latin. For this reason, terms from traditional grammar are always a bit suspect, and this one seems more incoherent than most. For instance, your Cambridge link lists (1) as a subject complement, but (2) as an object:

(1) He became a famous writer.
(2) He married a famous writer.

And sure, there is a semantic difference, and this might translate to a syntactic difference in some other language… but the two constructions are exactly parallel in English, and there is no grounds for distinguishing them syntactically. Furthermore, not a single one of their ‘subject complements’ are, in fact, a complement clause. For that matter, neither is their example of an ‘object complement’:

He makes me very angry.

Nor are the examples of ‘subject complements’ on Wikipedia — their ‘dependent clause’ is actually a headless relative. All of which just goes to confirm my opinion that most of ‘traditional grammar’ is rubbish.

[Vardelm]

(Sorry for the delay; I completely forgot to reply to your post…)

Not a problem, good sir!


Given this:

I hadn’t originally noticed this in Vardelm’s post, but this isn’t standard linguistic terminology. (Misleading terminology is a recurring problem with Wikipedia’s linguistic articles.) As far as I’m aware, a ‘subject complement’ is simply a complement clause in subject position (e.g. that Wikipedia misuses terminology bothers me), whereas an ‘object complement’ is one in object position (e.g. I am annoyed that Wikipedia misuses terminology). But even with these definitions, I haven’t really seen either term used with any regularity.

(Emphasis is mine.)

...plus this:

Ah, I should have specified academic sources. Your links use traditional English grammar, which basically ends up analysing English as a warped form of Latin. For this reason, terms from traditional grammar are always a bit suspect, and this one seems more incoherent than most.

I have to ask, if academic sources aren't even clear or uniform on this, why do we need to require academic sources for this? Yeah, I get the notion of academic vs non-academic sources & their relative value; I just completed grad school, after all! However, we're not engaged in academic writing here, so I would put value on a source more for its general usefulness than academic merits. Generally, those aren't at odds, but if the academics can't agree on something, I'm not going to worry about whether a description I write will hold up in a peer-reviewed journal.

To that end:

For instance, your Cambridge link lists (1) as a subject complement, but (2) as an object:

(1) He became a famous writer.
(2) He married a famous writer.

And sure, there is a semantic difference, and this might translate to a syntactic difference in some other language… but the two constructions are exactly parallel in English, and there is no grounds for distinguishing them syntactically.

Do the main predicate verbs involved (become and marry) not indicate a syntactic difference, in that "become" requires a complement, while "marry" allows/requires an object? Why is a syntactic difference even needed here? Are we not allowed to use the descriptors "subject complement" and "object complement" based simply on semantic properties? I mean, is semantics not a scholarly topic and thus doesn't warrant labels? (Yes, I'm exaggerating for effect here.)

My point about "usefulness" above is that here Wikipedia's description gave a very simple, clear, & easily applicable terminology to explain what's going on. "Subject compliments" describe the subject, while "object compliments" do so for the object. Qwynegold's original question was about complements and objects. It was useful to think in terms of complements as a whole clause/phrase, and that they describe the subject or object. That helped clarify that just because a complement may contain an object, that doesn't mean it's the object for the main predicate verb.

With that said:

Furthermore, not a single one of their ‘subject complements’ are, in fact, a complement clause. For that matter, neither is their example of an ‘object complement’:

He makes me very angry.

Nor are the examples of ‘subject complements’ on Wikipedia — their ‘dependent clause’ is actually a headless relative.

Maybe this SIL article agrees more with your understanding of relatives vs. complements?

https://glossary.sil.org/term/complement-clause

That article is written rather poorly such that it doesn't clearly distinguish between complements & relatives. It does a bit, but it requires re-reading a couple times. It does mention the irregular usage by academics between relative clauses & complement clauses. I think that shows the distinction between them is a tiny bit murky, and so for people who are new to or struggling with such concepts, it's not absolutely necessary for our purposes. Or, at least at Qwynegold's current stage of developing Omni-kan, it's not necessary. You basically go back to our previous posts and the Wikipedia article and replace "complement" with "relative" if you want, or maybe use them interchangeably. Academically correct? Nope; don't care too much at this point. Instead, it's for clarifying the issue at hand: a main predicate's object vs an object in a complement.

All of which just goes to confirm my opinion that most of ‘traditional grammar’ is rubbish.

Bob? Is that you?

[bradrn]

I have to ask, if academic sources aren't even clear or uniform on this, why do we need to require academic sources for this? Yeah, I get the notion of academic vs non-academic sources & their relative value; I just completed grad school, after all! However, we're not engaged in academic writing here, so I would put value on a source more for its general usefulness than academic merits. Generally, those aren't at odds, but if the academics can't agree on something, I'm not going to worry about whether a description I write will hold up in a peer-reviewed journal.

Mostly, I want these sources because academic linguists tend to be rather good at analysing the structure of language and postulating reasonable categories. Non-linguists, not so much.

For instance, your Cambridge link lists (1) as a subject complement, but (2) as an object:

(1) He became a famous writer.
(2) He married a famous writer.

And sure, there is a semantic difference, and this might translate to a syntactic difference in some other language… but the two constructions are exactly parallel in English, and there is no grounds for distinguishing them syntactically.

Do the main predicate verbs involved (become and marry) not indicate a syntactic difference, in that "become" requires a complement, while "marry" allows/requires an object? Why is a syntactic difference even needed here? Are we not allowed to use the descriptors "subject complement" and "object complement" based simply on semantic properties? I mean, is semantics not a scholarly topic and thus doesn't warrant labels? (Yes, I'm exaggerating for effect here.)

Well, this was originally a question about syntax. But you’re right that that’s not enough of a reason to reject it. More importantly, it just strikes me as a very arbitrary and pointless category to posit. Do ‘subject complements’ share anything in common beyond the definition itself? Does distinguishing ‘subject complements’ from non-‘subject complements’ enlighten us in any way? Does the distinction hold up outside English? If not, it’s a bad category.

I like how DeLancey (2004) puts it:

The question is, when we have a workshop on "ergativity" in various languages, are we pursuing a coherent typological phenomenon, or just collecting specimens? Not that there’s anything wrong with collecting and comparing specimens … But when we limit a collection to certain kinds of specimens, there’s a question whether a workshop on "ergativity" is analogous to an effort to collect, say, birds with talons -- an important taxonomic criterion --, birds that swim -- which is taxonomically only marginally relevant, but a very significant functional pattern --, or, say, birds that are blue, which will turn out to be pretty much a useless criterion for any biological purpose.

My point about "usefulness" above is that here Wikipedia's description gave a very simple, clear, & easily applicable terminology to explain what's going on. "Subject compliments" describe the subject, while "object compliments" do so for the object. Qwynegold's original question was about complements and objects. It was useful to think in terms of complements as a whole clause/phrase, and that they describe the subject or object. That helped clarify that just because a complement may contain an object, that doesn't mean it's the object for the main predicate verb.

The problem I have with this is that the word ‘complement’ is clearly being used for two completely different things here. Qwynegold’s question about ‘complements’ referred to subordinate clauses embedded in a larger clause as an argument of a verb. ‘Subject’ and ‘object complements’, by contrast, are merely specific types of what us linguists would call noun phrases. There is practically no connection.

Furthermore, not a single one of their ‘subject complements’ are, in fact, a complement clause. For that matter, neither is their example of an ‘object complement’:

He makes me very angry.

Nor are the examples of ‘subject complements’ on Wikipedia — their ‘dependent clause’ is actually a headless relative.

Maybe this SIL article agrees more with your understanding of relatives vs. complements?

https://glossary.sil.org/term/complement-clause

That SIL article agrees with me, in that a complement clause ‘is a notional sentence … that is an argument of a predicate’. (Though their third example is a headless relative, not a complement clause. R.M.W. Dixon calls such things ‘complementation strategies’, though the term sadly is not yet in wide used.)

That article is written rather poorly such that it doesn't clearly distinguish between complements & relatives. It does a bit, but it requires re-reading a couple times. It does mention the irregular usage by academics between relative clauses & complement clauses. I think that shows the distinction between them is a tiny bit murky, and so for people who are new to or struggling with such concepts, it's not absolutely necessary for our purposes. Or, at least at Qwynegold's current stage of developing Omni-kan, it's not necessary. You basically go back to our previous posts and the Wikipedia article and replace "complement" with "relative" if you want, or maybe use them interchangeably. Academically correct? Nope; don't care too much at this point. Instead, it's for clarifying the issue at hand: a main predicate's object vs an object in a complement.

Sorry, but I’m totally lost here… which article do you mean by ‘that article’? What ‘irregular usage’? How is ‘academically correct’ different from just, well, plain ‘correct’?

All of which just goes to confirm my opinion that most of ‘traditional grammar’ is rubbish.

Bob? Is that you?

Not at all. As far as I’m aware, I’m still bradrn. (Unless someone has taken over my mind without me knowing, of course.)

I should note it’s important to differentiate between ‘traditional grammar’ and the formalisms of modern linguistics—‘Basic Linguistic Theory’, as some people are starting to call it. The former was originally created for Latin by native speakers and scholars, and was then (mis)applied to most European languages in general and English in particular. The problem is, Latin syntax is quite different to English syntax, so various parts of English syntax ended up being shoehorned into bits of Latin grammar which seemed to express approximately the same thing. This is why terms like ‘adverb’ and ‘infinitive’ and, yes, ‘subject complement’ have ended up so ambiguous and difficult to understand.

(By the way, it’s also where rules like ‘no prepositions at the end of a sentence’ come from. In Latin, such constructions are ungrammatical. Ergo they must of course be ungrammatical in English as well! ‘Logic is a wonderful thing but doesn’t always beat actual thought’—Terry Pratchett.)

[Vardelm]

Mostly, I want these sources because academic linguists tend to be rather good at analysing the structure of language and postulating reasonable categories. Non-linguists, not so much.

Well, this was originally a question about syntax. But you’re right that that’s not enough of a reason to reject it. More importantly, it just strikes me as a very arbitrary and pointless category to posit. Do ‘subject complements’ share anything in common beyond the definition itself? Does distinguishing ‘subject complements’ from non-‘subject complements’ enlighten us in any way? Does the distinction hold up outside English? If not, it’s a bad category.

The problem I have with this is that the word ‘complement’ is clearly being used for two completely different things here. Qwynegold’s question about ‘complements’ referred to subordinate clauses embedded in a larger clause as an argument of a verb. ‘Subject’ and ‘object complements’, by contrast, are merely specific types of what us linguists would call noun phrases. There is practically no connection.

Furthermore, not a single one of their ‘subject complements’ are, in fact, a complement clause. For that matter, neither is their example of an ‘object complement’:

Generally, I'd agree that academic linguists are better than non-linguists (duh!). I hope it's obvious that I'm not against academics. But in this case, it seems that they haven't been great at postulating categories, or at least they haven't been great at agreeing on the categories for complements, etc. By your own admission, there is some debate & disagreement about this among academics.

Here's the thing: I'm not interested - at all - in debating linguistic theory. I think it would be a fun pursuit & career, but it's not what I've chosen. I'm interested in using linquistics in an artistic endeavor, as are most of us here. If academic linguists don't have a consensus on terminology & behavior for complements or whatever, then that's unfortunate, but I don't really care. In that light, if you can point to any articles that do explain complements, relative clauses, etc. etc. in a fairly straightforward manner that a majority - or even just a plurality! - of academic linguists would agree on, I'm happy to look. Until then, for a discussion about a conlang that requires only a very basic impression about complements, etc., I'm quite happy to look at a source like Wikipedia, even with its faults, and use it until the academics have enough consensus to provide information useful outside of academic bickering.

Sorry, but I’m totally lost here… which article do you mean by ‘that article’?

The article I linked: https://glossary.sil.org/term/complement-clause.

What ‘irregular usage’?

From the link:

The term complement clause is extended by some analysts to include clauses selected by nouns or adjectives.

How is ‘academically correct’ different from just, well, plain ‘correct’?

Just a higher standard for sourcing. Academics won't (and shouldn't!) use Wikipedia; as a conlanger I most certainly will.

Not at all. As far as I’m aware, I’m still bradrn. (Unless someone has taken over my mind without me knowing, of course.)

Oh good!

[bradrn]

Mostly, I want these sources because academic linguists tend to be rather good at analysing the structure of language and postulating reasonable categories. Non-linguists, not so much.

Well, this was originally a question about syntax. But you’re right that that’s not enough of a reason to reject it. More importantly, it just strikes me as a very arbitrary and pointless category to posit. Do ‘subject complements’ share anything in common beyond the definition itself? Does distinguishing ‘subject complements’ from non-‘subject complements’ enlighten us in any way? Does the distinction hold up outside English? If not, it’s a bad category.

The problem I have with this is that the word ‘complement’ is clearly being used for two completely different things here. Qwynegold’s question about ‘complements’ referred to subordinate clauses embedded in a larger clause as an argument of a verb. ‘Subject’ and ‘object complements’, by contrast, are merely specific types of what us linguists would call noun phrases. There is practically no connection.

Furthermore, not a single one of their ‘subject complements’ are, in fact, a complement clause. For that matter, neither is their example of an ‘object complement’:

Generally, I'd agree that academic linguists are better than non-linguists (duh!). I hope it's obvious that I'm not against academics. But in this case, it seems that they haven't been great at postulating categories, or at least they haven't been great at agreeing on the categories for complements, etc. By your own admission, there is some debate & disagreement about this among academics.

Well, yes, there is still some disagreement about this. But that’s only because getting two linguists to agree with each other is more or less a hopeless task. Linguists are by nature about as cooperative as sharks. (Even getting a linguist to agree with themself can be a bit difficult at times.)

That being said, there does seem to be a broad consensus on what sort of parameters are useful or not useful for classifying complement clauses. The key one appears to be deranking/finiteness (the two appear to be synonyms AFAICT); other important ones appear to be the relationship of the subordinated verb to its arguments, and the temporal connection between the main and complement clause.

(Actually, after some reflection, I must apologise for the comparison of linguists to sharks. I looked it up, and sharks are far more cooperative than I thought.)

Here's the thing: I'm not interested - at all - in debating linguistic theory. I think it would be a fun pursuit & career, but it's not what I've chosen. I'm interested in using linquistics in an artistic endeavor, as are most of us here. If academic linguists don't have a consensus on terminology & behavior for complements or whatever, then that's unfortunate, but I don't really care.

Personally, I study linguistic theory in large part because it helps me make better and more interesting conlangs. It also helps me explain ideas and analyse language more clearly. Bad linguistic theories do the opposite — if you use an unhelpful term, it doesn’t, well, help me understand what you’re saying.

In that light, if you can point to any articles that do explain complements, relative clauses, etc. etc. in a fairly straightforward manner that a majority - or even just a plurality! - of academic linguists would agree on, I'm happy to look.

Michael Noonan’s chapter on Complementation (from Language Typology and Syntactic Description) is pretty good, if you can get access. Don’t know about ‘agree’, but at least R.M.W. Dixon has referred to it favourably.

What ‘irregular usage’?

From the link:

The term complement clause is extended by some analysts to include clauses selected by nouns or adjectives.

Ah, yes. Noonan spends a whole section (well, 1.5 pages) talking about this. The article is referring to cases such as the fact that Joe is gone, which has something formally identical to a complement clause, but attached to a noun rather than a verb. The problem here is that this is a genuinely in-between case, so of course there will be disagreement over which category to include it in.

How is ‘academically correct’ different from just, well, plain ‘correct’?

Just a higher standard for sourcing. Academics won't (and shouldn't!) use Wikipedia; as a conlanger I most certainly will.

No, no, no, no, no! Wikipedia’s linguistic articles are terrible. I spent a good portion of my first few years of conlanging reading Wikipedia and getting horribly confused, until I realised that academic sources were not just more sensible but in many cases also easier to understand. A good typological volume — like Payne’s Describing Morphosyntax, or Dixon’s Basic Linguistic Theory, or ed. Shopen’s Language Typology and Syntactic Description — will cover far more content, far better than Wikipedia does.

[Qwynegold]

Note that poste mici puru saha "turn back the clock" is a subject complement here since it is describing the subject: mi "I". This might be part of what's tripping you up:

Aha. I had to reread that part in the WP article, because that's something I hadn't understood at all. So a subject complement is something that the subject is and becomes, while an object complement is something that the object becomes. And the complement co-occurs with the subject or object. This article only mentions subject/object complements that are predicative though. Are there other kinds of subject/object complement? (Disregarding what is says under the heading "As arguments", because there complement is just synonymous with argument.)

I hadn’t originally noticed this in Vardelm’s post, but this isn’t standard linguistic terminology.

Wait, I had also overlooked this. How is "turn back the clock" describing the subject?

[Qwynegold]

(Btw, I don't think there are much more "syntactically experted people".)

Thanks, but I'd point to Zompist (he's written a book, fer cryin' out loud!) and pretty much anyone who has looked into syntactic trees, transformations, etc.

I have actually taken a course in syntax, but I can't make any sense out of it. *shrug*

[Qwynegold]

Anyway, so it's clear now that adverbials always precede the thing they modify. Based on this, is there anything one can say about the order of prepositional phrases? I'm thinking specifically about prepositional phrases that describe location. I'll give a couple of example sentences.

Hmm, it might actually be fun to see if it's possible for others to construct a sentence in Omni-kan with minimal information. So if anyone wants to try this, I have provided English sentences dealing with location, and word lists, below. Try to translate these sentences into Omni-kan with help of the word list and any other information you may have found in this thread. I'll post my answers later.

1) They sink down like a duvet over the city.
Word list:

More: show

city - siti
duvet - korepiye
like - lo (ADVZ used at the end of the adverbial phrase)
of - te (the order is "possessee GEN possessor")
sink - tegkelam
they - okal
to - co
top - supere (N.)

2) I am lost here.
Word list:

More: show

at - ni
become - ti (V. or passive marker)
finish - pan (V. or perfective marker)
here - es lokus (lit. "this place")
I - mi
lose - ilta (V.)

[bradrn]

Anyway, so it's clear now that adverbials always precede the thing they modify.

Sorry, what? I think you’re using a different definition of ‘adverbial’ to me. As far as I’m aware, the term ‘adverbial’ (which does for once have a consistent usage) has always referred to a word or phrase which acts to modify a word or sentence, for instance: (Thompson & Longacre & Hwang 2007)

She mailed it yesterday
I get up when the sun rises
Carry this as I told you to
Having told a few bad jokes, Harvey proceeded to introduce the speaker

As you can see, only one of these examples precedes its modifier.

Based on this, is there anything one can say about the order of prepositional phrases? I'm thinking specifically about prepositional phrases that describe location. I'll give a couple of example sentences.

I think you forgot the example sentences here.

Hmm, it might actually be fun to see if it's possible for others to construct a sentence in Omni-kan with minimal information. So if anyone wants to try this, I have provided English sentences dealing with location, and word lists, below. Try to translate these sentences into Omni-kan with help of the word list and any other information you may have found in this thread. I'll post my answers later.

1) They sink down like a duvet over the city.
Word list:

More: show

city - siti
duvet - korepiye
like - lo (ADVZ used at the end of the adverbial phrase)
of - te (the order is "possessee GEN possessor")
sink - tegkelam
they - okal
to - co
top - supere (N.)

2) I am lost here.
Word list:

More: show

at - ni
become - ti (V. or passive marker)
finish - pan (V. or perfective marker)
here - es lokus (lit. "this place")
I - mi
lose - ilta (V.)

More: show

Okal

they

tegkelam

sink

co

to

siti

city

supere

top

lo

like

korepiye.

duvet

(Notes: I’m assuming possession is expressed via juxtaposition: siti supere = ‘top of the city’. Not sure how to position lo, especially given your note ‘used at the end of the adverbial phrase’, but used as a similative, head-initial seems more likely to me on the basis of the rest of the language.)

Es

this

lokus

place

mi

I

ti

become

ilta.

lose

(Notes: I’m guessing it’s fine to place the locational adverbial at the beginning of the sentence. Not sure how I’m supposed to use the perfective pan; I assume that it’s unneeded in this sentence, given that it’s more stative i.e. imperfective.)

[Qwynegold]

As you can see, only one of these examples precedes its modifier.

I mean, in Omni-kan they precede.

Based on this, is there anything one can say about the order of prepositional phrases? I'm thinking specifically about prepositional phrases that describe location. I'll give a couple of example sentences.

I think you forgot the example sentences here.

Those were the sentences 1) and 2) further down.

Okal

they

tegkelam

sink

co

to

siti

city

supere

top

lo

like

korepiye.

duvet

(Notes: I’m assuming possession is expressed via juxtaposition: siti supere = ‘top of the city’. Not sure how to position lo, especially given your note ‘used at the end of the adverbial phrase’, but used as a similative, head-initial seems more likely to me on the basis of the rest of the language.)

Here's my answer for 1)

More: show

Okal

they

korepiye

duvet

lo

like

tegkelam

sink

co

to

supere

top

te

GEN

siti.

city

I put "duvet like" before the verb, because it's modifying the verb.

Lo I had borrowed from Burmese where it's used in final position. I use it as an adjective that things are compounded to. Compound words are written apart and the head comes last. But I had recently been thinking, would it be better if lo was instead a preposition? Because then it would notify the reader/listener that the following word(s) is/are part of an adjective phrase. When lo is final the reader/listener may be lead to momentary garden paths. But if it was a preposition there's still the problem that you don't know where the adjective phrase ends.

Es

this

lokus

place

mi

I

ti

become

ilta.

lose

(Notes: I’m guessing it’s fine to place the locational adverbial at the beginning of the sentence. Not sure how I’m supposed to use the perfective pan; I assume that it’s unneeded in this sentence, given that it’s more stative i.e. imperfective.)

Here's my answer to 2)

More: show

Mi

1SG

pan

PFV

ti

become

ilta

lose

ni

at

es

this

lokus.

place

Hmm yeah, I'm also not sure if pan is really necessary here. I'm thinking like it helps make it clear that this is not a future action nor is it something that always happen when the person goes to some specific place. Hmm, I wasn't thinking of this action as imperfective at all. I was thinking that getting lost is an instantaneous action, and so because pan means "finish", when I have finished getting lost I am still lost.

I placed the location after the verb, because I had been treating locations as verb complements. You also had the intuition to put it after the verb in 1). So I need to make a modification to the rules and say that locational adverbials go after the verb. Hmm, I feel like locational adverbials are not the same thing as adverbials of manner. Because the manner ones are clearly modifying the verb, but to me it seems like the locational ones are not modifying anything at all.

[bradrn]

Okal

they

tegkelam

sink

co

to

siti

city

supere

top

lo

like

korepiye.

duvet

(Notes: I’m assuming possession is expressed via juxtaposition: siti supere = ‘top of the city’. Not sure how to position lo, especially given your note ‘used at the end of the adverbial phrase’, but used as a similative, head-initial seems more likely to me on the basis of the rest of the language.)

Here's my answer for 1)

More: show

Okal

they

korepiye

duvet

lo

like

tegkelam

sink

co

to

supere

top

te

GEN

siti.

city

I put "duvet like" before the verb, because it's modifying the verb.

Lo I had borrowed from Burmese where it's used in final position. I use it as an adjective that things are compounded to. Compound words are written apart and the head comes last. But I had recently been thinking, would it be better if lo was instead a preposition? Because then it would notify the reader/listener that the following word(s) is/are part of an adjective phrase. When lo is final the reader/listener may be lead to momentary garden paths. But if it was a preposition there's still the problem that you don't know where the adjective phrase ends.

I think it’s best to be consistent in an auxlang — either make everything head-final, or make everything head-initial, but I don’t think it’s a good idea to do a mixture. I don’t think any one approach is necessarily better than any other.

(Or you could even put one particle at the beginning and another one on the end, but that can become difficult to keep track of for speakers.)

Es

this

lokus

place

mi

I

ti

become

ilta.

lose

(Notes: I’m guessing it’s fine to place the locational adverbial at the beginning of the sentence. Not sure how I’m supposed to use the perfective pan; I assume that it’s unneeded in this sentence, given that it’s more stative i.e. imperfective.)

Here's my answer to 2)

More: show

Mi

1SG

pan

PFV

ti

become

ilta

lose

ni

at

es

this

lokus.

place

Hmm yeah, I'm also not sure if pan is really necessary here. I'm thinking like it helps make it clear that this is not a future action nor is it something that always happen when the person goes to some specific place.

From this description, pan doesn’t sound at all like a perfective to me. Future actions are very often perfective, or at least marked using a prototypically ‘perfective’ aspect — e.g. English ‘The bus arrives tomorrow’ is morphologically ‘present simple’ (=perfective), but semantically future. (The combination of perfective+present tends to do weird things, because an entire event seen as a whole doesn’t really ‘fit’ into a single present moment.) Meanwhile, an event which ‘always happens’ is habitual: technically imperfective, according to Comrie, but a very non-prototypical imperfective. If your pan specifically excludes those meanings, then to me it sounds more like a continuative aspect than anything else.

Hmm, I wasn't thinking of this action as imperfective at all. I was thinking that getting lost is an instantaneous action, and so because pan means "finish", when I have finished getting lost I am still lost.

Being lost is a State. Getting lost is an Accomplishment (i.e. telic durative). This is a difference in aktionsart, not aspect — though it might restrict the allowed aspects for these words, in that ‘being lost’ is less compatible with perfective or progressive aspect than is ‘getting lost’.

I placed the location after the verb, because I had been treating locations as verb complements. You also had the intuition to put it after the verb in 1). So I need to make a modification to the rules and say that locational adverbials go after the verb. Hmm, I feel like locational adverbials are not the same thing as adverbials of manner. Because the manner ones are clearly modifying the verb, but to me it seems like the locational ones are not modifying anything at all.

Really? I have a strong intuition that locational adverbs are acting as verbal modifiers.

(Some people distinguish verbal and sentential modifiers. Personally, I’ve never been able to see the difference.)

[Qwynegold]

I think it’s best to be consistent in an auxlang — either make everything head-final, or make everything head-initial, but I don’t think it’s a good idea to do a mixture. I don’t think any one approach is necessarily better than any other.

I agree that it would be more logical if everything was 100% a certain way. But I think that doesn't necessarily mean it's going to be easier to learn for most speakers. Anyway, the ordering for compounds is already set at this point, for reasons I don't quite remember rn... The other day I thought of some other reason why it would actually be better to have lo last, but now I can't remember it anymore. Ugh, this is one of those conlanging nightmares that never gets solved.

From this description, pan doesn’t sound at all like a perfective to me. Future actions are very often perfective, or at least marked using a prototypically ‘perfective’ aspect — e.g. English ‘The bus arrives tomorrow’ is morphologically ‘present simple’ (=perfective), but semantically future. (The combination of perfective+present tends to do weird things, because an entire event seen as a whole doesn’t really ‘fit’ into a single present moment.) Meanwhile, an event which ‘always happens’ is habitual: technically imperfective, according to Comrie, but a very non-prototypical imperfective. If your pan specifically excludes those meanings, then to me it sounds more like a continuative aspect than anything else.

Have I still completely misunderstood perfective aspect? Sigh, when I read about it I think I've understood it, but then when I see examples of how it's used in natlangs I'm like "why would you think of that action as perfective???"

Really? I have a strong intuition that locational adverbs are acting as verbal modifiers.

Huh. To me it just feels as if they were verb arguments (especially of motion verbs).

[Qwynegold]

Heh, I just saw your perfective thread. I'm gonna read and see if it's of any help...

[bradrn]

From this description, pan doesn’t sound at all like a perfective to me. Future actions are very often perfective, or at least marked using a prototypically ‘perfective’ aspect — e.g. English ‘The bus arrives tomorrow’ is morphologically ‘present simple’ (=perfective), but semantically future. (The combination of perfective+present tends to do weird things, because an entire event seen as a whole doesn’t really ‘fit’ into a single present moment.) Meanwhile, an event which ‘always happens’ is habitual: technically imperfective, according to Comrie, but a very non-prototypical imperfective. If your pan specifically excludes those meanings, then to me it sounds more like a continuative aspect than anything else.

Have I still completely misunderstood perfective aspect? Sigh, when I read about it I think I've understood it, but then when I see examples of how it's used in natlangs I'm like "why would you think of that action as perfective???"

Almost certainly you’re misunderstanding it. But that’s because basically no-one bothers explaining it properly. These days I tend to think about it in terms of prototypes, e.g. Dahl (1985):

A PFV verb will typically denote a single event, seen as an unanalysed whole, with a well-defined result or end-state, located in the past. More often than not, the event will be punctual, or at least, it will be seen as a single transition from one state to its opposite, the duration of which can be disregarded.

That is: the most prototypical perfective events are punctual, bounded, in the past, and involve change of state. On the other hand, the most prototypical progressive events must involve ongoing action, unboundedness, and occur in the present. These then form the ends of a continuum: perfective—habitual—stative—progressive, and which aspect is used for which meaning depends on their relative markednesses in each language. In English, the ‘perfective’ aspect is highly unmarked relative to the ‘imperfective’, so it is used for all of perfective, habitual and stative meanings. Slavic languages are the opposite: my understanding is that the Slavic ‘perfective’ is considered highly marked and thus can only be used for the most prototypical perfective events.

Really? I have a strong intuition that locational adverbs are acting as verbal modifiers.

Huh. To me it just feels as if they were verb arguments (especially of motion verbs).

This admittedly is true. It does ‘feel’ very strongly that e.g. ‘there’ in ‘I went there’ is an object. But it can’t be, because English ‘go’ cannot take any other ‘objects’. It ‘feels’ more like a modifier in other sentences such as e.g. ‘I cooked there’.