A sample of clause types

A forum for all topics related to constructed languages
Post Reply
chris_notts
hieroglyphic
hieroglyphic
Posts: 46
Joined: 14 Feb 2012 00:49

A sample of clause types

Post by chris_notts »

I'm (kindof) back after a long time away! I was re-reading "Intransitive Predication" by Stassen, and I thought it might make a good conlang quiz, one which I may or may not have done before. Without going into too many details, Stassen looks at patterns in the encoding of the following and how they compare with the typical verbal strategy in different languages:

1. Nominal predication

Examples of nominal predicates expressing identity and class membership, encoded on English with a copula verb:

Identity: he's the man I saw yesterday
Class membership: he's a man

Commonly encoded by a copula verb, copula particle (i.e. with distinct morphology/syntax compared to verbs), or by apposition of nouns. Some languages use the verbal strategy and apply person/TAM affixes/clitics to predicate nouns.

2. Locative predication

Expressions of location. Typically encoded by one or a small class of verbs together with a locative argument. Verbs may simultaenously encode posture or orientation of the figure. Sometimes expressed by just a noun and location with no verb present.

A separate but related class is existential clauses, which may be identical to locationals, differ in word order or pragmatic marking (compare English "A man is there" to "There is a man"), or use completely different supporting verbs.

3. Predicate adjectives

Typically marked either by the same strategy as predicate nouns (e.g. supported by a copula verb), or by the same strategy as verbs (taking verbal agreement and TAM marking). Stassen claims that in languages with inflectional tense marking on verbs predicate adjectives will tend to use the nominal strategy, othereise they'll tend to use the verbal strategy.

In a separate book, Stassen also looks at strategies for predicate possession, including:

Locational strategy = to/at Possessor there is/exists Possessee
With strategy = Possessor is with Possessee
Have strategy = Possessor has Possessee
Topic strategy = Possessor, Possessee exist (possibly with possessive marking on Possessee)

So, for your conlangs:

1. How is nominal predication expressed?
2. How is adjectival predication expressed?
3. How is locative predication expressed?
4. How are existentials expressed?
5. How is predicative possession expressed?
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: A sample of clause types

Post by Salmoneus »

chris_notts wrote: 19 Oct 2018 23:12 Stassen claims that in languages with inflectional tense marking on verbs predicate adjectives will tend to use the nominal strategy, othereise they'll tend to use the verbal strategy.
Oh, that's interesting! I haven't seen that claim before, though it would seem to make sense.
1. How is nominal predication expressed?
In Rawàng Ata, mostly by topic-comment structures. "Him, a man". At least in earlier versions there was also a copula particle, and I think I probably still agree with that, but it's used for emphasis and disambiguation, not as the default structure.

In Wenthish, mostly by copula verbs, although I suspect they may also use topic-comment with pseudo-copulas for some classifications ("Among men, stands him" or the like).
2. How is adjectival predication expressed?
In Rawàng Ata, almost entirely verbally. There is a very small class of true adjectives, and there are also locative nouns, but most 'adjectives' are just verbs. I can't remember how I deal with true adjectives predicatively, but I suspect it uses as "man, big man" ('the man is big') topic-comment pseudo-classificatory approach.

In Wenthish, I'm still not sure. You can just use the nominal presication strategy, but I think they may prefer rephrasing as an attributive ("He's a big man, the man"). I don't know.

3. How is locative predication expressed?
In Rawàng Ata, by default topic-comment, either plain ("cat, mat") or with a noun in the locative case ("cat, mat-at"). It's also possible to topicalise the location ("mat, cat"), possibly using a dummy verb where disambiguation is necessary ("mat, cat sleeps").

In Wenthish, mostly pseudo-copulas (derived from positional verbs) with prepostional phrases. "the cat sits on the mat".
4. How are existentials expressed?
In Rawàng Ata, just as they are. "Cat!" ('there's a cat'). Some sort of dummy verb can be used for disambiguation - probably in this case "a cat breathes".

In Wenthish... I don't actually know yet.
5. How is predicative possession expressed?
In Rawàng Ata, topic-comment. "Man, dog" ('the man has a dog'). Rawàng Ata, as you may have noticed, is not very in to syntax when it isn't needed. It does, however, have a hugely complicated system of attributive possession, so I suspect it often restructures sentences to employ that. "man, hammer tool-belonging-him". (Rawàng Ata uses possessive classifiers that look a lot like inflected verbs, so whether you parse it "hammer he-tool.owns" or "hammer tool-his" or whatever is sort of tricky).

In Wenthish, it's primarily prepositional ('the hammer is with the man'). Although there are verbs that can be used to be more specific, like 'the man owns the hammer' vs. 'the man is holding the hammer'.



Good questions.
User avatar
Reyzadren
greek
greek
Posts: 684
Joined: 14 May 2017 10:39
Contact:

Re: A sample of clause types

Post by Reyzadren »

1. Nominal predication:
Identity: Use the irregular verb/copula /azˤ/ (is) as a general word. For absolute 2-way identity, use /'azˤa/ instead.
Class membership: Use the verb /'vYʒa/, which broadly means "S subset O".

2. Adjectival predication:
/azˤ/ is used as a general word. However, adjectival verbs are also common instances, along with the rarer statefunction.

3. Locative predication:
/azˤ/ is then used with a preposition, though /'zˤidan/ (got-PASS) is also used just as much.

4. Existential:
Use /'zˤida/ (got) to show the presence of an object, without a "subject".

5. Predicative possession:
Use /'rətʃa/ (have).
Stassen claims that in languages with inflectional tense marking on verbs predicate adjectives will tend to use the nominal strategy, othereise they'll tend to use the verbal strategy.
Well, my griuskant conlang has virtually no inflectional tenses at all, yet it uses both nominal and verbal strategies equally according to different circumstances (though it tends towards nominals with just a slight edge).
Last edited by Reyzadren on 24 Oct 2018 00:10, edited 1 time in total.
Image conlang summary | Image griushkoent thread
chris_notts
hieroglyphic
hieroglyphic
Posts: 46
Joined: 14 Feb 2012 00:49

Re: A sample of clause types

Post by chris_notts »

I guess I'll do the conlang I'm working on (provisionally called 'Sint' but not sure if I'll change it). The text below is mostly copied from my messy incomplete grammar.

Nominal predication

In Sint, identity and nominal predication require the copula sa-, which is not morphologically a normal verb. The only
affixes which sa- can take are the gender agreement suffixes -n `masculine' and `plural/neuter'.
Agreement in gender occurs with both third persons and speech act participants. This agreement pattern differs from verbs,
which normally agree with their subjects in person and via prefixes.

In addition, the copula must occur between its two arguments, in contrast to the typical verb initial word order, and both
arguments must be present. For example:

hark san (šin) kiins
hark sa-n (šin) kiins
that.m be-m (nonspec.m) man
`he is a man'

helčaš sa tles nilai
helčaš sa tles n-i-lai
this.f be p.f 1exc-p-see
`this is the one (feminine) I saw'

The article used with the copula complement differs depending on the nature of the complement and
of the relationship being asserted. With identity assertions, the copular complement has a determinate (distal or proximate) article. With assertions of class membership, indeterminate articles are used: optionally if the copular complement is a simple noun, and obligatorily if it is a complex noun phrase.

There is a special negative form raik of the copula which does not show any agreement:

hark raik kiins
that.m not.be man
`he is not a man'

Adjectival Predication

Predicate adjectives take verbal morphology and act like verbs. Without further marking they are inchoative, and so
they typically combine with one of the stance suffixes to form a stative resultative. For example:

akalktsi
a-kalk-tsi
d-big-sit
`it/she is big'

Locative and Existential Clauses

Location is described a combination of pre-verbs and/or the verbal root \emph{e-} together with an obligatory posture suffix. For example:

neki o ke ikaar
n-e-ki o ke ikaar
1exc-loc-stand obl d.n camp
`I'm at the camp'

The same strategy is used for existentials, but often without any location specified:

etsi la šaaktıh
e-tsi la šaaktıh
loc-sit d.f god
`God exists'

Speech

Speech is marked by the quotative particle me-, which takes the same endings -n `masculine' and `plural/neuter' as the copula. It may be used alone to express speech, in which case the subject precedes the quotative and the speech follows:

ček men ``ninaa wai"
I quot-m 1exc-P-go now
`I (male) said ``I'm going now"'

me- is also used as a subordinator for speech, and another way of expressing speech is to combine it with the verb ji `do':

niji men ``ninaa wai"
1exc-p-do quot-m 1exc-P-go now
`I (male) said ``I'm going now"'

The supporting verb is obligatory if the clause is negated:

vanji men ``ninaa wai"
neg-d-1exc-do quot-m 1exc-P-go now
`I didn't say ``I'm going now'''

Possession

Possession is expressed by an existential clause with a possessed noun. For example:

etsi na čàmhım
loc-sit d.m dog-1sg
`I have a dog (lit: my dog exists)'
chris_notts
hieroglyphic
hieroglyphic
Posts: 46
Joined: 14 Feb 2012 00:49

Re: A sample of clause types

Post by chris_notts »

Salmoneus wrote: 19 Oct 2018 23:53 Oh, that's interesting! I haven't seen that claim before, though it would seem to make sense.
The explanation he offers is that tense is the least relevant of tense, aspect and mood to adjectives since they typically describe time-stable properties. Obligatory marking of tense therefore repels adjectives.

I guess that since adjectives have two functions, attributive and predicative, they ideally want to be minimally marked in at least one of them. If they full assimilate to the verb class and must be marked for tense then they lose both ways:

tall-PAST man -> predicative function marked
the man that tall-PAST -> attributive function marked

Of course, that doesn't explain why languages with tense wouldn't have a hybrid structure, since being forced to take tense marking isn't obviously worse than being forced to use a supporting verb / copula:

tall-PAST man -> verbal morphology in predicate function
the tall man -> non-verbal morphology in attributive function

I'll have to re-read that section of the book to remind myself is Stassen addresses the why in more detail.
akamchinjir
rupestrian
rupestrian
Posts: 2
Joined: 30 Sep 2018 14:56

Re: A sample of clause types

Post by akamchinjir »

I'll do this, for a language I'm calling Akiatu.

1. Nominal predication.

Class membership normally just uses the noun as a predicate:

Code: Select all

kai isaunaki
3s  healer
S/he is a healer
This can be supported (a bit emphatically) by the affirmative particle iti:

Code: Select all

kai iti isaunaki
3s  AFF healer
S/he is a healer
(iti can also occur with verbal predicates, but not with negation.)

For an identity statement, both nouns need to be definite. Context will often be enough. Subjects and topics have to be definite, so that's one way; demonstatives are also sometimes appropriate.

Code: Select all

isaunaki (iti) kai
healer    AFF  3s
The healer is her/him

Code: Select all

isaunaki wai, kai
healer   TOP  3s
The healer s/he is

Code: Select all

janaki kja  hau saraikuti kiwaita jaku wai, kai
person COMP 1s  yesterday see     PFV  TOP  3s
The person I saw yesterday is her/him

Code: Select all

kai ki  janaki kja  hau saraikuti kiwaita jaku=wati
3s  DET person COMP 1s  yesterday see     PFV =DEIC
S/he is that person I saw yesterday
2. Adjectives

There's a small closed class of true adjectives (adjectives that can directly modify a noun, without a relative clause), and a lot of stative verbs. The true adjectives can enter into a predicate only when modifying a head noun, so this reduces to the previous case:

Code: Select all

iti siwani siwi
AFF event  minor
It is an insignificant matter
(So you'll often get the equivalent of "It is a N Adj" rather than "The N is Adj.")

3. Locatives

Locative predicates require a verb, but no copula is available. The most neutral choice is a semantically and lexically appropriate posture verb.

Code: Select all

kawi akimau a   ikjamii kura
3p   stand  LOC river   bank
They are by the river
(I hope to come up with a fairly complex system for choosing which posture verb to use on a particular occasion, but I haven't done that yet.)

4. Existentials

An existential statement is like a locative statement with an indefinite subject---except that Akiatu isn't supposed to let an indefinite noun phrase be subject, so it has to cheat. The most neutral cheat uses simple inversion:

Code: Select all

ikjamii kura pijatu     ijaisa
river   bank be.hanging bat
There is a bat (/ are bats) by the river.
There's also a sort of impersonal passive that accomplishes much the same:

Code: Select all

ikjamii kura kiwaita   pijatu     a   ijaisa
river   bank see(PASS) be.hanging LOC bat
By the river there are bats hanging
(The verb will be less semantically bleached in this construction than with a simple inversion.)

5. Possession

This piggybacks off of the (inverting) existential structure, often using the (distal) deictic element wati to give the possessor a sort of locative sense:

Code: Select all

isaunaki=wati ijau anatu
healer  =DEIC sit  spear
The healer has a spear
rgj40q
hieroglyphic
hieroglyphic
Posts: 25
Joined: 06 Oct 2018 21:14

Re: A sample of clause types

Post by rgj40q »

1. How is nominal predication expressed? In Nuliziti, a noun gets a verb suffix, or a copula verb to be is also acceptable.

2. How is adjectival predication expressed? An adjective is a noun, so the same way.

3. How is locative predication expressed? Via the single-word verbs: be-at, be-inside-of, go-toward, go-from etc. Nonpredicative locationals are adverbs, which are in turn the nonfinite forms of the same verbs: be-at, go-toward etc.

4. How are existentials expressed? Noun gets a verb suffix, or phrases: <noun> exists, <noun> is-some.

5. How is predicative possession expressed? "Have strategy", or possession may be built into the word: be-friend-of, be-hand-of etc.
Iyionaku
mayan
mayan
Posts: 2102
Joined: 25 May 2014 14:17

Re: A sample of clause types

Post by Iyionaku »

:con: Yélian

1. How is nominal predication expressed?

Yélian generally has a verb for "to be" that covers most predicates. It's "balia" and highly irregular.

Simón bit iusorivar.
Simon COP.3SG.ANIM hiker-ambitious
Simon is an ambitious hiker.

2. How is adjectival predication expressed?

Generally similar, but note that all adjectives must be connected to the main verb.

U'neyʻir nelbut.
DEF.INAN=ball red-COP.3SG.INAN
The ball is red.

3. How is locative predication expressed?

Same, "balia" is used.

Sao tolar but pas pàcat.
2SG.POSS present COP.3SG.INAN on shelf
Your present is on the shelf.

4. How are existentials expressed?

There is a special constrution with the verbs "ciava" (to be there) and "cia" (not to be there), both can only be used in the 3rd person.

Vut zifeciavet avárlocan.
3SG.INDEF here-be_there-3SG cat-many
There are many cats here.

5. How is predicative possession expressed?

There is another irregular verb "biylra". Its 1SG and 2SG forms are identical to the verb "balia", so there can be ambiguous sentences.

Re vutret, re bai teriys.
1SG.REC INT, 1SG.REC have-1SG dog
I have a dog. Could also mean: I am a dog.
Wipe the glass. This is the usual way to start, even in the days, day and night, only a happy one.
User avatar
eldin raigmore
korean
korean
Posts: 6352
Joined: 14 Aug 2010 19:38
Location: SouthEast Michigan

Re: A sample of clause types

Post by eldin raigmore »

chris_notts wrote: 19 Oct 2018 23:12 ...
1. Nominal predication
...
2. Locative predication
...
3. Predicate adjectives
...
...strategies for predicate possession, including:
Locational strategy = to/at Possessor there is/exists Possessee
With strategy = Possessor is with Possessee
Have strategy = Possessor has Possessee
Topic strategy = Possessor, Possessee exist (possibly with possessive marking on Possessee)

So, for your conlangs:

1. How is nominal predication expressed?
2. How is adjectival predication expressed?
3. How is locative predication expressed?
4. How are existentials expressed?
5. How is predicative possession expressed?
I’m not ready to answer but I don’t want to forget.
Post Reply