Talking about Shaleyan morphosyntactic alignment

A forum for all topics related to constructed languages
Post Reply
Khemehekis
runic
runic
Posts: 2499
Joined: 14 Aug 2010 09:36
Location: California über alles

Talking about Shaleyan morphosyntactic alignment

Post by Khemehekis »

So . . . my conlang Shaleyan has a split-S system. Subjects come before the verb, and objects come afterwards (SVO).

With intransitive verbs:

Anika waphan.
Anika smile
Anika smiles.

Wana lepus aken.
pass_away old man
The old man passed away.

Transitive verbs follow the old familiar SVO pattern when there is both an agent and a patient:

Mophasha komen kuyu.
kelp_bunny eat kelp
The kelp bunny eats the kelp.

When you're using a normally transitive verb but there's only one noun argument attached to it, the verb comes after an agent (active voice) but before a patient (passive voice):

Mophasha komen.
kelp_bunny eat
The kelp bunny eats.

Komen kuyu.
eat kelp
The kelp is eaten.

When you have two verbs that both come after the noun/agent argument, or two that both come before the noun argument, simply put the two together and join them with ye (and):

Kiwa aphan ye nezol yeshi.
Kiwa talk and chew gum
Kiwa talked and chewed gum.

Ishuz ye shizel Dashash.
fall_asleep and dream Josh
Josh fell asleep and dreamt.

If you have two intransitive whose chronological order goes "with the wind" of usual before the noun/after the noun rules, simply place the first one before the noun and the second one after the noun. You do not need a conjunction:

Labon Mayem pep.
wake_up Mayem cry
Mayem woke up and cried.

If you have two intransitive whose chronological order goes "against the wind" of usual before the noun/after the noun rules,
state the subject, then the first/after-the-noun verb, then the conjunction ñad, then the second/before-the-noun verb, then repeat the noun or a pronoun referring thereto:

Pawolo nekash ñad ishuz al.
Pawolo close_one's_eyes and_then fall_asleep 3s.ANIM
Pawolo closed his eyes and fell asleep.

Questions:

1. How would you label the intransitive verbs that go after the noun (like "talk" and "smile") in my Shaleyan dictionary, and how would you label the intransitive verbs that go before the noun (like "pass away" and "dream")?

2. In the sentence Ishuz ye shizel Dashash. (Josh fell asleep and dreamt) above, should I say that "Josh" is the agent of "fall asleep" and "dream", or should I say that "Josh" is the patient thereof?

3. Is this really simply a split-S language, or do I have something else going on here as well?
♂♥♂♀

Squirrels chase koi . . . chase squirrels

My Kankonian-English dictionary: 67,000 words and counting

31,416: The number of the conlanging beast!
User avatar
Creyeditor
MVP
MVP
Posts: 4070
Joined: 14 Aug 2012 19:32

Re: Talking about Shaleyan morphosyntactic alignment

Post by Creyeditor »

I really want to answer this questions, but I will need some time to digest your post. And thanks for all the glossed examples, that really helps.
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
Omzinesý
runic
runic
Posts: 2890
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Talking about Shaleyan morphosyntactic alignment

Post by Omzinesý »

Khemehekis wrote: 28 Oct 2020 03:19
Questions:

1. How would you label the intransitive verbs that go after the noun (like "talk" and "smile") in my Shaleyan dictionary, and how would you label the intransitive verbs that go before the noun (like "pass away" and "dream")?
If there is a semantic component, one takes a generalized agent (actor) and the other takes a generalized patient (undergoer), I would call them unergative and unaccusative verbs, respectively. If not, verb class 1 and verb class 2 (?).
Khemehekis wrote: 28 Oct 2020 03:19 2. In the sentence Ishuz ye shizel Dashash. (Josh fell asleep and dreamt) above, should I say that "Josh" is the agent of "fall asleep" and "dream", or should I say that "Josh" is the patient thereof?
In general semantics 'the agent' and 'the patient' apply well to relations like 'to kill'. For many other relations, it is fuzzy.
In a particular language, grammar (or is it lexicon?) determines rules for that, and thus the language-internal categories 'Agent'/Actor and 'Patient/Undergoer'. So you basically can decide.
Khemehekis wrote: 28 Oct 2020 03:19 3. Is this really simply a split-S language, or do I have something else going on here as well?
Nom-Acc langs and Erg-Abs alignments are relatively well defined, though it's still questionable if it applies to all the whole language. Split alignment is much worse defined and there is many in-between cases. With this very broad intuitive concept of Split alignments, I say your alignment could be called a split alignment.

You could give numbers to your examples, so that we could refer to them.

The first examples seem to imply that there really are syntax categories Agent and Patient that also apply to references between clauses.
"If you have two intransitive whose chronological order goes "with the wind" of usual before the noun/after the noun rules, simply place the first one before the noun and the second one after the noun." This is, on the other hand, not based on the categories, because the noun is Patient in the first and Agent in the second clause, but just the position relative to the verb. If they are two verbs in one clause, that could maybe be "natural".



----
I still question the concept of 'syntactic transitivity'. In real languages, anaphoric rules are usually much more complex (based on semantics, context, pure ambiguity...) But Dixon, for example, speaks much of it, so it is part of the usual definitions.

I know there is a language with an Arg-Abs word order in Africa. I think a language with word-order-based split alignment does not exist, but that does not mean you could not many such a conlang.

Word order changes surely are not voices, but a simpler thing.
---


I think I just commented typological concepts of alignment rather than your lang :D
User avatar
Creyeditor
MVP
MVP
Posts: 4070
Joined: 14 Aug 2012 19:32

Re: Talking about Shaleyan morphosyntactic alignment

Post by Creyeditor »

First of all, this looks very interesting and original to me. I do not think any natlang that I know of does this and still it is naturalistic. In order to improve your description (and my understanding [:D] ) you could distinguish more sharply between semantic terms for thematic roles (agent, patient) and syntactic notions (like the single argument of an intransitive verb). I also think that you're active voice could just be termed "argument drop" since it basically follows the same alignment principles that the rest of the languages does.
Khemehekis wrote: 28 Oct 2020 03:19 1. How would you label the intransitive verbs that go after the noun (like "talk" and "smile") in my Shaleyan dictionary, and how would you label the intransitive verbs that go before the noun (like "pass away" and "dream")?
You could go with unergative and unaccusative, as was proposed before. I have the feeling that these are a bit counterintuitive and people in the typological literatur sometimes uses agentive and patientive verbs.
Khemehekis wrote: 28 Oct 2020 03:19 2. In the sentence Ishuz ye shizel Dashash. (Josh fell asleep and dreamt) above, should I say that "Josh" is the agent of "fall asleep" and "dream", or should I say that "Josh" is the patient thereof?
I would say that syntactically, Josh is the subject of both verbs, in the sense that in both cases it is the single argument of an intransitive verb. On the other hand, in your language specifically, if Josh follows the verb it must be a patient, right?
Khemehekis wrote: 28 Oct 2020 03:19 3. Is this really simply a split-S language, or do I have something else going on here as well?
I think it it a special case of a split-S language, because you use word order to distinguish semantic roles. It would be interesting to see what a transitive verb with two non-agent like arguments works in your language, e.g. verbs of perception.
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]
Khemehekis
runic
runic
Posts: 2499
Joined: 14 Aug 2010 09:36
Location: California über alles

Re: Talking about Shaleyan morphosyntactic alignment

Post by Khemehekis »

Omzinesý wrote: 28 Oct 2020 11:03
Khemehekis wrote: 28 Oct 2020 03:19
Questions:

1. How would you label the intransitive verbs that go after the noun (like "talk" and "smile") in my Shaleyan dictionary, and how would you label the intransitive verbs that go before the noun (like "pass away" and "dream")?
If there is a semantic component, one takes a generalized agent (actor) and the other takes a generalized patient (undergoer), I would call them unergative and unaccusative verbs, respectively. If not, verb class 1 and verb class 2 (?).
Thanks for the suggestions! I'm going with Creyeditor's suggestions, though. And yes there are semantic differences between the two types.

I just typed this up in my Shaleyan grammar document:

Shaleyan verbs fall into five classes. These are their names, and how they are indicated in the Shaleyan dictionary at this website:
  • Transitive verbs (T): Transitive verbs can take both a subject (agent) and an object (patient). These are verbs that often have both agent and patient, analogous to nominative and accusative respectively in languages like German: take, kick, throw, cook, eat, drink, wear, open, wash, create, absorb, stab. These verbs may sometimes lack a patient (but still have an agent), and other times may lack an agent (but still have a patient).
  • Agentive verbs (A): Agentive verbs are verbs that take only one noun or pronoun argument, and come after the noun/pronoun argument, that nominal argument being the agent. These are normally verbs that the agent actively plans and chooses to do: talk (in the sense of converse), yell, smile, walk, run, close one's eyes, apologize, get dressed.
  • Patientive verbs (P): Patientive verbs are verbs that take only one noun or pronoun argument, and come before the noun/pronoun argument, that nominal argument being the patient. These are normally verbs that happen to the patient naturally or involuntarily: fall, fall asleep, pass away, dream, fly (make a trajectory through the air), crash, break, blink.
  • Linking verbs (L): Linking verbs express identity or membership to a predicate noun or being, becoming, remaining, or seeming to a predicate adjective. These verbs act very much like transitive verbs, with the predicate noun or predicate adjective acting like a patient: be, stay, sound, feel (emotionally), feel (physically), smell, act, dress (the last verb, while taking an adverb in English, takes a predicate adjective in Shaleyan).
  • Helping verbs (H): Helping verbs, also known as auxiliary verbs, are always used with another verb (an action verb). These include the modals. These verbs and the action verbs to which they are linked will come after the agent if they modify a transitive verb or an agentive verb, but before the patient if they modify a patientive verb: should [deontic], should [epistemic], be supposed to, may, can, might, must [obligation], be willing to.
Khemehekis wrote: 28 Oct 2020 03:19 2. In the sentence Ishuz ye shizel Dashash. (Josh fell asleep and dreamt) above, should I say that "Josh" is the agent of "fall asleep" and "dream", or should I say that "Josh" is the patient thereof?
In general semantics 'the agent' and 'the patient' apply well to relations like 'to kill'. For many other relations, it is fuzzy. In a particular language, grammar (or is it lexicon?) determines rules for that, and thus the language-internal categories 'Agent'/Actor and 'Patient/Undergoer'. So you basically can decide.
Ah, I never realized how fuzzy it all was!
Khemehekis wrote: 28 Oct 2020 03:19 3. Is this really simply a split-S language, or do I have something else going on here as well?
Nom-Acc langs and Erg-Abs alignments are relatively well defined, though it's still questionable if it applies to all the whole language. Split alignment is much worse defined and there is many in-between cases. With this very broad intuitive concept of Split alignments, I say your alignment could be called a split alignment.
Yayyyyyy!
You could give numbers to your examples, so that we could refer to them.
I'll do that in my response to Creyeditor.
I still question the concept of 'syntactic transitivity'. In real languages, anaphoric rules are usually much more complex (based on semantics, context, pure ambiguity...) But Dixon, for example, speaks much of it, so it is part of the usual definitions.
Another good thing to keep in mind. I mean, popular linguists usually speak of "eat" as a transitive verb, but it's often used without a direct object in English (How does Jeff find time to eat?)
I know there is a language with an Arg-Abs word order in Africa. I think a language with word-order-based split alignment does not exist, but that does not mean you could not many such a conlang.
Ooh, beating ANADEW sounds fun! BTW, what does "Arg-Abs" mean? I couldn't find anything on it online. Do you mean an argument-absolutive or argumentative-absolutive language? Or did "Arg" mean "Erg"?
Word order changes surely are not voices, but a simpler thing.
OK, maybe I should say "corresponding to the active voice in English" and "corresponding to the passive voice in English"?
♂♥♂♀

Squirrels chase koi . . . chase squirrels

My Kankonian-English dictionary: 67,000 words and counting

31,416: The number of the conlanging beast!
Khemehekis
runic
runic
Posts: 2499
Joined: 14 Aug 2010 09:36
Location: California über alles

Re: Talking about Shaleyan morphosyntactic alignment

Post by Khemehekis »

Creyeditor wrote: 28 Oct 2020 12:38 First of all, this looks very interesting and original to me. I do not think any natlang that I know of does this and still it is naturalistic.
Thank you for the compliment! It's always fun to be beyond ANADEW (something that seldom seems to happen to conlangers). It's also nice to hear that it's naturalistic. Shaleyan is rather weird typologically: it's SVO, yet adjectives come before nouns, and the language has postpositions. And let's not even get into relative clauses!

I've made a lot of weird decisions for Shaleyan over the years, but the only one I really regret is having a gap for /t/ in the phonology. When I first decided on the phonology of Shaleyan, I wanted a language that would taste of "metal and nectar". I decided that /t/ didn't fit the aesthetic. So now I have /p b d k/ but no /t/ nor /g/ (although having all five of the "big six" plosives except /g/ is normal). On Christophe Grandsire's page where he reviews my conlang, he says there's no /t/, and since that's already up on the web, I can't change my language now! The only change I made to my phonology since beginning it was adding ñ, and that's in lots of common words now.
In order to improve your description (and my understanding [:D] ) you could distinguish more sharply between semantic terms for thematic roles (agent, patient) and syntactic notions (like the single argument of an intransitive verb). I also think that you're active voice could just be termed "argument drop" since it basically follows the same alignment principles that the rest of the languages does.
Argument drop! I like that!

On second thought, though, isn't my "passive" (sentence #3, "The kelp is eaten") an argument drop too, since I drop the agent argument? Or does that term not mean what I think it does?
Khemehekis wrote: 28 Oct 2020 03:19 1. How would you label the intransitive verbs that go after the noun (like "talk" and "smile") in my Shaleyan dictionary, and how would you label the intransitive verbs that go before the noun (like "pass away" and "dream")?
You could go with unergative and unaccusative, as was proposed before. I have the feeling that these are a bit counterintuitive and people in the typological literatur sometimes uses agentive and patientive verbs.
Oh, OK. I'll go with "agentive" and "patientive". I even came across those two terms when reading Wikipedia's article on active-stative languages.
Khemehekis wrote: 28 Oct 2020 03:19 2. In the sentence Ishuz ye shizel Dashash. (Josh fell asleep and dreamt) above, should I say that "Josh" is the agent of "fall asleep" and "dream", or should I say that "Josh" is the patient thereof?
I would say that syntactically, Josh is the subject of both verbs, in the sense that in both cases it is the single argument of an intransitive verb. On the other hand, in your language specifically, if Josh follows the verb it must be a patient, right?
I guess you're right.
Khemehekis wrote: 28 Oct 2020 03:19 3. Is this really simply a split-S language, or do I have something else going on here as well?
I think it it a special case of a split-S language, because you use word order to distinguish semantic roles.
Do most split-S natlangs have cases? Or at least adpositions reminiscent of Nom-Acc Japanese's particles? I should note that, despite having no inflection on nouns nor even pronouns, Shaleyan takes care of some (non-MSA) cases with postpositions. For instance, the postposition pa (for) is used to mark datives.
It would be interesting to see what a transitive verb with two non-agent like arguments works in your language, e.g. verbs of perception.
Interesting indeed! By verbs of perception, do you mean things like "I watched the ducks eat", "Brian heard the glass shatter", etc? Or things like "You know (that) you're going to lose", "I suspect (that) Julie loves me", etc?

BTW, here's a rewrite of the OP as the section Sentence structure in my document. It's the second section of my Shaleyan grammar, coming immediately after phonology. Just open the spoiler tag:
Spoiler:
Shaleyan is an SVO (subject-verb-object) language. Shaleyan syntax is based upon agents and patients. It is a split-S language, meaning that certain verbs that only take one noun or pronoun as an argument always treat said noun/pronoun as an agent (and therefore, in the case of Shaleyan, will follow the noun or pronoun), and certain other verbs that only take one noun or pronoun as an argument always treat said noun/pronoun as a patient (and therefore, in the case of Shaleyan, will precede the noun or pronoun). There are five types of verbs in Shaleyan: transitive, agentive, patientive, linking, and helping (read the section on verbs for more information on the kinds of verbs that fall into each type).

Transitive verbs follow the old familiar SVO pattern when there is both an agent and a patient: agent-verb-patient:

1. Mophasha komen kuyu.
kelp_bunny eat kelp
The kelp bunny eats the kelp.

When you're using a normally transitive verb but there's only one noun argument attached to it, the verb comes after an agent (in uses corresponding to the active voice in English) but before a patient (in uses corresponding to the passive voice in English):

2. Mophasha komen.
kelp_bunny eat
The kelp bunny eats.

3. Komen kuyu.
eat kelp
The kelp is eaten.

When the only action verb in the clause is an agentive verb, it comes after the agent noun/pronoun:

4. Anika waphan.
Anika smile
Anika smiles.

When the only action verb in the clause is a patientive verb, it comes before the patient noun/pronoun:

5. Wana lepus aken.
pass_away old man
The old man passed away.

When you have multiple verbs that both/all come after the noun/agent argument (like two agentive verbs or an agentive verb and a transitive verb), or multiple verbs that both/all come before the noun/patient argument, simply put the two together and join them with ye (and) (you might also use khub (or) or a similar conjunction):

6. Ab mey Eli yomak ye iñal.
1s with Ellie walk and laugh
Ellie walked and laughed with me.

7. Kiwa azoy ye nezol yeshi.
Kiwa talk and chew gum
Kiwa talked and chewed gum.

8. Ishuz ye shizel Dashash.
fall_asleep and dream Josh
Josh fell asleep and dreamt.

9. Wushush khub lisaw phudokoy lay?
shake or shiver hostage Q
Did the hostage shake or shiver?

If you have two monovalent verbs, one agentive and the other patientive, whose chronological order goes "with the wind" of usual before the noun/after the noun rules (e.g. the patientive action occurs before the agentive action), simply place the patientive one before the noun and the agentive one after the noun. You do not need a conjunction:

10. Labon Mayem pep.
wake_up Mayem cry
Mayem woke up and cried.

If you have two monovalent verbs, one agentive and the other patientive, whose chronological order goes "against the wind" of usual before the noun/after the noun rules (e.g. the agentive action occurs before the patientive action), state the subject, then the agentive, then the conjunction ñad, then the patientive verb, then repeat the noun or a pronoun referring thereto:

11. Pawolo nekash ñad ishuz al.
Pawolo close_one's_eyes and_then fall_asleep 3s.ANIM
Pawolo closed his eyes and fell asleep.
♂♥♂♀

Squirrels chase koi . . . chase squirrels

My Kankonian-English dictionary: 67,000 words and counting

31,416: The number of the conlanging beast!
User avatar
Creyeditor
MVP
MVP
Posts: 4070
Joined: 14 Aug 2012 19:32

Re: Talking about Shaleyan morphosyntactic alignment

Post by Creyeditor »

Khemehekis wrote: 29 Oct 2020 04:04 On second thought, though, isn't my "passive" (sentence #3, "The kelp is eaten") an argument drop too, since I drop the agent argument? Or does that term not mean what I think it does?
Yes, you are very much correct in this.
Khemehekis wrote: 28 Oct 2020 03:19 Do most split-S natlangs have cases? Or at least adpositions reminiscent of Nom-Acc Japanese's particles? I should note that, despite having no inflection on nouns nor even pronouns, Shaleyan takes care of some (non-MSA) cases with postpositions. For instance, the postposition pa (for) is used to mark datives.
According to WALS most split-S languages actually use verbal person marking, but some use case marking. None uses both in their sample.
It would be interesting to see what a transitive verb with two non-agent like arguments works in your language, e.g. verbs of perception.
Interesting indeed! By verbs of perception, do you mean things like "I watched the ducks eat", "Brian heard the glass shatter", etc? Or things like "You know (that) you're going to lose", "I suspect (that) Julie loves me", etc?
Yes, but also something like "I hear the neighbour." or "I see a duck." In a way, the subject in these sentences is not actively doing something, it is just receiving signals. Some natlangs actually code these verbs with the perceptor as an object and the perceived thing as a subject.
BTW, here's a rewrite of the OP as the section Sentence structure in my document. It's the second section of my Shaleyan grammar, coming immediately after phonology. Just open the spoiler tag:
This is much easier to understand, at least for me [:)] I also still like the "wind" metaphor.
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
Omzinesý
runic
runic
Posts: 2890
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Talking about Shaleyan morphosyntactic alignment

Post by Omzinesý »

Khemehekis wrote: 29 Oct 2020 03:22
I know there is a language with an Arg-Abs word order in Africa. I think a language with word-order-based split alignment does not exist, but that does not mean you could not many such a conlang.
Ooh, beating ANADEW sounds fun! BTW, what does "Arg-Abs" mean? I couldn't find anything on it online. Do you mean an argument-absolutive or argumentative-absolutive language? Or did "Arg" mean "Erg"?
Just a typo, and not the only one. I just mean an absolutive-ergative alignment.

BTW
Transitivity is polysemous.

Syntactic transitivity: A construction with a subject and a (direct) object. ('I shoot to a bear.' is not transitive, 'I see you.' is.)
Semantic transitivity: A relation with an agent and a patient. Somebody does something affecting somebody. ('I shoot to a bear.' is transitive, 'I see you.' is not.)
People don't usually explicitly specify with one they mean.
Khemehekis
runic
runic
Posts: 2499
Joined: 14 Aug 2010 09:36
Location: California über alles

Re: Talking about Shaleyan morphosyntactic alignment

Post by Khemehekis »

Creyeditor wrote: 29 Oct 2020 10:58 According to WALS most split-S languages actually use verbal person marking, but some use case marking. None uses both in their sample.
Oh, OK, I guess I had a unique idea. Maybe it's just my tendency to make isolating languages, and the fact that thinking that way is second-nature because I speak an L1 wherein nouns (but not pronouns) don't decline for case (unless you count the genitive -'s/-s' as a case).
Creyeditor wrote:
Khemehekis wrote:
Creyeditor wrote:It would be interesting to see what a transitive verb with two non-agent like arguments works in your language, e.g. verbs of perception.
Interesting indeed! By verbs of perception, do you mean things like "I watched the ducks eat", "Brian heard the glass shatter", etc? Or things like "You know (that) you're going to lose", "I suspect (that) Julie loves me", etc?
Yes, but also something like "I hear the neighbour." or "I see a duck." In a way, the subject in these sentences is not actively doing something, it is just receiving signals. Some natlangs actually code these verbs with the perceptor as an object and the perceived thing as a subject.
Good question!

Verbs of sensing/perception have their subjects and objects reversed from English, as the person/animal/being sensing is considered a patient, and the entity being sensed is considered an agent. Rather than "to see", the transitive verb ped means "to be seen by", and rather than "to taste", the transitive verb kisas means "to be tasted by":

12. Pawolo ped Eli.
Pawolo be_seen_by Ellie
Ellie saw Pawolo.

13. Kash kisas Bodusob.
Bodusian be_tasted_by kash
The Bodusian tasted the kash.

14. Añak mosus añi.
añak have_one's_mind_read_by human
The human read the añak's mind.

Now, avid students of Kankonian can see some Kankonian influence in this next part. Consider how prepositions fill most of the roles in the Kankonian sentence in this post:

KNK1. Safin dyu welas id wapishas ab homoses ad mem as orol.
keep by child-PL done_to secret-PL from parents to 3p PRS natural
It is natural for children to keep secrets from their parents.

Now, Shaleyan can do something similar. There's a twist, though: Shaleyan has postpositional phrases instead of prepositional phrases, so . . .

When a verb of perceiving takes the perceiver as the patient, a perceived action (such as the ducks flying in the sentence "I saw the ducks fly") becomes the agent; the verb here is treated as a gerund, and therefore a noun. The agents of these actions are followed by the postposition pho (by), and the patients of these actions are followed by the postposition ba (of).

Perceived actions that employ the objects of transitive verbs and agentive verbs follow their agents:

15. Lepuma pho komen ped wudus.
lepuma by eat be_seen_by sailor
The sailor saw the lepuma eat.

16. Lepuma pho layide ba komen ped wudus.
lepuma by layide of eat be_seen_by sailor
The sailor saw a lepuma eat a layide.

(A lepuma is a cartilaginous Shaleyan sea creature of the plesiosaur/Nessie/Lapras type. A layide is a Shaleyan shellfish.)

17. Zon pho humal ped phila.
zon by fly be_seen_by little_girl
The little girl saw the zons fly.

(Zons are the ornithologue class on Shaleya, in case you were wondering.)

But with the patientive verbs, the patient+verb construction is simply placed before the verb of perceiving; there are no postpositions required:

18. Kasak sanaki shebak ab.
shatter glass be_heard_by 1s
I heard the glass shatter.

More active verbs of perceiving on the other hand (to look at, to watch, to listen to, to sniff . . .) take the perceiver as the agent and the perceived entity as the patient:

19. Pawolo ñoyoy ash.
Pawolo listen_to ash
Pawolo is listening to ash music.

Pho and ba are used with perceived action clauses that employ verbs the same way they are with verbs like "to see" and "to hear", except that these clauses come after the verb:

20. Wudus keñe lepuma pho komen.
sailor watch lepuma by eat
The sailor watched the lepuma eat.

21. Deko keñe zakako pho depush ba dezez.
little_boy watch bulldozer by building of raze
The little boy watched the bulldozer raze the building.

When the perceived action of an active perception verb (like "listen to" or "watch") includes a transitive verb with a patient but no agent (i.e. the equivalent of an English passive), place the patient of the action with the postposition ba after it and no agent+pho phrase:

22. Deko keñe depush ba dezez.
little_boy watch building of raze
The little boy watched the building being razed. / The little boy watched the razing of the building.

When the perceived action of an active perception verb includes a patientive verb such as "to shatter", place the perceived action clause after the verb and use ba, making the patient of the verb the postpositional object of ba:

23. Kiwa ñoyoy sanaki ba kasak.
Kiwa listen_to glass of shatter
Kiwa listened to the glass shatter.

Clause complements ("I think that . . .", "She knows that . . .", "He said that . . .", "We suspect that . . .", etc.) use the conjunction sol much the same way English uses "that". Since sol is a conjunction, the clause complement simply comes after sol, and is never placed before the being who thinks/knows/says/suspects/etc. the way "zons fly" in the sentence "The little girl saw the zons fly" is. This applies even if the verb that governs the clause complement is a patientive verb (such as "know" or "think"). Within a clause complement, all the elements of the clause follow the usual rules of Shaleyan syntax:

24. Udob Shapho sol zon suhey bosakh humal.
know Shapho that zon again home fly
Shapho knows that the zons will fly home again.

25. Mahala leña sol wana al ba wabe.
Mahala say that pass_away 3s.ANIM of grandmother
Mahala said her grandmother passed away.

26. Sedon ab sol Dashim sayaphaw Nadali.
think 1s that Jim be_liked_by Natalie.
I think Natalie likes Jim.

27. Hakim pa Shosini dek sol Dañaw pa al ba masha pa al benu ñudish.
Hakim for Shosini tell that Dañaw for 3s.ANIM of daughter for 3s.ANIM give blender
Shosini told Hakim that she was getting her daughter a blender for Dañaw.

(Dañaw is a gift-giving holiday of the Shaleyan culture. It has not been corporatized the way Christmas has.)

28. Udob esh sol Kaña hel kolowa.
know 2s that Kaña be happy
You know Kaña is going to be happy.
♂♥♂♀

Squirrels chase koi . . . chase squirrels

My Kankonian-English dictionary: 67,000 words and counting

31,416: The number of the conlanging beast!
Post Reply