Exploring the secundative

[Ketumak]

This query relates to my own invader conlang, Classical Leheitak, but may be of broader interest. I'd like to learn more about how exactly secundative languages work, beyond the basic.

For those who don't know, in languages like English the direct object corresponds to the transitive patient and the ditransitive theme, whilst the indirect object is the ditransitive recipient. In secundative languages the primary object corresponds to the transitive patient and ditransitive recipient and the secondary object is the ditransitive theme.

I've been planning to use this alignment in Classical Leheitak, but have lately thought, there would be little scope for it as ditransitive verbs are rare in languages I'm familiar with. So I want to find ways to make some more of them and extend the scope for secundativity.

The classic ditransitive situation involves the transfer of an object:
X gave Y to Z

In a secundative language this becomes something like:
X gave Z "ka" Y


Do secundative natlangs extend this to:

- required locatives, e.g. placing objects?
X put the Y on the Z becoming X put.on Z "ka" the Y

- the directional sense of English "to" or "at" with intransitive verbs?
X went to Y becoming X went "ka" Y
X laughed at Y becoming X laughed "ka" Y

- benefactors that are not required (in English, at any rate)?
X did Y for Z becoming X did.for Z "ka" Y

- patients in the presence of an instrument?
X climbed the Y with a Z becoming X climbed.with Z "ka" the Y

Are there any other ways of adding ditransitive verbs to a language that would generate candidates for the secundative treatment? Thanks in advance.

[Pogostick Man]

I'm not a syntactician, but those all seem plausible to me. I could even see those being used without the "ka" phrases as shorthand for "he did it for Y/he gave it to Y".

Are there any other ways of adding ditransitive verbs to a language that would generate candidates for the secundative treatment? Thanks in advance.

Perhaps using words such as emotions as objects and then the "ka" phrase for the thing done as a statement of manner/mental state in which it is done?

[vec]

According to this article by Martin Haspelmath, a tripartite alignment with regards to ditransitives, is only attested in one language he could find. All languages with secundatives have identical nominatives and accusatives. In these cases, the nominative-accusative case is often termed primative.

Secundative alignments are either by overt marking (flagging in Haspelmath's parlance) or by indexing, which is syntactic. These can be mixed together. If you want an almost-tripartite secundative situation, the most likely scenario is indirective marking with secundative indexing. So the recipient goes in the direct object slot, but is unusually marked, while the theme is morphologically marked as if it were a direct object.

(He argues that actual tripartite languages, whether with regards to objects, or subjects, violate the near-universal of parsimony, which assumes that grammars generally don't contain blatantly unnecessary features. He further argues that horizontal alignments, where the theme and recipient are identical, but subject is distinct, violate the requirement for distinctiveness between objects.)

Languages will probably be on a sliding scale when it comes to secundatitivity. English has a handful of verbs that act as secundative, most famously provide: "He provided me with the documents".

Note that the secundative marker will often be the instrumental case, or something similar. For that reason, I find your final example (patients in the presence of instruments) to be the most unlikely, because instruments and themes are semantically very related. Because ditransitives are not that common, languages will be pretty unlikely to have a dedicated secundative case with no other uses (again, the "law" of parsimony). The secundative might be a general-purpose theme-marker: I can imagine a language where non-ergative subjects are marked with the secundative case. (Is that maybe basically Stative-Active alignment?).

I know this doesn't answer your question directly. These are just some thoughts.

[Ketumak]

I'm not a syntactician, but those all seem plausible to me. I could even see those being used without the "ka" phrases as shorthand for "he did it for Y/he gave it to Y".

Good to hear I'm on some of the right lines and I like your idea of optional omission of the "ka" phrase. I now agree with vec though about my last example being a stretch too far.

Perhaps using words such as emotions as objects and then the "ka" phrase for the thing done as a statement of manner/mental state in which it is done?

That sounds good, if I understand you correctly. Do you mean like:
He shouted "ka" anger

Classical Leheitak doesn't have a separate adverb class. I'd been thinking just use the adjective and position (immediately after the verb). A noun with preposition would work as well and would have the advantage that it could be moved around more and its referent would still be clear:
"Ka" anger, he shouted

According to this article by Martin Haspelmath

Aha! That's where it went. Thank you, I found this a couple of years ago and didn't keep a copy but have done this time. As a result of the earlier reading, I've already taken the point about secundativity going with indexing arguments on the verb - CL is a head-marking and active language that marks the subject and primary object on the verb.

Note that the secundative marker will often be the instrumental case, or something similar. For that reason, I find your final example (patients in the presence of instruments) to be the most unlikely, because instruments and themes are semantically very related. Because ditransitives are not that common, languages will be pretty unlikely to have a dedicated secundative case with no other uses (again, the "law" of parsimony).

I was originally planning to have a dedicated secundative preposition as Yoruba does, but your idea of reusing another one sounds more realistic. I'd have to think about which preposition though. Looking back at my examples above, I see some that would work with an instrumental preposition whilst others would be better with something else, not sure what yet.

I wasn't looking for a tripartite situation between P,T and R. I hope my examples above aren't pointing in that direction.