Handling polypersonal causatives and desideratives

[sasasha]

Just starting a new project, and could use some advice. Verbs in Bogtra take polypersonal prefixes that in their simplest form look something like this:

pah-

Code: Select all

slot#       3         2        1        
morpheme    p-        -a-      -h-
gloss       1S.AN.AG  PRES     3S.INAN.PAT

To which the verb stem is appended, e.g. pahtos 'I look at it'.

There is a set of causatives and desideratives that go in slots 4-5 (i.e., further to the left) and look something like this:

desiderative prefix: i-
causative: Cë-, where C is the relevant person affix of the new instigatory subject

ipahtos 'I want to look at it'.
tëpahtos 'You make me look at it'.

I'm thinking that I want to be able to 'refocus' my desiderative, in a similar way to the causative, so that the desiring is no longer being done by the original subject. This is done by adding the relevant person affix in slot 5:

tipahtos 'You want me to look at it'.

I'm then thinking that I want to be able to 'retense' these forms, so that the causation and the desiring do not have to occur in the same time frame as the original action. So far tense is being indicated in slot 2 of the prefix (the 'a' of pah-), with a simple 3-way distinction (a present, i future, u past).

The problem is I can't take this approach, because the same infixes already have different roles in slot 4. -i-, obviously, is signifying the desiderative already, whilst in this slot -a- signifies a progressive, and -u- a prohibitive. So I'm currently plumping for the creation of a new slot before this one, with the same patterns as slot 2, which would enable forms like:

tuipahtos 'You wanted me to look at it'.

I can't help but feel though that this is an unlikely strategy, since I was already planning on some two-way combinations of desiderative, prohibitive and progressive being possible indicated in much the same way. For instance, in that scheme, the above example could mean 'You must not want me to look at it.' (A bit obscure, I know, but better examples could be found). With the two slots 4 and 5 being right next to each other and both using exactly the same three infixes, it feels wrong to assign them completely different values. I could just change the infixes, but phonotactically I'm running out of options; I think it might be even weirder to have two sets of infixes for the same roles, just in different slots; and I've been trying to keep the actual number of infixes used as low as possible (plus I've simplified matters for this post: most other vowels have roles in these slots too). I could of course make all the infixes longer than a single phoneme and thus easier to distinguish from one another, but I'm trying to create something fairly concise...

My question: does anyone have any insights into how polypersonal systems handle causatives and desideratives and the like - are they often refocused and retensed in this way, and do they run into any similar issues? Where might I be best to look for some inspiration/help?

[bradrn]

A possibly stupid question: is there anything wrong with just declaring that slot 2 in causatives/desideratives refers to the tense of the causation/desire? So e.g. tipahtos is ‘You want me to look at it’, tipihtos (I think) is ‘You will want me to look at it’ etc. Or are you asking for something different?

[cedh]

I'm probably not answering your question directly, but maybe it would be helpful to think about a possible earlier stage of this language. Suppose the desiderative and causative were once auxiliary verbs that inflected in exactly the same way as the main verb. How would the relevant types of sentences have looked at that time? Maybe like this:

Code: Select all

*t-  u-   h-  i    p-  a-   h-  tos
 2:A-PAST-3:P-want 1:A-PRES-3:P-look
"you wanted me to look at it" (lit. "you wanted it, I look at it")

Or like this:

Code: Select all

*t-  u-   p-  i    p-  a-   h-  tos
 2:A-PAST-1:P-want 1:A-PRES-3:P-look
"you wanted me to look at it" (lit. "you wanted me, I look at it")

[sasasha]

A possibly stupid question: is there anything wrong with just declaring that slot 2 in causatives/desideratives refers to the tense of the causation/desire? So e.g. tipahtos is ‘You want me to look at it’, tipihtos (I think) is ‘You will want me to look at it’ etc. Or are you asking for something different?

That's not a stupid question at all, it hadn't actually occurred to me. But I did want to be able to use constructions where the two tenses don't correlate, like 'You will want me to have looked at it'.

However, I think I'm confusing myself unnecessarily. I can use a perfect infix (ö) in combination with the future tense to produce

tipihtos - you will want me to look at it
tipiöhtos - 'you will have wanted me to look at it' = maybe close enough (or maybe the perfect part is simply understood to apply to the verb).

I think aspectual information modifying the overall tense in slot 2 should be able to cover a wide enough array of nuances.

I'm probably not answering your question directly, but maybe it would be helpful to think about a possible earlier stage of this language. Suppose the desiderative and causative were once auxiliary verbs that inflected in exactly the same way as the main verb. How would the relevant types of sentences have looked at that time? Maybe like this:

Code: Select all

*t-  u-   h-  i    p-  a-   h-  tos
 2:A-PAST-3:P-want 1:A-PRES-3:P-look
"you wanted me to look at it" (lit. "you wanted it, I look at it")

Or like this:

Code: Select all

*t-  u-   p-  i    p-  a-   h-  tos
 2:A-PAST-1:P-want 1:A-PRES-3:P-look
"you wanted me to look at it" (lit. "you wanted me, I look at it")

This is really useful, thanks! That's definitely what I was thinking in terms of development, but I wasn't seeing it as clearly. I may use a construction like the first of your examples, treating the desiderative morpheme as an auxiliary verb, for any remaining situations in which certain information needs to be assigned specifically to the desiderative part, rather than the main clause. It's also given me some ideas for relative clauses.

Distressing aside: the 1st person singular agent pronoun is looking like it's going to have to be poo. I am considering changing a few things around to avoid that...

[Ares Land]

My question: does anyone have any insights into how polypersonal systems handle causatives and desideratives and the like - are they often refocused and retensed in this way, and do they run into any similar issues? Where might I be best to look for some inspiration/help?

In French we have:

Je veux le faire > I want to do it
Il veut que je le fasse He wants me to do it.
Il voulait que je le fasse He wanted me to do it.

Now, je veux le faire can be considered one word, and I think a case could be made that the other two are one word too (I'm pretty sure I just have one stressed syllable: fasse in any case)

Other than that, I second cedh's advice to look at it from a diachronic persective. Consider your problem here:

With the two slots 4 and 5 being right next to each other and both using exactly the same three infixes, it feels wrong to assign them completely different values

(1)Maybe i has a different etymon in both cases, and became identical through sound change, which is entirely plausible.
(2)Or maybe it's the same etymon, and the position changes the meaning, again very plausible. Especially for -i-, desiderative and future being so close to each other.

Considering French again, we have examples of both and we don't even notice it.
(1) j'aimais /ʒɛmɛ/ 'I loved' > j'aimerais /ʒɛmʁɛ/ 'I would love'
(2) j'ai aimé /ʒe.ɛmɛ/ 'I have loved' (subjunctive) > j'aimerai /ʒɛmʁe/

(BTW I'm using a fairly conservative transcription, many (European) speakers have ruthless mergers of ɛ and e)