Causatives in Tagalog (Help, anyone?)

[Jipí]

I'm working on figuring out causative constructions in my conlang at the moment. According to two typological articles I've read, its main way to form causatives is pretty batshit, but according to the Tagalog grammar I have, my conlang might actually kind of mirror its behavior. However, that grammar unfortunately only gives examples of causative sentences without their non-causative equivalents, and doesn't give glosses of anything either. Which sucks.

Here are two examples of causative sentences form that grammar:

Iniluha ni Nena ang pagkawala ng alahas niya.
'The loss of her jewelry made Nena shed tears.'

Ikinaluha ni Nena ang usok.
'The smoke made Nena shed tears.'

(Schachter and Otanes 313)

Schachter and Otanes (313–14) basically just explain that i- and ika- (as well as a bunch of other forms) are used as causative-focus (CF) prefixes corresponding to the actor-focus (AF) infix -um- (which would mean that causative-focus markers encode a kind of actor?). From looking things up in a dictionary I could gather that ni marks names as being in the actor role in goal-focussed clauses – which, if true, is rather interesting because according to those two articles on causative typology I've read that's kind of unusual (but is basically exactly what my conlang does!). Apparently you'd expect the causee (i.e. the actor of the incorporated clause) to be demoted in some way. I suppose that this is not necessarily true if a language marks case based more on semantics than syntax, which seems to be the case here and is definitely true of my conlang.

Now, with further help of a dictionary I assume the above examples break down as:

Code: Select all

I-  ni- luha       ni= Nena ang= pagkawala ng=  alahas  niya.
CF- ?-  shed.tears A=  NAME T=   loss      GEN= jewelry 3S

Ik<in>a- luha       ni= Nena ang usok.
CF<GF?>- shed.tears A=  NAME T=  smoke

Is this analysis correct? What are the -ni and -in- affixes doing there?! Are they in fact instances of the goal/object focus marker (note the remark on ni above)? Would that mean that the verb is marked for the ang-marked constituent being the topic/definite, with the causative-topic marker just piled on top of that, instead of the causative-topic marker itself corresponding to the ang-marked phrase? And what would the sentences be without the causer part, i.e. what would just Nena shed tears translate to?

-----
A = actor
CF = causative focus
GEN = genitive
GF = goal focus
T = topic

-----

• Schachter, Paul and Fe T. Otanes. 1972. Tagalog Reference Grammar. Berkeley: U of California P, 1983. 313–14. Google Books. Google, 2011. Web. 9 Jan. 2012. ‹http://books.google.com/books?id=E8tApLUNy94C›
• “Tagalog Dictionary.” SEAsite. Northern Illinois U, 25 Jun. 2001. Web. 24 Jan. 2012. ‹http://www.seasite.niu.edu/Tagalog/Dict ... iction.htm›

[Miekko]

Well, my conlang does this by ... []

Just kidding, and inb4 everyone else doing it and I don't get why people tend to think what their conlang does is pertinent when the question is specifically about some natural language etc. Ok, sorry, let's get back on track.

[clawgrip]

I don't get why people tend to think what their conlang does is pertinent when the question is specifically about some natural language

Background information on why he is asking the question in the first place I guess.

[kodé]

Code: Select all

I-  ni- luha       ni= Nena ang= pagkawala ng=  alahas  niya.
CF- ?-  shed.tears A=  NAME T=   loss      GEN= jewelry 3S

Ik<in>a- luha       ni= Nena ang usok.
CF<GF?>- shed.tears A=  NAME T=  smoke

Is this analysis correct? What are the -ni and -in- affixes doing there?! Are they in fact instances of the goal/object focus marker (note the remark on ni above)? Would that mean that the verb is marked for the ang-marked constituent being the topic/definite, with the causative-topic marker just piled on top of that, instead of the causative-topic marker itself corresponding to the ang-marked phrase? And what would the sentences be without the causer part, i.e. what would just Nena shed tears translate to?

I think the [-in-] might just be a prefix/infix for perfective aspect (it is in some related Austronesian Formosan languages). So the verbs would break down as:
[in-i-luha] 'PRF-CAUS1-shed.tears'
[ik<in>a-luha] 'CAUS2<PRF>-shed.tears'
But I'm not sure why the affix would be prefixed in one vowel-initial form and infixed in the other...

[Jipí]

You may be right, I didn't know that Tagalog had a perfective and how it is marked. See, that's exactly what I mean :< Grammar writers shouldn't assume that the reader knows everything about the language already.

Thus [the morphologic marker] in perfective formulations is realized in one of the following four ways, the particular realization that occurs in any given case depending upon the structure of the basic form of the verb, as specified in §5.20:

(1) the replacement of prefix-initial /m/ by /n/;
(2) the retention of the infix -um-;
(3) the addition of ni- or -in-;
(4) the addition of ni- or -in- plus the deletion of the suffix -in.

OK, so -in-/-ni are probably not the goal focus here, but the perfective aspect markers indeed. So the causative focus should actually work like I thought at first, before I got confused with those additional affixes. Neat. So a more correct glossing of the example sentences above probably is:

Code: Select all

I-  ni-  luha       ni= Nena ang= pagkawala ng=  alahas  niya.
CF1-PFV1-shed.tears A=  NAME T=   loss      GEN= jewelry 3S

Ik<in>a-  luha       ni= Nena ang usok.
CF2<PFV2>-shed.tears A=  NAME T=  smoke

I'd still like to know how just the phrase Nena sheds tears would look like, however, just to be able to compare (and maybe give an example in the blog article on this that I've planned).

Here's how my conlang does it; structural similarities should be apparent:

Code: Select all

Sā sip-ye  ang= Nena sundal teping-ye-na  yena.
CF cry-3SF A=   NAME loss.T jewel -PL-GEN 3SF.GEN
'The loss of her jewels makes Nena cry.'

Sā sip-ye  ang= Nena runu.
CF cry-3SF A=   NAME smoke.T
'The smoke makes Nena cry.'

Sip-ye  ang= Nena.
cry-3SF A=   NAME
'Nena cries.'

I don't get why people tend to think what their conlang does is pertinent when the question is specifically about some natural language

Background information on why he is asking the question in the first place I guess.

I complained on IRC that my linguistics questions here often do not get answered because everyone may give examples of how their conlang deals with stuff rather than answering my original question. It's great if a question makes people think how they'd deal with some feature or a related issue themselves, but it's crap if nobody cares about giving an answer to the original question.

[zompist]

Tagalog is weird. I'm pretty sure the key words break down as follows:

i-ni-luha
patient trigger - perfective - weep

trigger (marked by ang) is "loss"

ik<in>a-luha
reason trigger with infixed perfective - weep

trigger is "smoke", i.e. it's the reason for the tears

Ika- isn't the normal Tagalog causative, which is marked by pa-. These seem more like one more type of trigger than a causative per se.

I think the two sentences are different because a different trigger type is used. I- is used for several different triggers, including patient, benefactive, and instrumental.

(Too bad you didn't ask a week ago, when we had a Filipina caregiver at my parents' house...)

[Jipí]

A friend of mine is a Tagalog speaker, albeit not completely fluent. I could ask her next time I talk to her.

Your help notwithstanding, Zomp, but here's page 313 of the grammar, where it states all the causative-focus markers and where you can find the two examples I quoted: http://books.google.com/books?id=E8tApL ... #v=onepage. Maybe I should've quoted that bit about the causative-focus markers, too, for clarity, instead of just summarizing. If the sentence with "loss" had patient/goal-focus, it would say ^?'She wept the loss', which is not consistent with the explanations in that chapter, as it clearly says that i- in that case (and ika- in the other) marks causative-focus. Granted, in my research I read that intransitive verbs used in a transitive way may be understood as causatives in a number of languages including, for example, English, but I don't think that's the case here. Anyway, it was the perfective markers that confused me, because at least -in- is also used as a goal-focus marker IIRC.

[zompist]

Well, at least that confirms that the i- and ika- forms are different. Lots of overuse of i- there.

I got the "reason trigger" bit from Wikipedia, but I'm guessing that it's the same thing as what your grammar is calling "causative focus".

[Jipí]

Oh, another thing. The topic marker on the verb is usually understood to mark its referent NP as being definite, as far as I've understood other things I read. So in the case of the causative-topic marker on the verb that both examples above exhibit, what would happen if the causer wasn't definite, e.g. what if this NP was "a man", not "the man"? Would that NP be oblique (like an adverbial) then? Does this have any notable semantic implications?

[Jipí]

From looking things up in a dictionary I could gather that ni marks names as being in the actor role in goal-focussed clauses – which, if true, is rather interesting because according to those two articles on causative typology I've read that's kind of unusual (but is basically exactly what my conlang does!). Apparently you'd expect the causee (i.e. the actor of the incorporated clause) to be demoted in some way. I suppose that this is not necessarily true if a language marks case based more on semantics than syntax, which seems to be the case here and is definitely true of my conlang.

This thought of mine may actually be wrong, since ni + proper noun corresponds to ng + regular noun, and ng marks both non-topic actors and non-topic goals/patients/objects (from Schachter & Otanes 75):

Code: Select all

G<in>awa ng      modista    ang baro.
make<GF> A.NON-T dressmaker T   dress.
'The dressmaker made the dress.'

P<um>ili   ng      estudyante ang titser.
choose<AF> G.NON-T student    T   teacher.
'The teacher chose a student.'

I've also just seen that p. 314 of the grammar contains one pair of transitive example sentences at least:

• (1) With actor-topic:
  Nagbili sila ng kanilang kasangkapan dahil sa kahirapan nila.
  'They sold some of their furniture because of their poverty.'
  
  (2) With causative-topic:
  Ikinapagbili nila ng kanilang kasangkapan ang kahirapan nila.
  'Their poverty caused them to sell some of their furniture.'

Now, after tediously flipping back and forth in the grammar (because who needs interlinear glosses!!!), I came to the following speculative analysis:

Code: Select all

(1) With actor-topic:
Nag-   bili sila kanila=ng   kasangkapan ...
PFV\AT-sell 3P.T 3P.D  =POSS furniture
'They sold some of their furniture …'

(2) With causative-topic:
Ik<in>apag-bili nila     ng    kanila=ng   kasangkapan ang kahirapan nila.
CT<PFV>-   sell 3P.NON-T NON-T 3P.D  =POSS furniture   T   poverty   3S.GEN
'Their poverty caused them to sell some of their furniture.'

I marked nila with '3P.NON-T' and '3S.GEN' to reflect different contexts. The ng particle and its respective pronoun forms mark both possession and non-topic NPs. I have no idea why it says kanilang kasangkapan there, but for all I could find out, kanila is the directional/sa form of the 3pl pronoun. This may likely have something to do with possession marking as well, however.

The actor-topic 3pl pronoun sila in (1) becomes nila in the causative version of the sentence (2). That is, where the pronoun NP used to be marked as an actor (mag-…), it assumes its non-topic form. The NP with "their furniture" appears to be marked the same, though. Whether the causee is treated as an actor or a goal/patient/… is still not obvious to me. It simply doesn't seem to matter morphologically.