[Tlhingan-hol] Objects, direct and indirect

[De'vID]

De'vID:
>> We would obviously interpret {ghaH vIghojmoH} as "I teach him" and not
>> "I teach about(?) him", "I teach him as a subject of study" (like I'm
>> giving a lecture on Kahless).
>> [...]
>> If I saw {SuvwI' qIpmoH ra'wI'}, I'm not sure whether {SuvwI'} hit
>> someone or someone hit {SuvwI'}.

DloraH:
> Can you say "obviously", because you even go on to give an example of another possible
> interpretation?

Because {ghoj} is defined as "learn" and not "learn about". You can't
(I presume) say *{qeylIS ghoj ghojwI'} "the student learns Kahless".
The object of {ghoj} is that which is learned, and "Kahless" cannot be
learned. (I am of course talking grammatically, though of course
spiritually speaking the greatness of Kahless is also unfathomable.)
You can learn a lesson, a language, a skill, and so on, but you cannot
learn a person. Therefore, you cannot "cause-to-learn [i.e., teach] a
person-as-theme", though you can teach a person-as-patient, if I
understand SuStel's terminology.

DloraH:
> How would we say that you were teaching about him?

ghojwI'pu' vIghojmoHtaHvIS, qeylIS vInaD.
jIghojmoHtaHvIS, qeylIS lut vIja'.
SoQ vIjatlh. qeylIS yIn bop.

SuStel:
> Consider the context of the sentence. The pronoun will have an antecedent
> which you can presumably identify. That should be plenty of information to
> go on.
>
> Otherwise, no, you can't tell. {tlhInganpu' vIghojmoH} could mean "I teach
> Klingons (about something)" or "I teach (someone) about Klingons."

I'm surprised that you would say this.

In the first post of this thread, you wrote:
> Apparently you can't say *{qeylIS vIjatlh} "I speak to Kahless," but you CAN
> say {SoQ vIjatlh} "I speak a speech; I give a speech." You can also say
> {qajatlh} "I speak to you" and even {SoQ qajatlh} "I speak a speech to you;
> I give you a speech."

I think it's fairly obvious (though, apparently not to other people)
why the above is true. {jatlh} is "speak, say", not "speak to" or
"speak about". This constraints what nouns can sensibly be its object.
*{qeylIS vIjatlh} can't mean "I speak to Kahless". If it meant
anything, it would be "I speak Kahless", "I say Kahless". (But this is
different than "I say 'Kahless'." {qeylIS jIjatlh}.) But you can say
{SoQ vIjatlh} "I speak a lecture".

So {qeylIS vIjatlhmoH} can only mean "I cause Kahless to speak", not
"I cause some unspecified person to speak Kahless". I can even say
{SoQ qajatlhmoH} to Kahless; maybe I encourage him to rally the troops
on the eve of a battle.

I think similar constraints apply to {ghoj}. I can't say *{qeylIS
vIghoj} "I learn Kahless". I can say {qeylIS vIghojmoH}, which can
only mean "I cause Kahless to learn", "I teach Kahless". I can even
say to Kahless, {to' qaghojmoH} "I will teach you tactics".

Both of the following are possible:
{tlhInganpu' vIghojmoH} "I teach Klingons"
{tlhInganpu'vaD to' vIghojmoH} "I teach Klingons tactics", "I teach
tactics to Klingons"

I think André's analysis with the dative shifts is essentially
correct. The only hitch seems to be how to "get from" {tlhInganpu'vaD
vIghojmoH} "to" {tlhInganpu' vIghojmoH}, because they look
superficially similar, but as he's said, if you draw a syntax tree,
it's clear that what's happened is very different.

I think the justification for the apparent gap in André's analysis has
been provided by Marc Okrand, whom I am going to quote again:
> Since the object of jatlh is
> that which is spoken, and since "you" or "I" or "we" cannot be spoken (and
> therefore cannot be the object of the verb), if the verb is used with a
> pronominal prefix indicating a first- or second-person object, that first
> or second person is the indirect object.

Forget for the moment that we derive a rule, the "prefix trick", from
this, which is known only to apply to first- and second-person
objects. What Okrand is saying here is that, since *{SoH qajatlh} "I
speak you" is impossible, the way you'd analyse this is that it must
have been derived from {SoHvaD jIjatlh}. Applying the same idea,
*{tlhInganpu' ghoj} "he/she learns Klingons" is impossible.

Thus, if you see {tlhInganpu' vIghojmoH}, it must have been derived
from {ghoj tlhInganpu'}, that is, {tlhInganpu' vIghojmoH} can only
mean "I teach Klingons"; it could not have been derived from
{tlhInganpu' ghoj}, and thus it can't mean "I teach (someone else) to
learn Klingons". (There is only one possible syntax tree for the
sentence. It is not ambiguous.)

-- 
De'vID