Theiling Online    Sitemap    Conlang Mailing List HQ   

Tagalog - response to takatunu (was Re: Help: Ergative VOS

From:takatunu <takatunu@...>
Date:Tuesday, December 23, 2003, 20:51
JR <fuscian@OPTONLINE.NET <mailto:fuscian@...>> wrote:

Sorry for the delay, again.
I am honoured of being specifically addressed. However I believe that my posts along the "Triggeriness" thread responded long ago to all your questions and arguments. It is nevertheless a pleasure for me to confirm below. <<<< You're right, that's not a clear example at all. A better example of an agent would be "Nagtuturo ako" = 'I teach.'
>> Na-laglag ang libro. >> Intr-fell Abs book >> The book fell. > Same as above but as you are an active language speaker you would expect
> to be non-volitional.
It is non-volitional no matter what language you speak, and what else would anyone expect? The point is, it is not an agent in any sense, and it is clearly a patient.
The point is, it is not a "patient" in any sense. It is clearly a "sujet" (in French) or an "experiencer" (in English). <<<< The definition of particular thematic/semantic roles is universal, not language-specific, or [language-type]-specific. Volition is considered to be a contributing factor to agent-hood (see < n> ticRole.htm). That said, particular languages may group these roles slightly differently (e.g., conflating or separating things like "theme" and "experiencer"). Palmer calls the latter, in contrast, "grammatical roles."
That role system you apparently believe is *the* role system is only one among other proposed options. "Experiencer" is not a "slightly different" role and "theme" is a completely different thing coupled with "rheme" that has nothing to do with a role, except if you talk about the semantic "topic" (of the news for instance), in which case your claim that it's a "conflation" of an agent or patient role is only one more evidence that it's time for you to vary your readings. Volition is or is not a contributing factor to agenthood depending on what your definition of the role "agent" is (read below). Languages don't group universal roles "slightly differently", only linguists call them and group them differently. Palmer is not God, SIL is not the World Academy, verbal voices are not the only basis for languages. Certain native american languages tag as "patient" either the object of a verb or the subject of a verb whose subject is supposed to be deprived of will and tag as "agent" the subject of a verb whose subject is supposed to act on purpose (like in certain native american languages). The subjects of the same verb may swap roles according to their own semantic definition: "The man bows." "The bough bows." "The caterpillar bows." Other languages tag the subject of a verb whose subject is supposed to change the state of the object an "agent" and this object a "patient." Then you might get an "agent" deprived of will: "The nightmare frightened the man." Now there are also verbs of whom both subject and object are deprived of will, don't change their state and equally "experience" their relation: "The pot contains water." In such case, which one is the "agent", the "patient", the "experiencer", the "focus" or the whatever and according to what creteria? Really, the water is the content of the pot and the pot is the container of the water. Neither is experiencing more or willing more or causing more. A trigger system deals with all arguments in the same way and therefore is not underlied by some ur-universal verb-agent-patient system, but is a completely different system as Christophe tried to explain to you-to no avail apparently. To get this much, you need to step out of your verbal box and visualize each of your English verbs as a little play played by a group of actors, some of them endowed with will and some not, some changing in nature or state and some not-but without thinking that one is more important than another. Whether these roles really are *the universal* semantic roles is besides the point and anyway still depends on whether there are universal semantic roles at all. The fact that we agree to understand what other people call "agent" or "patient" or "experiencer" in their posts on this list does not mean that everybody agrees on how they are used. <<<< If you call it an actor, it seems like you're losing any connection to semantic roles, and thus to the role-marking-on-the-verb theory that I've been arguing against.
Forget about "actor". I tried to translate the French concept of "actant" collectively referring to "agent" and "sujet". The whole point of this thread and the only thing I was arguing against is the claim that the underlying structure of a trigger language is a verb-agent-patient system. That's a nonsense. I have no idea why you accuse me of ignoring "semantic roles". Ignoring semantic roles first implies that people agree on what and how many these roles are and how they work, which is not the case yet: you call a "patient" what I call an "experiencer" in the first place regardless what the bibliography that you generously offer in this post may summon. The role in trigger languages is not "marked on the verb" because their predicate is not a verb. "Nagtuturo" is not a verb. No more than its Hebrew counterpart "melamed" is. No more than the English noun "the teacher" is. All of them are what is called in French a "substantif" because they "substantiate" (impersonate) the predicate of the clause. That is what a trigger system is all about: Each argument may "impersonate" the predicate without needing a verb. There is as well a substantive meaning "the action/fact/process of teaching", but no verb "to teach" as such. Certain linguists conveniently define a set of verb-agent-patient roles by reference to verbal voices. Good for them. Now I gather from this thread that some of them label each trigger argument with these roles. This is presumably of not harm--if any use--since obviously the number and identity of arguments in a clause usually keeps being the same whatever the language the clause is translated in: The eating cat keeps being a cat and the eaten mouse a mouse. However, claiming that ergo a trigger predicate is truely a verb and that a trigger language is truely ERGATIVE (and *that* was the claim that triggered Chistophe's post, if I may this poor pun) is as much a nonsense as claiming that an ergative language is truely a trigger language. Verbal voices make up one of several possible systems expressing one predicate and one or more arguments, which is itself a way to express one or several pairs of entity+behaviour. Ergative (erg-abs) and active (nom-acc) languages express the predicate as a verb ("to teach") or a compound verb+substantive ("to be a teacher"). However, other languages work differently in whole or in part: They may express all or certain predicate-argument pairs as either: (i) the equation of a substantive (the teacher) and another substantive (me) (ii) or the attribution of a substantive (teaching, teacher) to another substantive (me). In this system, the predicate substantive is possibly linked with the argument substantives by juxtaposition, trigger tag, copula, you name it. This system subsists in languages based on a verbal voices: Hebrew present tense, French epithete adjectives, Japanese adjectival nouns, Hawaiian nominal clauses, etc., subsist and are solid parts of these languages.


Isaac Penzev <isaacp@...>