Topic/Focus (was: Valentine's Day Translations)

Tim Smith wrote:
>At 10:16 PM 2/12/99 +0100, Kristian Jensen wrote:
>>Alright then, me too!
>>
>>In the languages I speak:
>>     English: I love you
>>     Danish: Jeg elsker dig
>>     Tagalog: Mahal kita
>>
>>In my conlang, Boreanesian:
>>     Kele'aihkuh kih
>>     lit. "I'm your lover/endearer"
>>or
>>     Ke'aihkih kuh
>>     lit. "You're my love/dear"
>>or
>>     Kuhke'aihkih
>>     lit. "O you, my love/precious!" (addressing the loved-one)
>>
>>Once again, Boreanesian is highly sensitive to topic/focus. There
>>are two ways of translating the English expression "I love you",
>>depending on whether "I" or "you" is the topic/focus. In the first
>>example, "I" the agent is topicalized/focused. In the next two,
>>"you" the patient is topicalized/focused. The third way is the
>>most intimate of focusing the patient.
>
>This is very interesting.  But I'm a bit confused about what you
>mean by "topic/focus".  It's easy to be unclear about these two
>terms because they're used in different ways by different writers.
Well, this is sorta the reason why I chose to use both terms: they
are used differently by different people. But I did not intend to be
confusing - oops! Another reason why I have used both terms is
because Boreanesian is indeed sensitive to both topic and focus.
BTW, what you have written below is essentially how I understand the
difference between topic and focus:

>But most of the recent linguistic literature that I've seen, they
>have very different, almost opposite, meanings. The topic is what
>you're talking about; it's old, given information.  The focus is
>the new information that you're trying to convey about the topic,
>or (as some people use the term) a part of the new information that
>you're singling out for special emphasis.  IIRC, Matt Pearson once
>explained the difference by referring to the "as for" construction
>and the cleft construction in English as examples of topic-marking
>and focus-marking, respectively.  In the sentence, "as for John, he
>went to the library", John is the topic; John is who we're talking
>about, and what we're saying about him is that he went to the
>library.  But in the clefted sentence, "it was John who went to the
>library", John is the focus. The speaker is presupposing that the
>listener already knows that someone went to the library, and is
>telling them who it was.  (Though of course those are both very
>marked constructions; most of the time in English we just rely on
>intonation and context to mark topic and focus.)
>
>By these definitions (which I think are pretty standard today,
>although they're not very precise), I have the impression that in
>Tagalog the trigger can be either the topic or the focus.  From
>some stuff I've read that wasn't entirely clear, it sounded to me
>as if the trigger is normally the topic, but can be made the focus
>by putting it before the verb.  Does that sound right to you?  (I
>gather you're a native speaker of Tagalog.)
I'm not a native speaker, but I happen to have a mother who is a
Filipina (Ilocana to be precise), and she speaks to me in Tagalog
and I get away with Taglish 8-). Anyways, what you have said about
Tagalog sounds right to me.

>And if so, does Boreanesian work the same way?

Boreanesian works in a similar way, at least in the sense of
changing word order. The story is a bit more complex though. The
differences are that (syntactically speaking) the predicate and the
trigger are always at the beginning and end of a clause
respectively, and that the trigger is always topic while the focus
can be said to always be the predicate. This may sound
contradicting, but word order can still be changed even when the
triggers and predicates don't. For example, "the animal is crippled"
can be translated as follows depending on what is trigger:

a) PREDICATE = "cripple", TRIGGER = "animal"
     <ngehu'    'ehsetingh>
     a cripple  the animal

b) PREDICATE = "animal", TRIGGER = "cripple"
     <setingh   'ehngehu'>
     an animal  the cripple

Notice that the predicate is always the first word while the trigger
is always the last word of a clause. In example a), the topic of the
sentence is an animal. The addressee is presumed to already have the
animal in mind. On the other hand, example b) would be used instead
if the animal was mentioned for the first time. It might therefore
be possible to interpret predicates as the focus; i.e., new
information in the discourse that one is trying to convey about the
topic (ie., the trigger). So back in example a), the predicate
"cripple" is new information that is conveyed about the topic
"animal", while in example b), the "cripple" is conveyed as being an
"animal". If all information is new, then example a) would be used
by default. The most time-stable concepts (anything most nominal and
least verbal) are more inclined to be understood by the addressee as
older information than less time-stable concepts. In other words,
less time-stable concepts can be understood as being more prone to
contain new information.

This system is essentially a reflection of how I understand the work
of the so called Functional Sentence Perspective linguists of the
Prague School. According to these scholars, every clause has two
parts; the part that refers to what the addressee is presumed to
already have in mind, and the part that adds some new information.

The situation in Boreanesian is further complicated when there are
more than one arguments. In the examples given, there is only one
argument. In situations with more than two or more arguments,
different trigger situations come into play where it is possible to
promote different arguments to topic by changing the trigger alone
without involving the predicate. Of course, both triggers and
predicates can be changed in such situations as well by nominalizing
(or adding volition to) verbs. If it sounds complicated, then its
because it is. So I won't bother with more complex situations here.
(Unless you want me to, of course).

You might be wondering why I would always consider the first word of
Boreanesian clauses as predicates - especially when it seems that
the first word looks a lot like a noun. The question is, what
features are present that always makes the first word a predicate,
and what features present that always make other words just mere
arguments? Well, its difficult to tease apart verbal and nominal
morphology in Boreanesian. Most words can be used as both a
predicate and/or an argument. Both can also be marked by the same
homophonous particle; the bound phase determiner. Lack of it marks a
phrase as unbound in phase. In predicates, the bound and unbound
phases are semantically correlated with the perfective and
imperfective aspects respectively. In arguments, bound and unbound
with definite and indefinite respectively. When applied to
Boreanesian clauses, the phase of the first word always indicates
the aspect of the entire clause, thereby making the first word of a
clause behave more like a predicate. Note for instance that in
examples a) and b), the first word is unbound so both examples would
be read as imperfective. The boundness in phase of the last word
says nothing about the aspect of the clause but rather, the
definiteness of the trigger.

How is _that_ for complexity in one area that arises from simplicity
in another? What do you think? I'd like some input since everything
is still quite experimental.

-Kristian-