tlhIngan-Hol Archive: Thu Jan 29 08:53:26 1998
[Date Prev][Date Next][Thread Prev][Thread Next]
Re: Locatives and {-bogh} (was Re: KLBC Poetry)
-----BEGIN PGP SIGNED MESSAGE-----
>Date: Tue, 27 Jan 1998 18:47:18 -0800 (PST)
>From: "Andeen, Eric" <[email protected]>
>
>Qermaq has argued that in a {-bogh} clause, any noun with a type 5
>suffix is the head noun, and charghwI' has argued (passionately: reH
>nong charghwI') that the only type 5 suffix that makes any sense is
>-'e'. I mostly agree with charghwI' on this one.
I don't think there's much room to argue with charghwI' on this point.
Okrand himself said that he couldn't see the head-noun of a relative clause
being anything other than the subject or the object, and thus if the head
is flagged at all, it would be with -'e'.
The {meQtaHbogh qachDaq Suv qoHpu' neH} example is not a counterexample to
this: the -bogh clause is {meQtaHbogh qach}, and the head noun is the
*subject* of that clause. The reason it has the -Daq on it is because it
functions as a locative in the *matrix* clause (the surrounding sentence).
This is admittedly weird, to have this noun wearing two hats (a subject in
one sentence and a locative in another--though we don't bat an eye at the
subject of a relative clause being the object of a matrix clause. Probably
because subjects and objects are both unmarked in Klingon, while everything
else is marked with suffixes), but does not comprise any contradiction to
what we know. It says that the head-noun of a -bogh clause--its subject or
object only, mind you--can fill other roles in the matrix, that's all. In
fact, it further strengthens the case that strictly speaking, an outside
linguist describing the Klingon language (NOT respecting the traditional
classical analysis) would probably call most of the type-5 noun suffixes
(with the possible exception of -'e') postpositions and not suffixes (they
migrate to ends of noun-clauses modified by verbs, etc). (Please, don't
get me wrong. I'm not arguing with Okrand. We call -Ha' a rover when it
doesn't rove and all. Just saying that from a linguistic perspective, -Daq
is really more properly a postposition, not a suffix).
>Generalizing a bit, -'e' is the ONLY type 5 suffix that can EVER work to
>mark the head noun if the noun comes before the verb: if the noun has a
>type 5 suffix other than -'e', it cannot be the object. Even if the
>rules on type 5 suffices are relaxed a bit in {-bogh} clauses, there is
>still no reasonable way to make it the object: how could you tell
>whether it is the object or just an extraneous noun. What would it mean
>if it were the object anyway? <qachDaq Qaw'bogh nawlogh SoplI' HoD>
>could mean, if we accept -Daq on an object "The captain was eating in
>the building which the squadron destroyed", turning the whole {-bogh}
>clause into a locative. However, I firmly believe that the locative
><qachDaq> would glom on to the nearest verb and act (as it should) like
>a locative, and the correct meaning of this sentence would wind up as
>"The captain is eating the squadron which destroyed (no object) in the
>building". Yummy. Trying to put a type 5 suffix other than -'e' on the
>object of a {-bogh} clause doesn't work because there's no way to tell
>that the noun isn't just a free floating type 5 suffix noun like it
>appears to be.
This is a significant point. I, too, believe that I would be far more
likely, at first hearing, to think "The captain was eating the squadron
which destroyed in the building" when I heard {qachDaq Qaw'bogh nawlogh
SoplI' HoD}. At least briefly. The ambiguity I'd see would be between
whether or not the locative belonged inside or outside the relative clause
(i.e. whether it's "The captain was eating (the squadron that destroyed in
the building)" or "in the building, the captain was eating the squadron
that destroyed")--an ambiguity MO has told us exists in Klingon (in one of
the interviews he said that the joke about "I shot an elephant in my
pajamas... and how he got in my pajamas I couldn't guess" would work in
Klingon, at least as regards ambiguity of relatives). I almost certainly
would not think "the captain was eating in the building which the sqadron
destroyed," {meQtaHbogh qachDaq} example notwithstanding. This reading
doesn't contradict the edict against oblique (not subject or object)
head-nouns, since the building is still the object of the relative
clause, but it just doesn't seem too likely. There are too many
syntactically sensible other alternatives (and you can come up with many
cases in which the other alternatives are also semantically sensible), and
- -Daq-flagged noun screws up the structure so royally. Consider this: I'll
diagram some of the sentences we have looked at and mark where the relative
clauses and main clauses begin and end. I hope I can do it clearly:
(1a)
rel clause
/--------------\
qachDaq Qaw'bogh nawlogh SoplI' HoD
\---------------------------------/
main clause
This is the meaning "In the building, the captain was eating the squadron
which destroyed (stuff in general)." Note how the main clause is
interrupted by the relative clause, but contains it whole; the clause
simply drops into a place inside it. Neat recursive containment, the
mental stack just pushes the "qachDaq" information in, parses the relative
clause, pops back, and all's well--and linguistically brains really DO work
with (fairly shallow) stacks, apparently.
(1b)
/----------------------\
qachDaq Qaw'bogh nawlogh SoplI' HoD
\--------/
This is the meaning "The captain was eating (the squadron which destroyed
in the building)." Strictly speaking, the main clause bar should include
the whole sentence, but that wouldn't be very informative. Basically, the
whole relative clause exists first, gets parsed and remembered, and then
passed as the first element in the main clause. Also continuous, actually
simpler than above. Here there's no mixing of clauses, and no
interruption. The other one has a small interruption, but just one, and
for a whole notion.
(2)
/-------------\
meQtaHbogh qachDaq Suv qoHpu' neH
\----------------/
This one IS intriguing. But put this way we see it has a certain amount of
logic to it. The relative clause is parsed, and passed as the first
element of the main clause. Neither clause is broken or discontinuous,
all's quite clean and sensible. The only weird thing is that the
clause-break actually occurs mid-word, between qach and -Daq. However, if
you think of -Daq as a postposition, that's not really so serious. It's
strange, and it taught us something new about Klingon, but frankly it's
something I don't mind learning, and now that I've thought it through, it
works.
Now let's look at the latest:
(1c)
/--\ /--------------\
qachDaq Qaw'bogh nawlogh SoplI' HoD
\-/ \--------/
(This is the meaning "the captain was eating in the restaurant which the
squadron destroyed")
Now do you see why this is ugly? The clauses are horrendously
discontinuous: the {qach} belongs to the relative clause, followed by
{-Daq} which doesn't... and then comes the rest of the relative clause!
That -Daq is a main-clause intrusion jammed in the middle of the relative
clause. This isn't the kind of interruption that languages allow; the
brackets don't nest, so to speak. I can't think "OK, parsing clause... ok,
remember where I was in this clause, parse relative clause... finished with
that, where was I, continue...": it's "parse clause... parse relative
clause... remember where I was in relative clause, parse some main
clause... back to the SAME relative clause again, parse more..." Ugh. This
isn't stack-compatible, and we DO think in terms of stacks.
This introduces some seeming assymmetry, in that objects of relative
clauses can "wear two hats" while subjects can't. But that's due to
Klingon's rigid word-order, which specifies assymettry: subjects and
objects can go at the end or beginning of a clause (respectively), but
oblique objects can only go at the beginning.
Actually, if I wanted to say "the captain was eating in the restaurant
which the squadron destroyed," I think at this point what I'd like to see
is this:
/-------------------\
?qach Qaw'bogh nawloghDaq SoplI' HoD
\------------/
Admittedly ambiguous between that meaning and "The captain was eating in
the squadron which destroyed the building," but that's normal head-noun
ambiguity of relative clauses, which we know exists in Klingon (cf. {Hov
ghajbe'bogh ram rur pegh ghajbe'bogh jaj}).
This is definitely speculative and controversial, and I wouldn't use it
without Okrand's say-so, but frankly it makes sense to me, despite its
seeming illogic. It relies rather heavily on the view of -Daq as a
postposition, a view which is unproven, though supported by the grammar we
know so far.
I'm going to regret mailing this, since I think it would make a great
HolQeD article. Whoever roundtables this, can we get this letter in?
~mark
-----BEGIN PGP SIGNATURE-----
Version: 2.6.2
Comment: Processed by Mailcrypt 3.4, an Emacs/PGP interface
iQB1AwUBNNCz8MppGeTJXWZ9AQFJ2wMAiYnOwaX+ocrFhr9JgZKWAbJtCOfY+m0g
cjAZnoBV81tW7Tc1WAnSFECE6k4mu3mqwi7ZKdLtPJS7q7jbLM4W/8tI99OBT2QQ
91JhW/tP9BVHn/HKRL83A2KfUuzOxcvA
=Zc3g
-----END PGP SIGNATURE-----