tlhIngan-Hol Archive: Mon Feb 02 14:25:28 1998
[Date Prev][Date Next][Thread Prev][Thread Next]
Re: Locatives and {-bogh} (was Re: KLBC Poetry)
According to David Trimboli:
>
> Keep reading, I come up with something I think is really cool towards the
> bottom!
>
> -----Original Message-----
> From: William H. Martin <[email protected]>
> To: Multiple recipients of list <[email protected]>
> Date: Sunday, February 01, 1998 2:45 PM
> Subject: Re: Locatives and {-bogh} (was Re: KLBC Poetry)
>
>
> >Given your preferred method of applying Type 5 suffix to the
> >head noun of a relative clause even when it is the object, we
> >lose the ability to disambiguate the grammar. By your method:
>
> I, for one, don't claim that any Type 5 suffix can mark the head noun of the
> relative clause. However, it just happens to always work that way.
It works that way if the Type 5 noun is functioning as a Type 5
noun in the main clause. My point is...
Okay, I just tried to describe this and it got very convoluted,
so I'll try to be VERY careful. We know that you cannot have a
Type 5 noun which uses its Type 5 grammatical function within
the relative clause and is also the head noun. That doesn't
mean that a Type 5 noun can't function as such within a
relative clause. It just means it can't do that AND be the head
noun. Example:
qachDaq Sopbogh HoD puq vIlegh.
I'm not saying, "I saw it in the building which the captain's
child ate." [That would assume that {qach} was head noun of the
relative clause, qachDaq is the locative for the main clause
and {HoD puq} is a noun-noun construction.] Using ~mark's
tools,that would be:
/---...---------------\ Relative clause
qachDaq Sopbogh HoD puq vIlegh.
/------.................-----/ Main clause
I'm also not saying, "I saw the child in the building which the
captain ate." [That would assume that {qach} is the head noun
of the relative clause, {qachDaq} is a locative for the main
clause and that {HoD puq} is not a noun-noun construction.]
/---...-----------\ Relative clause
qachDaq Sopbogh HoD puq vIlegh.
\------............----------/ Main clause
In this case, {qachDaq} is being used within the
relative clause as a locative. There is no rule saying I can't
do that. It is simply that I cannot use it as a locative
within the relative clause AND ALSO USE IT AS THE HEAD NOUN.
Here, what I WANT to say is, "I saw the captain's child who ate
in the building." {qachDaq} is the locative for the relative
clause. {puq} is the head noun of the relative clause and
object of the main verb.
/---------------------\ Relative Clause
qachDaq Sopbogh HoD puq vIlegh.
\--------/ Main Clause
Please notice that what I want is actually the simplest grammar
of anything else here. Ocham's Razor would indicate that my
intended meaning wins, but by your suggestion, there would be
no way to indicate the difference between these very different
meanings.
If we use the Type 5 "postposition" idea and combine it with
Krankor's use of {-'e'} to disambiguate, we can have these
different meanings more clearly stated:
qachDaq Sopbogh HoD puq vIlegh.
I saw the captain's child who ate in the restaurant.
qach'e' Sopbogh HoDDaq puq vIlegh.
In the restaurant which the captain ate, I saw the child.
qach'e' Sopbogh HoD puqDaq vIlegh.
In the restaurant which the captain's child ate, I saw it.
Again, with ~mark's tools:
/---------------------\ Relative Clause
qachDaq Sopbogh HoD puq vIlegh.
\--------/ Main Clause.
/-----------------\ Relative Clause
qach'e' Sopbogh HoDDaq puq vIlegh.
\---...............-------------/ Main Clause
/---------------------\ Relative Clause
qach'e' Sopbogh HoD puqDaq vIlegh.
\---...................---------/ Main Clause
This would work to make these three very different meanings
very easy to distinguish. In fact, I now see yet another
meaning that could exist and be differentiated: "I saw the
captain's child in the building which was eaten.
/------------\ Relative Clause
qach'e' SopboghDaq HoD puq vIlegh.
\---...........-----------------/ Main Clause
Your choice for grammatical rules would make all four meanings
stated in the same one sentence. Mine makes the meaning clear
between them.
> The head noun is that which is going to have a place in both the relative
> clause and the main clause. If the head noun's role is locative in the main
> clause (it is forbidden for it to be locative in the relative clause -- ship
> in which I fled) then the head noun WILL receive a Type 5 suffix.
It is not forbidden for it to be locative in the relative
clause. It is only forbidden for it to both be locative in the
relative clause AND be head noun of the relative clause. That
is why your grammar falls apart, unless we do newly forbid the
use of a locative within the relative clause. If we do that,
then we lose the ability to express a LOT of different
thoughts, like the one I offered above. If, while standing in
the grocery store, I see the captain's child whom I had seen
earlier eating in a restaurant, I can't say, "I saw the
captain's child who ate in the restaurant," because you have
disallowed the use of a locative within a relative clause.
I see this as a problem.
> 'u' SepmeyDaq luSovbe'lu'bogh maleng.
> We travel in unknown regions of the universe.
I also see THIS as a problem. Yuck.
> {Sepmey} happens to be the head noun, and it happens to get the Type 5
> suffix. It will always work out that way when using any Type 5.
Again, you assume that the locative has to act within the main
clause and can't act within the relative clause, even if it is
not head noun. I don't understand where you got that
restriction.
> puqvaD qIpbogh yaS ghe'naQ vIbom.
> I sing an opera to the officer who hit the child.
You mean, "I sing an opera to the child who was hit by the
officer."
You also might mean, "I sing the officer-who-hits's opera to
the child." You also might mean, "I sing the officer's opera to
the child who was hit." Meanwhile, this would not be ambiguous
if the grammar more explicitly marked the boundaries of the
relative clause, as it would if you said:
puq'e' qIpbogh yaSvaD ghe'naQ vIbom.
I sing an opera for the child who was hit by the officer.
puqvaD qIpbogh yaS ghe'naQ vIbom.
I sing the officer-who-hits's opera to the child.
puq'e' qIpboghvaD yaS ghe'naQ vIbom.
I sing the officer's opera to the child who was hit.
Can't you appreciate how much clearer this is? There are SO
many potential confusions cleared up by marking the boundaries
of the relative clause with the Type 5 as postposition with
{-'e'} used to mark the head noun whenever it is not the same
noun that carries the Type 5 suffix. It is not that weird and
it is VERY MUCH CLEARER.
> This shows that I sing to the child. Since relative clauses cannot take
> oblique nouns (like those with {-vaD}), the {puqvaD} must be playing a role
> in the main sentence.
I don't accept that repeated presumption. That is definitely
NOT what Okrand has told us.
> But {qIpbogh yaS} seems to be a relative clause
> without a connection to the main sentence, until you see that it's head noun
> is {puq}. Thus, the head noun will ALWAYS be the one with the Type 5
> suffix, if there is one.
Not true. {yaS} could be the first noun of the noun-noun
construction {yaS ghe'naQ}, which is head noun of the relative
clause and object of the main clause. That is very much a
connection to the main clause.
> The Type 5 suffix is NEVER a requirement of the relative clause. It is a
> requirement of the main sentence (except {-'e'} nouns, which are entirely
> optional in any case).
I completely disagree. The Type 5 suffix should definitely be
allowed to function within the relative clause so long as that
noun is not the head noun of the relative clause.
> ><A>vaD <B>bogh <C>na' <D>taH E<Hey>.
> >
> >This could mean as you wish it to mean, "The apparent <E> is
> ><D>ing for the <A> which is <B>ed by the definite <C>." It
> >could also mean, "The apparent <E> is <D>ing the definite <C>
> >which <B>s for <A>." Without knowing what the words mean and
> >which of these two meanings makes more sense in context, you
> >can't tell what is being said here.
>
> Unless you can see the ship in which I fled, a more accurate description of
> the alternate reading would be "For <A>, the apparent <E> is <D>ing the
> definite <C> which <B>s." But this is just a little ambiguity which occurs
> often enough in Klingon.
>
> You're simply showing an example where it's not clear if <C> is the object
> of the main clause or the subject of the relative clause, or both. You're
> inventing this ambiguity.
>
> jatlh tlhIngan jIH
> He said, "I am a Klingon."
> The Klingon said, "Me."
>
> Which is it?
That argument could be created against ANY ambiguity. You are
proving nothing. Yes, ambiguity exists in the langauge. That
doesn't mean we should value ambiguity so highly that we clutch
every opportunity we can to preserve it.
> >You might interpret that as {-vaD} acting as {-'e'} within the
> >relative clause, but for me, that is not necessary.
>
> The {-vaD} noun, by process of elimination, is the ONLY possible candidate
> for the head noun. Since it's always true, it's not difficult to remember
> when parsing a new sentence.
Well, it was not true in your own example. The object of the
main clause could have been the head noun of the relative
clause. I think this can be quite common.
> >There is no
> >need for a universal rule for Type 5s to do this. Were that the
> >case, there would be conflict in my previous example because
> >{-vaD} and {-'e'} both exist in the same relative clause.
>
> If you wrote such a sentence, I'd assume that {-'e'} was topicalizing its
> noun, its original function. For example:
>
> puq'e' qIpbogh yaSvaD ghe'naQ vIbom.
> I sing an opera to the officer who hit the CHILD (as opposed to hitting
> someone else).
You could use the same empty argument against the Krankor rule
in any setting. Okrand and the rest of us have accepted the
idea that {-'e'} in a relative clause indicates the head noun
of the clause. If you accept that and accept that the Type 5,
which functions as link between the relative clause and the
main clause should follow the entire clause (postposition),
regardless of its head noun, then you'd know that the Type 5
applies to the whole clause which modifies the head noun,
linking it to the main clause.
> >The only new rule necessary arises when we want the object of
> >the relative clause to be the head noun of the relative clause
> >and the Type 5 noun in the main clause. That new rule combines
> >two old rules. One is that {-'e'} disambiguates the head noun of
> >a relative clause when both explicit subject and object of the
> >verb within the relative clause exist. This is the Krankor Rule.
>
> {puqvaD qIpbogh yaS ghe'naQ vIbom}. {puq} HAS to be the head noun, because
> it HAS to be the indirect object of the main clause as well as being the
> object of the relative clause.
/-----------------\ Relative Clause
puqvaD qIpbogh yaS ghe'naQ vIbom.
\------............------------/ Main Clause
This interpretation is actually simpler than yours, not
requiring one to ignore any Type 5 suffix within a clause while
interpreting the clause. This interpretation doesn't violate
any rules of grammar. The meaning this interpretation offers is
not remotely related to the meaning your interpretation offers,
and if you disallow this interpretation, you essentially say
there is no way to express this meaning at all, since this is
the ONLY way this meaning could be expressed.
> It is automatically disambiguated because of
> this. No extra rule is necessary.
>
> *********HEY, COOL!********
NOT!
> Is it possible that the "Krankor Rule" actually came about BECAUSE Type 5'd
> nouns must always be the head noun?!?!? Every other Type 5 noun must be the
> head noun by its very nature. {-'e'} was left out all on its own. Perhaps
> over time, Klingons started to use {-'e'} in the same way as all of the
> other Type 5 nouns. That is, they thought, "Well, since any other Type 5
> noun in a relative clause means it must be the head noun, let's use {-'e'},
> another Type 5, to distinguish the head noun when it's NOT an oblique part
> of the main clause.
>
> In other words, they saw
>
> puq qIpbogh yaS vIlegh
>
> and couldn't decide if it meant "I see the officer who hit the child" or "I
> see the child whom the officer hit." Because it wasn't being used as an
> oblique part of the sentence, there was no indication as to which one was
> right. So, they decided to add {-'e'}. This would work in the grammar of
> the sentence, yet would still fulfill the "Type 5s mean it's the head noun"
> pattern, even though it wasn't technically true. Now, of course, it's been
> used so much that it has become a permanent rule of the language.
>
> In other words, {-'e'} was the LATEST addition to the head noun
> disambiguators!!!!
>
> Wow, I LIKE this idea!!!!!!!!
wejpuH. The problem with it is that you are now claiming that,
following your model, when one places {-'e'} to disambiguate
the head noun of a relative clause, you are making that noun
the topic of the main clause. If {-Daq} and {-vaD} perform
their root function in the main clause, I'd expect {-'e'} to do
the same thing. I doubt you want that effect.
In fact, I suspect what you'd want is to have the {-'e'}
function within the relative clause rather than within the main
clause, which is exactly what I don't want to be eliminated for
{-Daq} and {-vaD} or {-mo'} or any OTHER Type 5 suffix.
They should be allowed to be used either within the relative
clause or the main clause. If (except for {-'e'}, which we
already know is a wee bit strange among Type 5 suffixes) they
are applied to the whole clause instead of to the head noun,
then the boundaries of the relative clause can be easily
clarified. This would be a good thing.
Trust me on this.
> SuStel
> Stardate 98089.0
charghwI'