tlhIngan-Hol Archive: Fri Nov 05 14:22:37 1999

Back to archive top level

To this year's listing



[Date Prev][Date Next][Thread Prev][Thread Next]

Re: the scope of {-be'}



On Thu, 4 Nov 1999 22:48:47 -0500 Alan Anderson 
<[email protected]> wrote:

> [regarding {maqarbe'chu'}]
> 
> jIja'pu':
> >That pseudo-ambiguity is just as present in a simple sentence like
> >{maqar}.  Maybe it's the "we" part that's the focus of the assertion.
> >But it's more likely not to be, since we have explicit pronouns that
> >are generally used to provide that focus.
> 
> ja' charghwI':
> >But there is no way to apply negation to a pronoun unless it is
> >being treated as a verb. To use that in this kind of
> >construction, we have to go to the perversity of:
> >
> >qarwI'pu' maHbe'.
> >
> >Yuck.
> 
> maqarbe'chu' maH.  It is *we* who are completely inaccurate.
> I don't see why you reject this as a valid way of emphasizing
> the subject in the negation of "we are absolutely accurate".

I hate to point this out to you, but in your example, {maH} is 
AFTER {-be'}, so by your globalized rule, you are explicitly 
removing "we" from the negation, since it negates that which 
preceeds it and does not negate that which follows it. This is 
the kind of hogwash I try to avoid by making the far simpler 
observation that, except for that single example which predated 
{batlhHa'} as a valid word, in ALL cases, the local 
interpretation of negation definitely works, while the 
globalized interpretation only works some of the time, making it 
pretty much useless.
 
> jIja'pu':
> >...The most likely intent is that the rover is modifying only
> >the suffix it immediately follows, and the adverbial is not being
> >negated.  I'd go so far as to say that the presence of another suffix
> >after the rover is a very powerful indication that the {-be'} isn't
> >intended to negate the entire statement; if it were, it would come at
> >the end of the verb.
> 
> ja' charghwI':
> >That is a very interesting and convoluted argument without any
> >justification.
> 
> Pardon?  I just said that if a suffix appears after a rover, the
> rover obviously isn't intended to apply to the entire statement.
> What's convoluted about that?  My justification is the trivial
> observation that rovers apply to what comes *before* them, and
> thus if something in a statement comes *after* one, the rover
> cannot affect the entire statement.

Like the {maH} in {maqarbe'chu' maH}?
 
> >So, since it does seem absurd that suffixes following {-be'}
> >should be somehow excluded from a globalized {-be'}...
> 
> You're either misunderstanding or misrepresenting the idea.  It's
> not absurd at all to exclude suffixes after a rover from the effect
> of that rover; that's how rovers work, after all.  I think you're
> overreacting to the term "global" and pretending that there's some
> argument that a rover can reach in both directions.

No. I'm merely saying that if a suffix applies to only that 
grammatical element immediately preceeding it, then that suffix 
can be easily used and easily interpreted in all cases, but if 
you instead insist that it acts like some kind of boundary 
dividing that which is negated (everything to the left) from 
that which is not negated (everything to the right), then 
suddenly, the somewhat arbitrary order of suffixes becomes 
artificially significant. The same is true of whether or not you 
choose to place a dependent clause before or after a verb with 
{-be'}. And what if the verb has more than one {-be'} in it? And 
what if there is a {-be'} in both the main clause and in a 
dependent clause? Does it make a difference then whether the 
main clause comes first or the dependent one? And what about 
{-meH} clauses that are forced to come before the main clause? 
Or what about idiomatic {pumDI'} clauses forced to come before 
the main clause?

The whole thing becomes a stupid mess. The accidental effects of 
this kind of short-sighted model of the way {-be'} works does 
nothing good for the language.
 
> >, you respond
> >by saying that it must be the case that if {-be'} is followed by
> >any other suffixes, it must have the local interpretation, but
> >if it is the last suffix on a verb, it probably uses the
> >globalized interpretation.
> 
> No, I merely said that the presense of another suffix after the
> rover is a very powerful indication that the {-be'} isn't intended
> to negate the entire statement; if it were, it would come at the
> end of the verb.  I had just stated that it's more likely that the
> rover is modifying only the immediately preceding thing anyway, so
> there's not real reason to look for a less likely interpretation in
> the first place unless the likely interpretation fails to make sense.

In this case, the English is straining to make sense.
 
> >By implication, this means that all
> >verbs with {-be'} and a Type 9 suffix must use only the
> >localized interpretation of {-be'} because {-be'} can't follow
> >Type 9 suffixes.
> 
> Where did I say that the "localized interpretation" must be forced
> by the presence of a subsequent suffix?  I certainly didn't intend
> anything of the sort.

I see our misunderstanding.
 
> >And you've also said that in simple cases like
> >{maqarbe'} the globalized version doesn't work so well, either.
> 
> I believe I said something about the two interpretations coming up
> with the same basic idea in simple cases like this.
> 
> >I wonder how many other exceptions that we need to make to this
> >rule before it starts to actually do anything useful.
> 
> I'm not seeing these "exceptions" that you think I'm putting forth.

[No meaningful response.]
 
> >If it really is the case that {batlh bIHeghbe'} is the only
> >cannon example of Okrand doing this, and this example was
> >created before Okrand introduced the option of using {batlhHa'},
> >then might we come to the conclusion that the idea of using
> >{-be'} to negate more than a single grammatical element
> >preceeding it is a bit outdated and move on?
> >[...]
> >My point is that given the extremely limited utility of this
> >small scope of statements, we would lose less by ignoring their
> >existance entirely than we gain by respecting them enough to
> >make a lot of otherwise clear statements suddenly ambiguous
> >because we can't tell the scope of {-be'}. There is nothing that
> >can't be said quite clearly with the local-only interpretation
> >of {-be'} which is so much better said with the globalized
> >{-be'} that can justify the muddiness of a language that can
> >never know exactly what it is negating,...
> 
> But there are so many other things that we "can never know exactly"
> that you don't seem to be worried about.  The order of suffixes is
> fixed, so how can we know that the speaker didn't intend something
> that would violate the suffix order if strictly translated from an
> English sentence?  Your {maqarbe'} arguments prompted me to show a
> basic uncertainty in {maqar} that exactly matches what you proposed
> as the basic problem with negating more than a single syllable.  It
> is even hard to tell exactly what a purpose clause is modifying if
> it comes before the object of a sentence -- is it the object, or is
> it the verb?

The suffix {-be'} gets a bit more usage than {-meH} and effects 
a lot more sentences, many of which are expressing extremely 
basic concepts. It is more important to be clear in these more 
common instances. I can find a way to cast around {-meH} 
ambiguities. If the globalized {-be'} is always considered 
valid, then I have no way to cast around problems this creates.
 
> The answer always seems to come down to the fact that this is not a
> mathematically pure language, and there's always going to be some
> measure of assumption in the interpretation of any utterance.  For
> fluent speakers who share a similar experience base, there's not a
> problem.  We deal with fundamental ambiguities in our native tongue
> constantly without even being aware of it.  So why make such a big
> deal about being confronted by a characteristic of natural languages?

Ambiguities tend to exist primarily in unusual statements or 
ones that try to represent fine levels of meaning that are 
relatively uncommon. The {-meH} problem is acceptable because it 
doesn't happen all that often. The fixed order of the suffixes 
is tollerable because the problem areas don't happen all that 
often. {wej} is annoying because it is right on the verge of 
being too common to ignore. There is more pressure to do 
something about it.

And {-be'} is more common than any of these. And THAT is why it 
is a problem that bothers me quite a bit. Anything this 
fundamental to basic expression of thought needs to be as clear 
as possible.

Your argument against my interest in clarity is a valid one when 
we are talking about locatives in relative clauses. While these 
are ugly and ambiguous to the point of becoming meaningless, 
they don't happen very often and you can always split the 
relative clause off to become a separate preceeding sentence 
with some the former head noun repeated with {-vam} on it in the 
second sentence. I was not tolerant enough, just because I saw 
something that could have been beautifully versatile, expressive 
and clear, and that was ignored in favor of something ugly, 
vague and nearly useless.

But that is its strength. It is so useless, nobody ever uses it, 
so it is not really a problem.

But {-be'} is QUITE useful. It is a necessity. The language 
needs to do it well.
 
> >especially since we can
> >now negate adverbs (a relatively new feature of the language).
> 
> [Saying "we can now negate adverbs" goes a bit too far, I think.  We
> have some "negated adverbs" at our disposal, true.  But Maltz didn't
> like {*vajHa'}, and there was some interesting debate over what the
> theoretical {*wejHa'} might mean.]

I'm curious about how you think these very same adverbs can be 
"negated" by {-be'} in a following verb more meaningfully than 
they can be negated by {-Ha'}.
 
> >> I too think the grammatical sogginess that this implies is...distasteful.
> >> However, I just don't see it as that big a deal.  I'm not troubled by
> >> the presence of hard-to-interpret constructions in English grammar.  I
> >> have no problem accepting that they exist in Klingon as well.  The only
> >> time I'd be upset is if someone relied on them without giving other cues
> >> to his intent.
> >
> >Why go to that trouble? Do you really think it is very Klingon
> >to have to say something and then explain what you mean by it?
> >This seems especially distasteful for something as elemental to
> >the language as negation.
> 
> I don't begrudge you your wish for clarity in the language.  But I do
> take offense -- slightly -- at your insistence that things which make
> for less clarity should be ignored.  Worse, you seem to want actively
> to attack any usage of them, making what appear to be grammatically
> motivated pronouncements over what is arguably just a matter of style.
> 
> I'll make a deal with you.  You go on reading and writing sentences
> using the assumption that rovers can only act on the morpheme that
> immediately precedes them, and I'll go on writing using whatever I
> think gets my idea across effectively.  We'll both be responsible,
> jointly, for the process of communication.  Okay?

Deal. Meanwhile, this didn't get started because of your usage. 
It got started because of a pronouncement of how grammar works, 
all spoken in English.
 
> -- ghunchu'wI' 'utlh

charghwI' 'utlh



Back to archive top level