tlhIngan-Hol Archive: Wed Jan 11 08:43:04 2006

Back to archive top level

To this year's listing

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

Re: transitivity

Terrence Donnelly ([email protected]) [KLI Member] 

--- QeS 'utlh <[email protected]> wrote:

> ghItlhpu' ter'eS, ja':
>> Every verb has at least one argument: the
> > subject, which is a 
> >valency of one.
> 
> In some languages, verbs of weather are arguably
> zero-valent.
>

Conceded.  In fact, Okrand's exclamation {SIS} (which
I had forgotten) is zero-valent.
 
> 
> >So, the answer to the question "Is {jISop}
> >transitive or intransitive?" is 
> >"Both: {Sop} is always semantically bivalent/
> >transitive;
> 
> Why couldn't the *semantic* valency differ between
> languages as well as the 
> grammatical valency?  Particularly with a verb like
> "to eat", where the 
> semantics could be focusing on the action ("he's
> eating") or on the object 
> ("he's eating the fish"). 

I do think that would vary from language to language,
but
the point is, i think, that the semantic valency of a 
particular predicate remains consistent within a
language
and doesn't change willy-nilly without some sort of
valency changing operation to trigger it. But the
point
as I understand it, is that the "full" semantic
valency of
a verb will contain all the arguments that a language
considers necessary to prototypically express its
meaning, 
and that grammatical valency will control which need
to be expressed in a given phrase. So your examples of
the changing
valency of "eat" above are only grammatical
adjustments,
not semantic.  If a language really wanted to make a
semantic distinction between focussing on the action
or
on the object, they would be two separate verbs
(albeit maybe
closely related).

>And what do you do about
> {yIn} "to live"? Judging 
> from what Okrand has said in previous interviews,
> {yIn} is underlyingly 
> bivalent, although in practice the "object" of the
> action is hardly ever 
> expressed. It could be argued that any non-stative
> verb in Klingon is 
> theoretically able to take an object (including such
> fundamental 
> "intransitives" as {Qong} and {yIt}).
> 

I don't follow this.  When did Okrand say that {yIn}
is bivalent?  Why do you say that any non-stative can
take an object? For that matter, why do you call
{Qong}
a non-stative?

> >There are no semantically trivalent verbs in
> >Klingon,
> 
> Depends on how you define "trivalent". Following
> your argument, I would have 
> said that "to give" is semantically trivalent
> (taking the underlying 
> arguments as Agent, Patient and Goal), but due to
> the inherent bivalent 
> limitation on licensing of arguments in the Klingon
> verb, only two of those 
> arguments ever surface: the agent as subject, and
> the patient as object. 
> However, with the prefix trick, the verb licenses
> the Goal argument instead, 
> allowing the Patient to stand in normal object
> position syntactically and 
> the Agent to be, well, the Agent.
> 
> >but the prefix trick is a grammatical valency
> >increasing
> >operation: {paq qanob} "I give you a book"! Fun
> stuff!
> 

I would have to say that the fundamental bivalent
nature
of Klingon grammar (nouns can be only subject, object
or "everything else") argues against predicates being
seen
as semantically anything more than bivalent (why have
an
argument that can't normally be expressed?).  The
prefix 
trick might be considered a mechanism to let that
trivalency express itself, but it is of such limited
applicability (subject
and/or indirect object can only be 1st or 2nd person,
object
can only be 3rd person) that i think it truly is a
"trick":
something clever imposed on top of normal grammar
(almost as
if someone were trying to come up with an explanation
after
the fact for an embarrassing slip of the tongue!).

>Very fun. But also, very hard to resolve... {{:)

Only if you make it so! 8+)

> 
> QeS 'utlh

-- ter'eS

Back to archive top level