Re: (chat) Yoda's word order

At 8:18 PM +0200 05/28/02, Christian Thalmann wrote:
>Dirk Elzinga wrote:
>
>>  Since 'finden' is the "information-carrying verb," the sentence is
>>  properly seen as SOV with a second position auxiliary, which is
>>  obligatory in main clauses for this tense/aspect combination. In that
>>  respect, it is not unlike Uto-Aztecan languages like Luiseño and
>>  Tohono O'odham, which must also place tense/aspect information in the
>>  second position of the sentence. The difference is that lexical
>>  (i.e., information-carrying) verbs in German may carry all of the
>>  tense/aspect information themselves for certain tense and aspect
>>  combinations and do not require auxiliary support in those cases.
Before I go further, two caveats. 1) I'm not a syntactician, but I am
interested in Germanic in general (as well as Uto-Aztecan). 2) I have
no great stake in proving the correctness of the analysis I
summarized in my last post. It makes sense to me, and it shows
interesting similarities to an otherwise unrelated language family
(Uto-Aztecan), which I feel is an argument in its favor.

>Still, the majority of German sentences contain only one single
>verb, so the multiverbal construction should be viewed as an
>exception, not the rule.
Except that the construction under question *is* the rule when
periphrastic tense/aspect categories are called for. Just because
they are less common doesn't make them any less rule-governed.

>Furthermore, the V2 rule (finite verb at second position in the
>sentence) is so well-established an consistent in German that it
>seems very much forced to squeeze it into a SOV cage.
But I'm not squeezing it into a SOV cage; tense/aspect categories are
realized by a second position element. If that is the lexical verb,
you have surface V2 order. If tense and aspect are realized by an
auxiliary in second position, the lexical verb is final.

>To treat
>the infinite verb like the "true" verb sounds like an Americanism to
>me; seeing as English has very little typographical distinction
>between finite and infinite verbs, the average American probably
>considers the distinction far less important than a native speaker
>of German.
In my reply to your post, I quoted you as saying that 'finden' was
the "information-carrying verb" and supplied a more concise label
(lexical) to express the same idea. I'm not aware that the position
is particularly "American", nor am I sure what that is supposed to
mean in this context.

I'm also not sure how "important" either German speakers or English
speakers actually do consider the constructions. When English
speakers are made aware of the distinction between finite and
non-finite verbs, they can pick them out pretty consistently in spite
of the paucity of inflectional morphology; at least, that's been my
experience when teaching them. At any rate, the argument by "native
speaker intuition" carries rather low weight for me.

>Mark's homepage claims that "German is basically SOV, but a finite
>verb (anything but a participle or an infinitive) appears after the
>subject in a main clause."  This scheme seems to suggest that the
>infinite verb is an invariant of German sentence structure, with
>finite verbs occasionally appearing in weird places as an exception.
>However, finite verbs are the invariant in German sentences (no
>sentence is grammatically correct without a finite verb), while
>infinite parts of the verbal phrase are the exception.
Here's how I understand it (Rosenfelder's wording aside). Tense and
aspect, which are obligatory in finite clauses, are realized on a
second position element. This may be an auxiliary, as required by
certain tense/aspect combinations. In such a case, the auxiliary is
in second position, and the lexical verb is final. Other tense and
aspect combinations may be realized on a lexical verb. In such a
case, the lexical verb appears in second position.

>While the claim "All elephants are green except for the four-legged
>ones" might be logically true (albeit weird), the claim that
>"Elephants are *basically* green, except for the four-legged ones"
>is definitely false, since it designates a wrong "normal state"
>while treating the true normal state like an exception.
The normal state in German main clauses is that tense and aspect are
marked on the element appearing in second position. This
generalization is IMO superior to a SVO analysis of German, since it
accounts not only for SVO order when it occurs, but also for SAuxOV
order, which also occurs. It doesn't privilege one order over the
other since it describes both.

Dirk
--
Dirk Elzinga                  Dirk_Elzinga@byu.edu

Man deth swa he byth thonne he mot swa he wile.
'A man does as he is when he can do what he wants.'

- Old English Proverb