tlhIngan-Hol Archive: Tue Jun 09 09:23:59 1998

Back to archive top level

To this year's listing

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

RE: Infinitives and Subjects (was RE: Online Lexicon of Linguistic Terminology)


charghwI'vo':

> You just reminded me to try looking up "infinitive" in my
> Concise Oxford Dictionary. I know it is not a linguistic
> authority source, but worth checking out:
>
> "(Verb-form) that expresses verbal notion without predicating it
> of any subject (e.g. ask, to ask)."

Well, there *are* cases where infinitival forms (in English) can take
subjects.  In these cases, the subject noun must receive case marking (yes,
case; see below) from another element, such as a preposition or an
"Exceptional Case Marking" (ECM) verb.

	For John to leave now would be foolish.
	I believe John to be guilty.

'John' is definitely the subject of the infinitive, but receiving case from
another element (P 'for' in the first, ECM verb 'believe' in the second).
What's important is the lack of tense (at least, from the perspective of the
theory with which I am familiar; not so say other theories don't have
equally valid interpretations).  In Chomskian syntax (here it comes), all
overt NPs must have case; one possible case assigner is Tense (assigning
case to the subject; the verb itself assigns case to the object).  If a verb
is non-finite (no tense), then either there is no (overt) subject, or the
subject NP has to get case elsewhere.

That's ultimately why I was looking for a more accurate term, such as
"impersonal."  It describes a verb form that does not assign an argument
role (say, AGENT) to a subject.  The verb may have other forms in which it
does assign those roles.  Like with the passive form, it assigns the object
role, and no subject role; the object raises to the position of the subject.
If there's a subject, it gets case from another element, usually a
preposition.

Ack, this is exactly the path I was trying to avoid going down!

> Also note that Holtej's repeated reference to infinitives not
> having reference to tense seems odd, since I would think that
> "to have given" was an infinitive form. I'm sure there is
> simply something that I don't understand.

Alternatively, it may just be an artifact of Chomskian syntax; more general
desecriptions of grammar probably don't make the distinction that I am
describing.  (The SIL glossary I pointed out, for example, does not.)  What
can I say?  I've spent the past 7 years studying one theory.  Maybe when I
finish this degree, I can broaden my horizons.  :)

As for "to have given," this actually argues for the interpretation of
"tense" as a subject case-marker.  This verb form is inflected for aspect,
not tense.  No tense, no case-marking, no subject.  * "John to have given
the ball is surprising."  But with a case-assigning preposition it's fine:
"For John to have given the ball is surprising."

But in defense of the term "impersonal," it's not a theory-specific term.
Non-chomskians will understand what it means (at least, those with
familiarity with the syntax of verbal argument structure, linguistics is a
very broad field).

Jeez, no wonder there's confusion.  Even Oxford's got it wrong.  No, an
infinitival form is simply not conjugated (I say for tense; more generally
(or, less theory-specifically), for person, number, etc.).  Infinitives
*can* have subjects, in special cases.

Sorry, guys, don't mean to keep this dreadful linguistic thread going!

> charghwI'

--Holtej

Back to archive top level