Pet peeve

[Herb Stahlke]

Larry,

The to-infinitive developed from a semantic contrast in Old
English between two different sets of verbs, a contrast that still
persists in ModE in the difference between "will" as a modal and
"will" as a desiderative verb, a doublet in which the former
descends directly from OE "willan" and the latter from ME
"willen", which itself is from OE "willan", -an marking the
infinitive in OE.  The verbs that were to become ModE modals did
not use "to:".  Verbs of causation, intention, motion, perception,
etc. were variable, sometimes taking "to:" and sometimes not.
Verbs of purpose regularly took infinitives with "to:", and "to:"
also was used if the infinitive was a complement to a noun or
adjective or functioned as a nominal itself.  In late OE, "to:"
generalized to more infinitives.  The suffix continued well into
ME, although the /n/ was increasingly dropped.  It's still present
in Chaucer, as in the line "Thanne longen folk to goon on
pilgrimages."  While to-infinitives are more widely distributed in
ModE, there is still a complementarity between modals, perception
verbs, and a few semi-modals taking the infinitive without "to"
and most other uses requiring the "to".  I don't see any reason
why the analysis would have changed significantly from OE times.
Details of verb classes have shifted, but surprisingly little.  As
for the use of the term "base form", that's a consequence of the
decision, or lack of decision, to define the infinitive as marked
only by "to".  It's morphologically useful in that it draws
together a variety of uses of the uninflected verb, but it's not
of much use in a grammatical description.  I agree, BTW, that
"marked/unmarked infinitive" is not standard usage, but then
"split infinitive" is, so that measure doesn't buy us much.
Obviously, I'm pretty much restating Jespersen's case in MEG, Vol.
5.

Herb

>>> laurence.horn at YALE.EDU 10/31/00 02:56AM >>>
At 3:39 PM -0500 10/31/00, Herb Stahlke wrote:
>I haven't run into, or at least noticed, this aberration.
>However, as others will undoubtedly note, split infinitives are
>perfectly acceptable, another 18th c. prescriptivism surviving
too
>long.  In "would change", by the way, "change" is an
infinitive.
>Modals take unmarked infinitives.  The "to" is not what makes a
>form an infinitive.  It's just one way of marking that status.
>
>Herb Stahlke
>
Well, that's really a question of definition.  On some
definitions,
I'm pretty sure the majority of them in both theoretical and
descriptive analyses, only the "to ____" forms are in fact
infinitives, the others--while non-finite--aren't.  (-ing forms,
as
in present participles, are also non-finite but obviously not
infinitives, and likewise for imperatives and subjunctives.)  As
for
the form that follows the modal, I've heard/seen that referred to
as
the base form.  I don't believe "marked/unmarked infinitive" is
standard usage, although I would have no trouble if a particular
treatise explicitly defined them that way.

larry