Continuous vs. Repetitive inflection in ASL

[Don E. Newkirk]

Sorry to jump in after a week's discussion, but I was out of town (on
another planet--um, er,
rural Ohio).  Since I have a long-standing interest in ASL morphology, I
will stick my foot
firmly in {the door|my mouth}, and have half a go at it.

PREDICATE ROOTS AND STEMS

First, surface form is only an indirect identifier of the predicate
root--temporal aspect
inflections in ASL are applied to stems which encode, depending on the root
and its
semantics, one of the following dichotomies:

        Stative vs. Punctual
        Imperfective vs. Perfective.

These categorical distinctions are almost the same, but not quite: the
boundaries are sometimes slightly different, and it is possible to encode a
predicate
that is Imperfective, though based on an inherently Punctual stem, for
example.

Note that root predicates can be Stative or Punctual, while whole verbs
with their
inflections can be Stative, Punctual, Imperfective, and/or Perfective.

How does what seems like a Stative and Perfective predicate such as KILL,
then,
do under inflection?  Let's leave my (partial) answer till after my next
digression.

THE NATURE OF TEMPORAL ASPECT INFLECTIONS IN ASL

I have stated in several places that the inflectional system in ASL is
quite a bit
different, at one or two levels deeper than the surface, than the
descriptions to be
found in Klima and Bellugi 1979 (_The Signs of Language_).  I take quite
seriously
Supalla and Newport's (and Newport and Supalla's) descriptions of verbs of
motion
and location, and the implications of their findings for the rest of ASL
predicative
morphology.  At the heart of their analysis (perhaps I lend it more
importance than
they) are families of similar verbs, differing in the presence or absence
of
formationally and semantically significant movement fragments:

        core stasis vs. directional path movement
        hold end manner vs. continuous end manner vs. restrained end manner

Significantly, a semantic root encoded by handshape, contacting region, and
location, can be amplified into an inflecting stem by the affixation of
some movement
pieces, or by a kind of ablaut that alternates specific semantic types,
moderated by
specific movement changes.

LOOK-AT has a path movement to a hold; BE-LOOKING only has stasis (I
discuss
obligatory epenthetic initial unmarked path movement to a static core
elsewhere).
It seems fair to posit the construction of LOOK-AT from BE-LOOKING, where
the
added path movement makes the verb Punctual.  Note that both LOOK-AT and
BE-LOOKING are both 2-argument verbs, so the path does not add
transitivity, as
some have argued.

Add simple repetition to these 2 verb stems, and what happens?  Egad, in
the first
case one gets the Habitual aspect inflection described in K & B, while in
the other
case one gets the Durational aspect inflection--made with non-tense,
non-fast
circular reduplication cycles!  Semantically, one can say that over time,
one
extends the Punctual act by making it habitually, while one extends the
Stative
act by, well, extending it--in both cases through reduplication (which I
also argue is
an all-purpose Stativizer in ASL).  You can't "repeat" true stasis, but
remember the
epenthetic initial path movement that is obligatory for ASL verbs with
static cores?
That makes a sequence with the static core in BE-LOOKING, which, when
repeated,
leads naturally to small circles.

Take the same two stems, repeat them, pull the resulting surface forms
apart,
and then make sure that the outward path movement is tense, the turn-around
segment is quite tense (in LOOK-AT, the hold end manner becomes a tense
hold; in BE-LOOKING, the continuous end at the turn-around to the
transition back
to the beginning locus becomes restrained: super tense).  The result is
what in
K & B are called Iterative and Continuative.  The process is not identical,
exactly,
and a true Continuative can be built on the LOOK-AT stem;  note that the
 Continuative
inflection requires a Stative stem, so--you guessed it--you must
reduplicate LOOK-AT
first, then add the dynamics of the Continuative inflection, with a path
profile very
similar to that of the Continuative form of INTERRUPT:

In INTERRUPT[Continuative], straight path to a hold on the base hand
becomes
tense straight path + restrained sliding of the active hand in contact with
the base
hand + non-tense return to the beginning.

In LOOK-AT[Continuative], straight path to a hold in space becomes tense
straight path,
full-sized, followed immediately by a short repetition of the Citation-Form
movement,
ending at the indexic point of the Direct Object;  there is now a
restrained turn-
around (down, then back toward the beginning locus), followed by a normal
transition back to the start.

There are, then, independent inflections with the following semantics:

        Punctual Stem           Static Stem             Meaning
        -----------------------------------------------------------------------
---------------------------------------------------------------------------
        Habitual                        Durational              Simple extension of episode
        Iterative                       Continuative            Long (often negatively viewed) extension of
episode

K & B mention more inflections that can be added to this matrix.  For
instance:
        Incessant               Obsessive               Names are descriptive; 1st describes repeated acts,
                                                        2nd describes State of having to do X.

The Incessant and Obsessive are formed by adding tension to already
shortened
stem movements.  Non-manual gestures usually accompany these two,
indicating negative
affect.

While I believe that the Habitual and Durational of K & B are reflexes of a
single
simple reduplicative inflection, built on Punctual and Stative stems, I am
not entirely
sure about the Iterative and Continuative, though their places in this
paradigm
are strongly suggestive that they are also alternating stem-based reflexes
of a single
process.

KILL -- FINALLY

Since temporal aspect inflections in ASL _never_ entail lateral spatial
displacement;  the reason native
signers are apt to add such displacement to KILL [Durational] is that the
latter (without sweep) is semantically
suspect at best.  KILL is punctual, in ASL as in English, despite endless
in-class discussions of whether a
victim poisoned on Monday, who died on Friday, was actually _killed_ on
Monday by his assailant,
or on Friday by the inexorable workings of the toxins.  Note that stative
(imperfective) phrases in English
containing this verb, such as "I thought I was killing him" do not
translate in ASL with the Durational
aspect inflection (there is an inchoative--"Inceptive" in K & B--that
translates this better, though
independently one might render the translation back into English as "I
thought I would kill him".

Simple (usually) ipsilateral spatial displacement marks plural direct
object (in a 2-argument verb) or plural indirect
object (in a 3-argument verb).  Number marking in ASL verbs is quite
independent of temporal aspect, and
combines with temporal aspect inflections to produce a rich variety of
surface forms.  A topic for another thread.

Because Habitual and Durational are *not* independent inflections, and
Iterative and Continuative may not
be independent inflections, the application of the list of inflections
found in K & B (and elsewhere)
arbitrarily to any verb in ASL stands a fair chance of producing an
anomalous or marginal form.

MORPHOLOGY AND SEMANTICS IN ASL

In summary, temporal aspect inflections in ASL are coded underlyingly as
invariant sequences of segmental
and suprasegmental building blocks, but in the process of deriving surface
forms, as well as surface semantics,
the outcome may vary from verb to verb, and from stem to stem.  The
bundling of a specific form with a specific
meaning, in ASL as in any language, is liable to formational as well as
semantic irregularities, usually reconcilable
in diachronic and synchronic terms.

On Friday, May 07, 1999 12:05 PM, Tane Akamatsu [SMTP:tanea at IBM.NET] wrote:
> Sandra,
>
> I'm wondering, from your gut (since you are a native signer), who is
being
> killed in KILL-durative, and who is being killed in KILL-continuative?  I
know
> that *grammatically* the sweep vs. the same-place difference exists.  To
these
> non-native eyes, the sweep implies that many different people/insects
were
> killed over some period of time (i.e., serial killer, as someone
mentioned
> earlier), and with the same-place, it seems like the same person/insect
is
> being killed over a period of time, which makes no sense, as someone else
> pointed out.  Would a long-slow-tortuous death be implied in the latter?
>
> Tane Akamatsu
>
> Sandra K. Wood wrote:
>
> > I've been following your discussion on 'KILL' (durative vs.
continuative).
> > I agree with Don's assessment pretty much.  However, I'd like to add my
> > observation as a native signer and ASL linguist which may (or may not)
> > help distinguish between these two semantically.  When I sign KILL
> > (durative), I move my hands from left to right over the neutral space
in
> > front of me, while signing the durative aspect of KILL at the same
time.
> > When I sign KILL (continuative), I simply keep the signing in the same
> > space while signing the continuative aspect.  I think this adds another
> > dimension to the layer of semantic information represented by the type
of
> > movement that durative and continuative make.
> >
> > What do you think, Don and whoever else is reading this? :)
> >
> > Sandra K. Wood
> > Purdue University
> > ASL Linguistics Lab
> > swood at omni.cc.purdue.edu
>