Agreement or indicating verbs?

Sat May 29 05:42:19 UTC 1999

Thanks to everyone so far who has contributed to the discussion regarding
agreement or indicating verbs.

I would like to respond to Gary Morgan's comments posted to the list via
Bencie Woll.

I agree that the evidence from first language acquisition, signed language
in Nicaragua (ISN vs LSN), and aphasia which you mention is extremely
interesting and seems to very strongly support the mainstream analysis of
agreement verbs and the syntactic use of space. I have a few concerns about
some of this evidence, however, and I'm familiar with other types of
research which seem to complicate the picture a little.

(1) You referred to studies by Newport (1990) and Mayberry (1993). I find
the evidence from studies by Mayberry (1993) much more persuasive than the
evidence put forward by Newport (1990). The study by Mayberry (1993) you
mention differentiates between late first language learners of ASL and
second language learners, for example, whereas the paper by Newport (1990)
does not. It also provides information about standard deviations for its
test scores, so that the difference between native signers and other groups
is clear both in terms of their mean scores and in the range of scores
across the members of each group. Newport's work came in for considerable
criticism in this regard from Bialystok & Hakuta (1994).

(2) In regard to your points about the use of agreement verbs in signing
children, it is worth noting recent work by Casey (1998). Although her work
confirms the findings of previous researchers that deaf children produce
more uninflected than inflected forms of agreement verbs before age 3;0, her
data suggest that directionality in both gestures and signs is used
throughout acquisition. She also points out that research has shown that
deaf children (aged 1:4-4;1) with no exposure to a signed language produce
action gestures containing directional movement and spatial displacement to
indicate arguments (Goldin-Meadow & Mylander, 1990). Deaf children only
exposed to Signing Exact English (SEE 2) produce directional forms of SEE 2
verbs (Supalla, 1991). The use of directionality in verbs, she concludes,
has gestural origins, but is grammaticalised in ASL.

(3) You mention work on deaf aphasics and the types of dissociation reported
by Poizner & Kegl (1992) and their colleagues. In work by McNeill & Pedelty
(1995), they make mention of research which found that hearing patients with
aphasia (of the non-fluent type) display a disruption of the abstract use of
gestures, such as the use of abstract pointing, although they are still able
to make use of mime. This leads them to suggest that the underlying
dissociation with both signing and hearing aphasic patients is along the
concrete-abstract dimension, rather than linguistic versus non-linguistic.

(4) Another study by Casey & Kluender (1998) showed that, when deprived of
speech, hearing subjects with no exposure to a signed language produce
action gestures which indicate arguments of the verb, similar to agreement
verbs in signed languages. Similar findings have been reported by Dufour
(1993).

(5) Senghas (1994) found that ISN signers produced more locative and person
agreement than signers of LSN (but they also describe more events in more
detail than signers of LSN), but they both produced similar amounts of
handling classifiers. What does this suggest to you about the linguistic
status of such forms?

I don't want to argue that the use of space in agreement verbs and the
directional use of gestures by non-signers are fundamentally the same thing,
but I find characterising the difference somewhat difficult. The difference
between deictic gestures and pronominal pointing in signed languages, for
example, seems to be one of degree, rather than kind. There is perhaps why
there is disagreement about the number of pronouns in signed languages and
if specific forms other than a first person pronoun exist (Meier, 1990;
Engberg-Pedersen, 1993). In my own work on polymorphemic verbs of motion and
location in Auslan, producing an inventory of articulator (a.k.a
"classifier") morphemes is relatively straightforward, but producing an
exhaustive list of morphemes expressed by location and movement is
altogether a more difficult task.

Like Sherman Wilcox and others, I find I'm very much in the "lumper" rather
than "splitter" category. Sherman commented in a response to my question
(not posted to the SLLING-L) that it is not easy to divide up the world into
"linguistic stuff" and "gesture stuff". There may be a continuum, with
prototypical gestures at one end, and prototypical signs at the other, but a
very messy middle. There seems no principled way, he suggests, to
distinguish gesture from linguistic elements (especially, I would add, when
we know so little about the acquisition, development and loss of signed
languages, and even less about these aspects of gesture). I find Liddell's
work interesting and challenging, but I agree with those that suggest it is
predicated upon a definition of "linguistic" versus "non-linguistic" that
not everyone is comfortable with. In relation to the evidence from aphasia,
for example, I'm intrigued by the suggestion from the work of McNeill &
Pedelty (1995) that it is a general feature like compositionality and
abstractness (present in some uses of gesture, and in all languages) that it
is undercut by left hemisphere damage, rather than some specifically
linguistic skill.

Adam Schembri

PS How little we really understand about gesture is highlighted by the
recent surprising finding by Iverson and Goldin-Meadow (Nature 396, 228 19
November, 1998). Their data suggests that congenitally blind children
produce just as much co-verbal gesture as sighted children, even when
speaking to other blind children! Yet children who are blind from birth
presumably never see such gestures and so have no model for gesturing. Some
uses of gestures thus appear to require neither a model nor an observant
partner, they conclude.

>From: "B.Woll" <b.woll at CITY.AC.UK>
>Reply-To: "For the discussion of linguistics and signed languages."
>      <SLLING-L at ADMIN.HUMBERC.ON.CA>
>To: SLLING-L at ADMIN.HUMBERC.ON.CA
>Subject: Re: agreement of indicating verbs?
>Date: Thu, 27 May 1999 13:33:51 +0100
>
>>From Gary Morgan via Bencie Woll:
>
>Thanks Adam Schembri for the original prompt for discussion,
>During Scott Liddel's TISLR talk a lot of things came to me
>regarding the differences between an agreement verb in a sign
>language and a gesture produced by a non-signer.  Although from a
>formational perspective they may look the same, as linguists or
>psychologists we have to go deeper than this.  I will try to outline
>some of these in no order of importance. They overlap with those
>points Rachel Mayberry made at the end of Scott Liddel's talk at
>TISLR.
>
>There is work from Elissa  Newport (1) which describes the
>differences between native signers (exposure to sign as a young
>child) and late signers (first exposure as an older child or young
>adult) in the use of verb morphology.  Learning sign late doesn't
>stop you from having a high level of competency but having early
>exposure gives you an advantage in the productivity of the
>morphological system.  Non-linguistic gesture is equally
>accessible to both populations of signers, yet it is the native
>signers who have most linguistic competence with signs because
>they have applied a child-orientated linguistic analysis to the input.
>This work was supported by Rachel Mayberry's claim (2) that any
>language input during infancy signed or spoken will make a
>difference in the final competence of the signer.  Another
>compelling argument comes from Judy Kegl's description of the
>difference between LSN and ISN (Nicaraguan Sign Languages)
>signers in their use of all aspects of verb and noun morphology
>(3).  Again if verb morphology is based on non-linguistic information
>why would there be a difference in these populations?
>
>Strong evidence for the linguistic nature of agreement verbs comes
>from the over-generalisations of verb argument structure young
>children make in their spontaneous signing.  Bellugi et al (4)
>reported on non-adult like use of verb inflections in ASL  10 years
>ago.
>
>A more recent example comes from British Sign Language (5).
>One group of BSL agreement verbs are produced in two parts.  The
>first part of the verb-pair moves away from the signer towards a
>location indexed previously as a third person, the second part of
>the pair moves back onto the signer's own body.  An example
>utterance would be BOY WASH-OUT GIRL WASH-FACE 'the
>boy washes the girl's face'.  Adult native signers always indicate in
>the first verb what the action is and in the second verb, what the
>affected body part is, and without fail they appear in this order.
>You 'could' say this is entirely gesture-based as it is an enactment
>of the action using surrogates and indicating gestures.  However
>looking at the child BSL data you get a different picture.  Children
>from 3 to 4 years old reverse the order of the verb pairs painting
>their own faces first then painting a third person in a location in
>from or to the side of them. Example *BOY WASH-FACE WASH-
>OUT 'the boy washed, washed him'. They do this at an age where
>they show good comprehension of this structure in adult-produced
>utterances.
>
>If they were approaching verb morphology based on non-linguistic
>information, a reversal of this type is difficult to explain as it would
>represent  a gesture schema they will never have seen in their
>environment (ie. non-iconic).  Other evidence for the linguistic
>underpinning to agreement verbs comes from signers who have
>suffered brain injury,  especially the types of dissociations reported
>by Poisner, Kegl and colleagues (6).
>
>The above studies suggest therefore that although a sign may have
>formational similarities to a non-linguistic gesture it hasn't
>necessarily the same underlying representation in the mind of the
>user.
>
>(1) Newport, E. (1990). Maturational constraints on language
>learning. Cognitive Science 14, 11-28
>(2) Mayberry, R.I. (1993). First language acquisition after child-
>hood differs from second-language acquisition: The case of
>American Sign Language. Journal of Speech and Hearing Research
>36, 1258-1270
>(3) Kegl, J. (1994). The Nicaraguan Sign Language project: an
>overview. Signpost, 24-31
>(4) Bellugi, U., van Hoek, L., Lillo-Martin, D. & O'Grady, L. (1989).
>The acquisition of syntax and space in young deaf signers. In K.
>Mogford-Bevan & D. Bishop (Eds.), Language development in
>exceptional circumstances. Cambridge: L.E.A.
>(5) Poizner, H. & Kegl, J. (1992). The neural basis of language and
>motor behaviour: Evidence from American Sign Language.
>Aphasiology 6, 219-256
>(6) Morgan, G., Herman, R. & Woll, B. (1999). Event packaging in
>British Sign Language: The development of serial verb
>constructions. Paper to be presented at the IASCL congress, San
>Sebastian, Spain, July 1999.
>
>
>
>
>
>
>-------
>Gary Morgan
>Dept of Linguistics, UCL, London
>tel: 0171 4193162 (voice/text)
>fax: 0171 3834108
>
>
>
>Bencie Woll
>b.woll at city.ac.uk
>Chair of Sign Language and Deaf Studies
>City University, Northampton Square
>London EC1V 0HB, UK
>Tel: +44 (0)171 477 8354 (voice) +44 (0)171 477 8314 (text)
>Fax: +44 (0)171 477 8577 or 8354

______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com