new sign transcription system

[Dan I. SLOBIN]

Dear Angus,
        We're delighted that you've already examined our system.  The
reason we want to have a strictly morphological transcription (on the
utterance line) is that this is the level of language acquisition that our
research focuses on--that is, the assignment of meaning to linguistic
elements.  In this way, we are consistent with the long tradition of child
language studies of spoken languages.  We found it awkward, for example,
to introduce a notation such as "cl'V" for the two-legs classifier, while
knowing that it is a morphological element.  By analogy, one does not
represent the PLURAL morpheme by "-s"  in morphology-level transcriptions
of English, but rather as PL.  To be sure, phonological information is
often relevant, and can be provided on one of the dependent tiers (%pho).
If we were studying the acquisition of phonology, of course, we'd use a
totally different transcription system (and some candidates already
exist).
        The reason for putting semantic elements in caps is to be able to
search for them independently of other symbols, thus carrying out word and
morpheme counts automatically, using CHAT programs.  The glosses in
parentheses are not part of our analysis, but are only there to aid
legibility.  That is, for example, RIDE_MOUNTED is a cue to the reader
about how to untangle the following list of symbols.  But we give no
linguistic significance to these glosses, and the transcriber is free to
use whatever gloss seems to be a useful pointer to the meaning of the
following sign notation.  What first pushed us to abandon spoken-language
glosses was our work with ASL and SLN:  we found that we were worrying
about the comparable meanings of English and Dutch words in the glosses,
rather than the actual signs.  It was a great relief to leave that
tradition behind, and discover the rich polymorphemic structure of the
lexical items themselves.
        I like your suggestion that we explore ASCII notations for sign
phonology.  This would be very useful on our %pho tier, where we are
definitely concerned with childish and idiosyncratic formations, as well
as those of hearing parents who are learning to sign with their children.
What ASCII system would you recommend for phonology?

Many thanks for your comments.
Best regards,
-Dan

On Wed, 23 Jun 1999, Angus Grieve-Smith wrote:

> On Sun, 20 Jun 1999, Dan I. SLOBIN wrote:
>
> > This is still work in progress, and we look forward to feedback in
> > further development and application of the system.
>
>         This is a very useful project!  I hope I can provide some useful
> feedback.
>
>
> >         - linear representation on a continuous typed line, using
> >           only ASCII characters
>
>         This is very important for all transcription/writing systems.  It
> is possible to use non-ASCII systems (such as SignFont or SignWriting),
> but the convenience of ASCII for email and various searching and
> concordance programs is significant.
>
>
> >         - consistently morphological representation (avoiding both
> >           phonological representation and glosses in English or
> >           other spoken languages)
>
>         Okay, I don't understand this.  The expamples in the .pdf file
> were full of glosses.  Do you mean not using glosses to indicate bound
> morphemes?  And why do you want to avoid phonological representation?
>
>
> >         - full representation of elements of polymorphemic verbs
> >         - representation of manual and non-manual elements
> >         - representation of gaze direction, role shift, visual attention
> >         - representation of gestures and other communicative acts
> >         - notation of characteristics of adult-child interaction
> >           (child-directed signing, errors, overlap, self-correction)
>
>         These all seem to work.  After a few minutes of study I was able
> to figure out most of what was going on in those transcripts, still
> needing to go back and look at the explanations several times, but I think
> if I worked with this system on a regular basis I could pick it up.  The
> codes used are not particularly transparent, but that's fine.
>
>         As a longtime supporter of phonological notation, I have to add a
> couple of words about the use of glosses.  I'm sure you've all heard the
> various arguments both ways, so I won't go into them.  I just want to say
> that your system is in fact compatible with ASCII-based phonological
> notations, so you might want to make the "English gloss" requirement more
> flexible.  The second point is that the use of all caps in glossing is
> nothing more than a convention, and in fact it's the same arbitrary
> convention that means "shouting" in email.  I'd say it's easier to dispose
> of that convention than to tell everyone they have to use the "+k" flag in
> their searches.
>
>         Again, it's great to have something that will allow us to put
> signed-language data into CHILDES!  Looking forward to seeing everyone's
> corpora on line soon.
>
> --
>                                         -Angus B. Grieve-Smith
>                                         Linguistics Department
>                                         University of New Mexico
>                                         grvsmth at unm.edu
>