[sw-l] SignWriter - feature "gloss" input method

[Steve Slevinski]

Hi Barbara,

I used to be a proponent of glossing, but no longer.

I didn't understand international concerns, but I'm learning.
In theory, I support Unicode.  I still have testing to do.
Each sign will have mutliple keywords and multi-lingual descriptions.

Glossing has been removed from the SignPuddle 2.0 beta. Instead each
sign is given a sequencial number starting with 1. The beta is usable,
but the terms are confused.  Val and I are discussing all of these terms
and ideas.  We don't agree on all of the terms, but we agree on the
direction and we are having fun along the way.

Here is a section about the ideas and terms we are discussing.  This is
a rough draft to explain my thinking.

Drawing Sequence
Each sign has an order in which the symbols were drawn in SignMaker.  In
theory, this could be the same order used when drawing by hand.  Many
start with the head when drawing for proper centering.

Sort Sequence
Create a Sign Spelling Sequence by clicking on the symbols displayed in
the draw sequence.  The sort sequence uses the SSS id numbers of the
symbols in the IMWA.  Read up on Sutton's SignSpelling Guildlines for an
offical SignWriting document.  An example SSS id number would be
01-01-001-01-01-01, not to be confused with a Base Symbol which could
have an SSS id number like 01-01-001-01.  The SSS id numbers lend
themselves very well to sorting.

Detailed Spelling Sequence
After a dictionary grows to more than 200 signs, a simple sort sequence
is not enough.  We now need to use a Detailed Spelling Sequence.
Detailed sorting uses additional symbols not found in the draw
sequence.  New sections of the IMWA and additional SignWriting rules
must be learned for correct detailed spelling sequences.

Statistical Symbol Subsets (aka alphabets)
Using the whole IMWA in SignMaker is possible, but a subset may be
beneficial for beginners.  SignMaker will work will the whole IMWA or
unique subsets based on usage.  Additionally, PDFs could be created for
each signed languages.  These alphabets are not dictated or given to the
deaf.  These alphabets are discovered by reflecting in the puddle.

SignPuddle 2.0 is still in beta.  It could be used today and I may
create an IMWA puddle to see how it scales.  But it does not use gloss.

-Steve






Barbara Pennacchi wrote:

>
> On 31.05.05 11:56, Tomáš Klapka wrote:
>
>> My idea is to add one more method of input, "gloss" input method.
>
>
> I'm not talking about signpuddle specifically, I'm talking about the
> concept of "glossing" signs in general.
>
> I'm sorry, but after so many years of working side by side with sign
> language researchers and teachers, my opinion about the (ab)use of
> glosses  in almost any sign-related field is very very very very low.
>
> I *do* really wish we'd all start weaning ourselves from these. I do
> really want do wean myself, at least (grin)
>
> Glosses do not solve the problem of "rebuilding" in one's own mind a
> specific sign made by someone else, in absence of visually-supporting
> material. Glosses tend to add more ambiguity. Glosses do not help
> people  visualize the 4dimensionality of sign language.
>
> And how about the presence of multiple synonyms? Would glosses such
> as  "DROP" "LET-IT-FALL" "ABANDON" point to the same or to different
> signs?
>
> And how about using glosses from a strongly "gendered" language such
> as  spanish, french or italian (which don't have the neutral gender)
> for sign  language that tend instead to be genderless?
>
> Furthermore, in some cases, the use of glosses is suspected of
> influencing  non-deaf people into translating sentences from their own
> mono-dimensional  language into our 4dimensional language, word by
> word, before actually  signing the sentence.... (I'm being sarcastic
> here, folks)
>
> But this is only my opinion. And I'm having a real bad day here at
> work so  forgive me for my bluntness. Nothing personal here, ok?
>
> (BTW: it bugs me that Sign Puddle forces me to give a gloss to signs
> in  order to save 'em, but I know it is because of technical reasons
> --  "Unicode, where the heck art thou?")
>