Re: SignWriting

> SignWriting is also difficult to write.
>
> Not necessarily more than those that learn writing Chinese.
Learning how to write Chinese is difficult. It "only" takes like 6.5
years of schooling, and when students go abroad for college, they
quickly forget how to write many characters. In fact, /within/ China,
I'd say nearly everyone including civil servants dealing with
registration matters (!) whom I've ever asked to and seen write down a
sentence or a bunch of expressions for me stumbled or scribbled outlines
for certain characters or got components wrong in a way that didn't look
like they were just abbreviating. I'm not exaggerating.

> It appears difficult to those that don't know sign languages. But for
> the others, yes it requires training, just like those that learn to
> read and write any oral language of the world (we call this ability
> "li[t]eracy").
Do you think the shapes are amenable to fast writing by pen?

> Don't forget the purpose of this script : not just communicate in a
> written form but also help learning the sign languages themselves.
For that it's probably useful.

> And it's certanily more economic than storing and sending videos, that
> are also highly dependant on encoding and rendering technologies.
> But if technologies are available, it will be easy to render them
> correctly, or to convert it into an animated person showing the actual
> signs (not more complicated than decoding and rendering a video.
You might be underestimating the difficulty of CG (computer graphics).
I'd say: perhaps not in a way that looks realistic, but probably in a
way that the important features get conveyed – some day.

> The same technogies will also allow converting them (based on
> dictionaries) to sequences of words representing the oral language
> (but most probably the encoded sign-writing would be sent along with a
> transcription of the associated oral language)
Grammars of sign languages are a somewhat different; are you sure that
dealing with relative positioning in 3D is that easy? What's easily
expressible in a sign languages can be hard to express in an oral or
written language. I won't speculate further, as this is really not my area.

What I was thinking of was a way of serializing the heavily multimodal
gesture sequence. Output of oral languages seem suitable for linear
representations. The exceptions are called "suprasegmentals", but as I
understand /most/ of oral language is encoded in linearly arrangeable
segments (phones). An orthography that is fast to write (this requires
brevity of the signs, which in turn requires a fair level of
abstraction) would be good. I'll be curious to hear of systems that
aren't clumsy.

Stephan
Received on Sun Apr 21 2013 - 18:09:02 CDT