Antw: Re: [SLLING-L] What's the deal with SignWriting?

[Franz Dotter]

Dear Andrew,

I was not clear enough: We have to differentiate between scientific notation for different purposes. Just to list the basic ones:
1. a notation (=transcription) in order to show the formative/phonetic-phonologic features of (the signs/productions of) a language: Here we have technical names for the distinctive features of sounds (like "nasal" or "plosive") and the IPA-system of alphabetic signs for sounds more or less commonly accepted by all linguists. For sign languages we have only HamNoSys which is comparable in exactness (but this is rather complex to use because there is no internal ordering of the features of a sign). You are right that the Stokoe notation does not allow to represent all features like HamNoSys does. Therefore we may take it as a scientifically older notation which may be used for some purposes where its elements are sufficient.

2. a notation of language production for the morphological level (i.e. ignoring the exact transcription of sounds and only representing ALL morphemes: For spoken languages there is a typological norm for such representation which contains 3 lines with different functions:
first line: gives an "impression" of the utterance (= language analysed) either in a broad phonetic or phonological representation (often avoiding IPA characters which are not part of a normal alphabetical writing), or only in writing
second line: gives a morpheme-for-morpheme representation of the content in the language of analysis (lexical morphemes in normal letters, grammatical ones in capitals)
third line: gives a free translation into the language of analysis.

Elena Pizzuto and colleagues where right in pointing out that many respective sign language notations lack an equivalent of the first line (most ooften there is only a glossing in the language of analysis and you can get no idea of the analysed signing itself). SignWriting could help filling this gap if it was accepted.

So far scientific notations which explicitly represent linguistic units of different levels in an individualised manner.

Lets change to writing systems now: No writing system is primarily linguistically oriented but it more or less represents words. So is SignWriting: It is neither intended to show every single formative feature nor to show every single morpheme of a sign. It represents signs and you have the possibility to add or let away some additional writing elements (like apostrophs or accents, diacritical marks in some writing system for spoken language or complex vs. more simple representation of Chinese ideographic signs).
Therefore SignWriting is fine for the "first line" but no one can see the linguistic features/elements of any sign language DIRECTLY (= without further analysis of the sign represented) from it (which is the most important property of a scientific system). As it follows a depicting strategy (which - besides - you can also find in HamNoSys) it may serve more (but not completely) like a phonologic transcription and does not darken production features as some writing systems for spoken languages do. And therefore it could serve as an important tool (because you can get a rather exact illustration of the signing), but - from its definition and strategies - it is not an analytical tool itself.

best regards

Franz

>>> "Andrew Pidkameny" <pidkameny at gmail.com> 13.01.09 19.19 Uhr >>>
Franz,

I want to know more about why you claim that Stokoe Notation and
HamNoSys are more appropriate scientific notation systems than
SignWriting. Is it because SignWriting can be produced with varying
levels of precision/abstraction? It would seem that a higher level of
abstraction, as is found in Stokoe Notation and HamNoSys, would make
it more difficult to accurately convey phonetic information. And from
what I've heard, Stoke Notation is at least as imprecise as
SignWriting, as it lacks descriptors for nonmanuals.

What say you?

~Drew

> Then you continue:
> "It can be used in scientific annotation as
> much as any other transcription method and it can be integrated in
> scientific annotations (for instance in combination with time markers or
> annotations)."
> Here I disagree: Scientific transcription always has to represent all the
> features of any behavioral activity demanded/important for/on different, but
> definable levels of abstraction. That means, as long as you want to describe
> production phenomena of language, you have to use a scientific transcription
> (where - for sign languages - you have e.g. the choice between the Stokoe or
> the HamNoSys system, different approaches for different levels of
> abstraction, but scientifically defined). Economic everyday witing does not
> obey this claim. Naturally, as we sometimes use the written language system
> (the writing code) in order to illustrate some facts of language where
> production phenomena are not necessary (e.g. in syntax), you can adopt the
> SignWriting code also for that (often/sometimes enriched with representtions
> of grammatical categories like - for spoken language e.g. "do_3PSG" for
> "does"). But in this method you do not follow the rules of written language
> (be they normative
>  or not) but you only use a written form for representing speech/signing. My
> consequence: As a linguist I CANNOT use e.g. German written language as a
> scientific tool for describing every day spoken German. I can use elements
> from written German or English to gloss the semantics of e.g. a spoken bantu
> sentence, adding some signs for grammatical elements, following some
> conventional linguistic rules.
> In other words: You have always to differentiate between the area of objects
> which you describe (that could be spoken/signed or written communication and
> the area of scientific description where you use some certain
> representations to show some level of analysis. Spoken/signed (as the
> primary code) or written (as the secondary code) language always belong to
> the object area. For the descriptional area you have to define the use of
> your descriptional elements in a way we need not do for elements used in the
> object area.
>
> Best
>
> Franz
_______________________________________________
SLLING-L mailing list
SLLING-L at majordomo.valenciacc.edu
http://majordomo.valenciacc.edu/mailman/listinfo/slling-l


_______________________________________________
SLLING-L mailing list
SLLING-L at majordomo.valenciacc.edu
http://majordomo.valenciacc.edu/mailman/listinfo/slling-l