COMPARISON continued...

2. Hand Shape
Hand Shape most clearly shows the difference between the schematics of SSW and the taxonomic approach of SN. The traditional taxonomic approach establishes a limited number of distinct shapes and gives each one a symbol. To write a Sign one looks at the whole hand shape, matches it with whichever one of the set it resembles most, and gives it that symbol. In contrast, the schematic approach used by SSW doesn't care how many shapes there are, as it represents each little part of the hand independently. For a new shape, SSW merely reproduces it, while SN is forced to put it with one of the existing symbols, even if it doesn't quite match.

For the Hand Shape parameter SN chose arbitrary shapes, the letters of the ASL manual alphabet and number system. This is natural enough and serves a mnemonic function but of course is a problem when applied to languages other than ASL. The hand shape used in ASL for <F> is used to represent <T> in Dutch Sign, in Sweden the hand shape for <B> is that of ASL's <A>, and in many cultures the ASL <F> is an obscene gesture (Miller 199). What's more, some signed languages have no manual alphabet at all (Chinese), or completely different ones (England, Australia). For them these letters are truly arbitrary, and of course, for children just learning to read, they are a burdensome task to memorize. The schematic alternative of SSW eliminates all these problems (Figure 9a).

Figure 9a:
SSW Hand Shapes are schematic diagrams

SN kept the number of hand shapes down to 19 by such devices as treating the <A>, <S>, and <T> as one hand shape. Stokoe has been criticized for this as these three are clearly distinctive, and others have found many more distinctive hand shapes in ASL: Battison found 45, Klima about 40, Newkirk 54 (Valli & Lucas 185), and Johnson and Liddel 150 (Liddell & Johnson 223). I think Stokoe merely erred on the side of a popular script. "A hundred and fifty letters is more than we'd want," was his apparent assumption.

In SSW the number of hand shapes is effectively the number of positions the hand can take. The symbols are really 2-3 line segments per digit that join each other and the palm straight or at angles depending on the positions of the bones. The many thousands of possible combinations are organized into ten groups that correspond to the ASL numbers 1-10, and within each group the fingers can be crossed in four or five ways, the thumb bent out or straight in either of three planes and so on (Figure 9b).

Figure 9b:
SSW has a different symbol for each hand position

How much phonetic detail can these scripts show? As we have seen the original SN has proven inadequate in this respect, and I think Wilbur's criticism still holds for later versions. Although there is virtually no published discussion of SSW, it fares much better, at least as regards hand shape. Of the many feature analyses of ASL, perhaps the most ambitious is the work of Johnson and Liddell at Gallaudet University. Although their study deals in detail only with Hand Shape, SSW seems able to show every phonetic contrast they describe.