Is there a cross-linguistic subdivision of phones in signed languages akin to how all spoken languages have vowel and consonant phones?

Question

After reading Joe Martin's enlightening answer to the question "Are there counterparts to phones and phonetics for signed languages?" I immediately began to wonder how much further spoken and signed phones' similarities might go.

For instance in all spoken languages the phones can be divided into vowels and consonants (plus the small category of semivowels).

Further all spoken languages' vowels can be analysed various ways such as monophthong / diphthong / triphthong; high / low + front / back + rounded or not + some other features. Consonants are usually graphed with the major features of "place of articulation" on the X-axis and "manner of articulation" on the Y-axis, and other features being treated in various ways.

So my question is Are there any generally accepted counterparts to these systems that are applied to all signed languages generally?

(And don't worry, I'm not asking whether sign languages have "voiced" and "unvoiced", or any other exact counterpart to any specific aspect of spoken languages, if signed languages are divided into one or five parts where spoken languages are broken into the two of vowel and consonant is just as interesting.)

Answer 1

One idea that's been going around at least since the 80s is that you can distinguish between Holds and Moves. For instance, many signs have a Hold-Move-Hold structure -- you start with one handshape at one spot in the signing space, move your hand along a designated path, and end up with another handshape in another spot. It's been suggested that a Hold-Move-Hold sign is akin to a Consonant-Vowel-Consonant syllable in spoken language.

Holds and Moves can also be described using sets of distinctive features, just as consonants and vowels are in spoken languages. For instance, every Hold has a handshape component. A signed language has a fixed inventory of handshapes (here's the set for ASL), and you can find minimal pairs for handshape in signed languages just like you can find minimal pairs for voicing or point-of-articulation in spoken languages. For signs where the active hand touches your other hand or your body, there's an inventory of possible points of contact. And so on.

Just like in spoken languages, different signed languages make use of different feature combinations. F'rinstance, a friend of mine who works on this stuff tells me that in Ban Khor Sign Language, spoken in Thailand, there are signs in which you touch your face with the pinky side of a flat hand (what's called the B handshape). In ASL, the B hand only ever makes contact with your face on the thumb-and-index-finger side. So there are phonological patterns that are possible in Ban Khor Sign Language and not in ASL, and probably vice versa.

You might be interested in Liddell and Johnson's "ASL: The Phonological Base" (you can read some good chunks of it here at Google Books), which goes into some detail on this stuff: the Hold/Move structure of signs, the features that constitute Holds and Moves, etc. I don't think it's the state of the art anymore -- if there are any experts out there reading this, I hope they'll suggest something more current -- but it's a pretty cool, very readable article, and it gives you a feel for what a careful description of signed phonology might look like.

Answer 2

All the same methods of analysis that are used with spoken languages apply successfully to signed languages. Most work with sign is at a subphonemic level, in keeping with modern trends toward nonlinear phonology. The analysis of a HMH sign as one syllable utilizes a sonority scale based on perceptual salience, and equates Movements with vowels and Holds with consonants. Other models of visual phonology utilize autosegments, with separate tiers for configuration features { +/- backed tongue root, raised apex, bent finger, palm upward ... } and for Location / Place features { alveolar ridge, teeth, lip, nose, forehead … } each tier operating independently. Brentari, D. (1998). A Prosodic Model of Sign Language Phonology. MIT Press, outlines different models of sign phonology starting on page 83, and discusses syllables from page 71.

In the Movement-Hold Model, the phoneme /t/ would be a Hold-Movement-Hold structure, with beginning and ending positions of the tongue joined by a Movement. The auditory signature of a stop consonant is dead silence: what's perceived, via changes in formant frequency, is the movement of the articulator, either upward [eat] or downwards [tea]. As it moves, the articulator has to be in some configuration, and since this tongue-shape occurs simultaneously with the movement component, we get acoustically different /t/s if we change it. This co-articulation is the difference between dark L and clear L, or between two signs with different hand-shapes.
Note: A big difference in the two media is that no morphemes show up in speech at this level of detail, but a HMH sign often includes several morphemes, maybe an entire sentence.