Hyla,

Not all classifiers can be used to describe things; that is to say, they are not general base forms to be applied to everything. Using the stone column example, I could say that -li and -la are only applicable to inantimate objects, and that if we were talking about animate objects, it would have to be -bi and -ba.

The classifier used in the CAR example is the classifier for any vehicle. It isn't polysynthetic because the signer uses the handshape and motion very distinctly different when it is "classifying" compared to merely being used as a sign.

You could classify something as being flat and smooth, but logically, that classifier would only apply to physically existing objects that could fit into that category (sheet of paper, calm waters, Pergo flooring, and even a metaphorically-used poker face).

When signing or classifing CAR, you could signify a long car by using a specific arm movement (farther from the body), puckered lips, and a head tilt to the side. You could convey a short car by holding the sign or classifier in tight to the body, keeping your head pulled up and back, and piercing the lips together tightly.

My apologies for the longwindedness. In all the on-line discussions I've had with others (including graduate classes in ASL linguistics), we always struggle with written forms of conveying such intricate visual detail. I'm still waiting for the digital video revolution...

Brandon