the vowel /a/

[Daniel]

Can you give an example of both? Are you saying you'd pick the feature that is most likely to be positive? Or the one that is most likely to be negative?

[MalFet]

The convention is to use the label of the "marked" feature. What exactly that means varies from theory to theory, of course, but at minimum it implies some kind of (usually phonotactic though at least perceptual) salience. I'm not sure exactly what you mean by "most likely" to be positive or negative. Are you talking about corpus frequency distributions? That's not really a mainstream analysis.

As far as examples go, I'd have to look them up. Off the top of my head, I believe several of the Athabaskan languages with small vowel inventories use a binary distinction with [front] marked.

[Daniel]

Thanks. I hadn't come across it that way before.

Of course it seems a bit odd to me to claim one is marked and that the other is not, but maybe there are complex reasons for this.


As for "most likely" I was referring to the distribution of vowels-- maybe 3 are back and 2 are front.