Add capability to set phonological features for a segment based on environment

[ddaspit]

There are cases where a user might want to treat a consonant as a vowel or vice-versa in a certain phonological environment. A UI for this is probably not currently necessary, since it is a rare use case.