Editing Autosegmental phonology (section)

=== Distinctive features ===
{{Main|Distinctive features}}
Rather than classify segments using the categories given in the [[International Phonetic Alphabet]], the autosegmental formalism makes use of distinctive features, which provide greater granularity and make identification of [[natural classes]] easier. A segment is identified by a +/− dichotomy of a series of binary features, some of which are subfeatures of unary features ([[place of articulation]], in particular, is identified by unary features indicating the active articulator, and binary subfeatures that distinguish further). For example, [p], the voiceless bilabial stop, is indicated [−sonorant, −continuant, −voice, labial], and the set of voiced coronal stops can be indicated [−sonorant, −continuant, +voice, coronal].
[[File:Dependent features.svg|thumb|175px|An alveolar segment using dependent features]]

==== Feature dependencies ====
For unary features to be fully specified, it is necessary to include binary subfeatures that correspond to them. In the autosegmental formalism, this is depicted by placing the binary subfeature at a horizontal offset from the unary feature and connecting them with a line. The next top-level feature in the segment would then be connected to the unary feature as well as opposed to the tone.