- In classical phonemics, phonemes and their possible combinations are central.
- In standard generative phonology, distinctive features are central. A stream of speech is portrayed as linear sequence of discrete sound-segments. Each segment is composed of simultaneously occurring features.
- In non-linear models of phonology, a stream of speech is represented as multidimensional, not simply as a linear sequence of sound segments. These non-linear models grew out of generative phonology:
o autosegmental phonology
o metrical phonology
o lexical phonology