Phonology Theories: Autosegmental to Optimality Theory | Lecture

Added:

Exemplar Theory and Usage-Based Phonology

While Autosegmental Phonology and Optimality Theory (OT) represent a continuous lineage within the formalist, generative tradition, Exemplar Theory and Usage-Based Phonology offer a fundamentally different, functionalist alternative. Generative theories assume that phonological knowledge consists of abstract, symbolic, and categorical representations governed by rules or universal, innate constraints. In contrast, Exemplar Theory posits that phonological representations are not abstract, but are composed of highly detailed, concrete memory traces (exemplars) of actual speech experiences. Under this view, phonological patterns and structures emerge dynamically from frequency of use, phonetic detail, and general cognitive categorization mechanisms, rather than from a pre-determined, universal grammar. This perspective challenges the necessity of abstract underlying representations (central to Autosegmental Theory) and universal, violable constraints (central to OT), arguing instead that language is shaped by usage, motor control, and auditory perception.

A foundational understanding of basic phonetics and phonology, including the International Phonetic Alphabet (IPA), phonemes, allophones, and distinctive features.
Familiarity with classical rule-based generative phonology, particularly the framework established in Chomsky and Halle's 'The Sound Pattern of English' (SPE).
Basic comprehension of syllable structure, including the division of syllables into onset, nucleus, and coda, and how these structures interact with prosody.
Practical application of Optimality Theory (OT) mechanics, including constructing constraint tableaux and determining strict dominance hierarchies (GEN, CON, and EVAL).
Analysis of tonal languages, vowel harmony, and nasal spreading using multi-tiered Autosegmental representations.
Exploration of advanced and alternative constraint-based frameworks, such as Harmonic Grammar, Stochastic OT, and MaxEnt (Maximum Entropy) grammar models.
Investigation of phonological acquisition and learnability, specifically how children acquire constraint rankings using algorithms like the Gradual Learning Algorithm (GLA).
119 views3likes2:27:51@LinguisticsIslandOriginal Release: 2025-08-14

Autosegmental phonology, proposed by Goldsmith, represents suprasegmental features like tone on separate tiers from segments, addressing limitations of generative phonology which bundles all features together. This theory explains phenomena such as tone stability (tones surviving segment deletion), floating tones (tones without underlying segments), and toneless segments (segments without underlying tones) through multi-tier representation. The theory employs the Obligatory Contour Principle (OCP) to prevent adjacent identical autosegments and uses association conventions (mapping, dumping, spreading) to link tones to segments, though it faces limitations in explaining long-distance assimilation.