Abstract

This study analyzes the prosodic structure of the Myanmar language using Metrical theory to illustrate all possible deep structures. The findings reveal distinct differences in mora placement, particularly in duration and intensity. The average vowel duration is 33 ms for a one-mora (µ) syllable and 165 ms on average for a two-mora (µµ) syllable. The average vowel intensity is 48 dB for a one-mora (µ) syllable and 54 dB on average for a two-mora (µµ) syllable. Both values indicate that the weight of Myanmar moraicity depends on the syllable type, which has two categories: light (µ) and heavy (µµ). Regarding syllable length and vowel duration, the equality of weight represented by a mora is proposed as σ(μi(V)) = σ(μi(VV)). The metrical analysis reveals the results on the prosodic structure that syllable weight is distributed to the diphthongal vowel and coda under possible moraic constraints. Seven patterns are identified: [Cəμ]σ, [CVμμ]σ, [CVμVμ]σ, [CVμCμ]σ, [CVμVCμ]σ, [CVμCμ]σ, and [CVμVμC]σ. The moraic weights change according to vowel diphthongization and coda presence. The stress assignment in Myanmar prosodic structures includes both syllabic and moraic trochees. This study provides a comprehensive analysis of Myanmar’s prosodic structure, emphasizing the strengths of metrical frameworks. Metrical theory integrates acoustic data to examine moraic weight effectively, whereas Optimality Theory (OT) does not incorporate these aspects. The prosodic structure has significant implications for further language technology development. It enables the advancement of prosodic tagging through language data, allowing machine translation (MT), speech recognition (ASR), and text-to-speech (TTS) systems to recognize prosodic features.