Constraints on Homorganic Cluster Lengthening in Early English: A Unified Account of Vowel Lengthening and Shortening

Abstract

Homoragnic Cluster Lengthening (HCL) has often been trea ted as exceptional or problematic in a holistic analysis of the quantitative changes of vowels in Middle English (ME). This paper aims to explore what constraints actively participate in HCL and to reconsider whether or not HCL is really incompatible with other quantitative changes of vowels, especially Pre-Cluster Shortening (PCS). This study starts with the assumption that the difference between HCL and PCS goes back to the different combination of the consonant clusters preceding a vowel. It is proposed that HCL is a consequence of the phonetic tendency that vowels and sonorants are longer before a voiced obstruent than before a voiceless one. The fact that only vowels are perceived as phonernically long in ME is captured by a conjoined constraint. Unlike previous analyses, it is shown that in an Optimality-theoretic analysis many causal factors are not mutually exclusive, but ranked with one another, and it is also demonstrated that HCL and the seemingly contradictory change, PCS, can be unified under the same hierarchy of constraint.