Chapter 3 argued that the potential conflict between letter-based and speech-based identity in Old English alliteration should be resolved in favor of the linguistic principle, an approach which in its turn allowed a re-evaluation of the segmental histories of the velars. Proceeding from the same premise, the primacy of the linguistic signal over convention and the written medium, I now turn to the evidence and the arguments for reconstructing the status of the stressed syllable onset in the history of English. In Modern English the realization of a filled Onset in stressed syllables is optional and can be subject to segmental and prosodic variability. Unlike the voluminous literature on segmental quality and quantity, the empirical and theoretical coverage of Onset is quite limited. The presence or absence of a glottal stop in vowel-initial stressed syllables in Old English has been the only issue debated in the literature, with no single position becoming generally accepted. There has been no comparable interest in the Middle English evidence for vowel-initial syllables, nor has the diachronic picture ever been compared to the Modern English data on Onset. A further question that needs to be addressed is the probability of hybridization of the native prosodic patterns with the typologically different Romance syllable structure, parallel to the much discussed hybridization of the stress system. These are the issues that chapter 4 will examine.

The vowel alliteration riddle

Throughout Germanic verse, vowels alliterated with each other promiscuously, “Alle Vokale stoben durcheinander” (Heusler 1925: 95).